初创公司尽调
尽调报告 Healthcare / medical devices / neural engineering Clinical-stage private company 2026-06-28

Science Corporation

临床阶段神经接口公司,拥有 NEJM 背书的视觉修复植入物,并率先打通 BCI 商业化路径

Science Corporation 拿到了植入式视觉恢复领域迄今最强的公开临床证据;但在 CE Mark 结果和欧盟商业牵引得到确认前,$1.5B 入场估值跑在期望价值前面。

封面要素

PRIMA 关键试验成功 03
80% patients with meaningful visual acuity gain [CO027]
CE 认证申请 04
Submitted June 2025 [CO031]
PRIMAvera 试验入组 05
38 patients at 17 sites (EU + US) [CO025]
员工人数 06
~150 employees (analyst estimate, low confidence) [CO034]
Series C 完成 07
March 2026 $230M, oversubscribed [CO019, CV001]
FDA 突破性设备认定 08
Granted 2023 designation (inherited via Pixium acquisition) [CO024]

公司概况

Science Corporation 是一家临床阶段神经工程公司,2021 年 7 月由 Max Hodak(Neuralink 前联合创始人兼总裁)和 Alan Mardinly 博士(Harvard 神经生物学博士、Neuralink 前研究负责人)创立,总部位于加州 Alameda。公司近期商业化主线是 PRIMA:一款面向年龄相关性黄斑变性导致的晚期地图样萎缩的光伏视网膜下假体。2025 年 10 月,NEJM 发表的 PRIMAvera 关键试验显示,80% 受试者在 12 个月时取得有意义的视力改善。Science 已于 2025 年 6 月提交 CE 认证和 FDA 申请,目标是在 2026 年启动欧盟商业化,成为第一家把视觉修复产品推向市场的 BCI 公司。公司累计融资约 $490M,其中包括 2026 年 3 月以约 $1.5B 投后估值完成、超额认购的 $230M Series C。除 PRIMA 外,Science 还在开发生物混合神经接口架构、Science Eye 光遗传学平台,并在北卡罗来纳州 Research Triangle Park 运营自有 MEMS 晶圆厂。

官网
science.xyz
成立时间
2021-07-01
创始人
Max Hodak, Alan Mardinly, PhD
创立地点
Alameda, California, USA
总部
Alameda, California
产品
PRIMA 光伏视网膜下假体(2×2 mm、厚 30 µm),用于地图样萎缩 / AMD,配合带红外投影仪的专用眼镜和 AI 口袋处理器使用;Science Eye 光遗传学视觉修复平台;搭载工程化神经元的生物混合神经接口芯片;自有 MEMS 晶圆厂,对外提供第三方 BCI 制造服务。
客户
AMD 导致的晚期地图样萎缩患者,其次是视网膜色素变性和 Stargardt 病患者;欧盟和美国市场的视网膜外科医生与眼科中心。
商业模式
临床阶段、尚未商业化的医疗科技公司;计划在 CE 认证放行后于欧盟销售设备(ASP 估计为每台 €80K–€90K),放量取决于报销;晶圆厂服务作为次要收入;美国上市取决于 FDA PMA 和 CMS RAPID 覆盖路径。
阶段
Clinical-stage private company (pre-revenue)
融资情况
2026 年 3 月完成 $230M Series C,投后估值约 $1.5B;自 2021 年以来,Series A、Series B、2025 年 4 月 Khosla 领投的 $104M 可转债和 Series C 合计融资约 $490M;所有 Series C 投资者均为老股东。
[CO001, CO002, CO003, CO019, CO021, CO022, CO025, CO027]

执行摘要

主要优势

  • PRIMAvera 枢纽试验发表于 NEJM,覆盖 17 个中心、38 名患者,患者成功率 80%——这是植入式视觉恢复设备迄今公开发表的最高质量临床证据。
  • 商业化先发位置明确:CE Mark 申请正在审评,可能在 2026 年中作出决定;若获批,Science Corp 将成为全球第一家把产品推向市场的 BCI 公司。
  • FDA Breakthrough Device 认定叠加 2026 年 4 月 RAPID Coverage Pathway,形成 FDA 批准与 Medicare 覆盖同步推进的机制,显著压缩历史上的获批后报销缺口。
  • 创始人与运营团队有经验:Hodak、Mardinly 带来 Neuralink 背景,监管终局前又加入来自 Butterfly Network、具备商业 medtech 深度的 Shahida。
  • Research Triangle Park 的自有 MEMS foundry 带来制造差异化,降低供应链风险,也创造此前视网膜假体公司没有的平台可选性。
  • 资本储备充足(累计融资 $490M,$230M Series C 超额认购),公司称现金跑道足以支撑到 CE Mark 商业发布,短期没有立刻稀释压力。

主要风险

  • 三家已获 CE 标志的前代视网膜假体公司——Second Sight(Argus II)、Pixium Vision(PRIMA)和 Retina Implant AG——都拿到监管许可却商业失败,行业先验是 0/3。
  • EU MDR 下 CE Mark 决定时间不确定;若被拒或 RFI 把审评拖过 2027 年 Q2,将实质打击商业化时间线和投资者信心。
  • 报销基础设施缺位:Anthem 2026 年政策把所有人工视网膜设备列为 investigational;欧盟关键市场按国别推进 HTA,通常在 CE Mark 后还要 12–36 个月。
  • 公司收入为零,股权结构优先权堆叠、可转债转换条款和清算瀑布均未披露,真实投资者经济性不透明;$1.5B post-money 标记可能高估普通股价值。
  • 关键人集中在 CEO Max Hodak 身上:科学愿景、投资人关系、监管策略和主要公开发言人角色都由他同时承担。
  • 平台化分心风险存在:biohybrid 神经接口、Science Eye optogenetic 项目和 foundry 建设并行推进,在需要集中执行 PRIMA 的阶段分散管理层注意力和资本。

未决问题

  • BSI 的 CE Mark 结果,以及德国 / 欧盟首次商业发布时点,是近期信息量最高的催化剂;截至 2026 年 6 月状态仍未确认。
  • 股权结构、优先权堆叠深度、参与权和清算瀑布未公开披露;Series C close 时可转债转换机制仍未知。
  • 美国 FDA PMA 决定仍在等待;与 FDA/CMS 就 RAPID Coverage Pathway 互动的时间线未公开确认;FDA 原始记录中的 Breakthrough Device 认定应独立核验。
  • 欧盟发布的定价和报销条款未披露;德国首发市场准入模型、ASP 和手术报销代码可用性都是关键未知项。
  • 12 个月以上 PRIMA 长期安全性和耐久性数据尚未发表;仍需要 36 个月随访,以及植入眼与对侧眼地理萎缩进展数据。
  • 员工数(约 150 人)来自单一分析师估计;官方员工数和组织结构未公开披露。

目录

Chapter 01

01公司概览

1.1 身份、使命与创立

Science Corporation 是一家临床阶段神经工程公司,2021 年注册成立,总部位于加州 Alameda。公司由 Max Hodak 和 Alan Mardinly 博士联合创立:Hodak 曾联合创立 Neuralink,并在 2021 年 5 月前担任总裁;Mardinly 拥有 Harvard 神经生物学博士学位,曾任 Neuralink 研究负责人。公司的底层判断是:大脑是人体唯一不能移植的器官,也是医学的终极前沿。Science 近期商业化载体是 PRIMA,一款光伏视网膜下脑机接口(BCI)视网膜假体,面向年龄相关性黄斑变性导致的晚期地图样萎缩(GA)。除 PRIMA 外,公司还在同步推进生物混合神经接口架构、Science Eye 光遗传学平台和位于北卡罗来纳州 Research Triangle Park 的 MEMS 半导体晶圆厂:前者试图让工程化神经元在芯片上生长,以提升大脑整合效果;Science Eye 面向更广泛的视觉修复;晶圆厂则提供自有制造能力。创始团队选择保持私有、商业化前状态,一边搭建多模态神经接口组合,一边围绕监管里程碑安排外部产品发布。公司对外始终把使命表述为直接接入大脑这个信息处理器官,以取得传统医学无法达到的效应量。 [CO001, CO002, CO003, CO004, CO005, CO014]

快照 KPI 表
指标数值 / 状态截至日期置信度缺口 / 尽调路径
成立日期2021 年 7 月2021-07-01多份公司官方沟通已确认
总部美国加利福尼亚州 Alameda2026-06-28每篇主要新闻稿和新闻报道均确认
累计融资~$490M2026-03-05来自公司 Series C 公告;各轮构成金额为近似值
Series C 后隐含估值~$1.5B2026-03-05分析师来源报道;公司未官方确认精确数字
员工数约 150 名员工2026-03-05单一分析师估计;公司尚未正式披露员工数
收入商业化前;$02026-06-28截至 2026 年中,尚无产品上市,也未获得监管批准

估值为分析师报道的隐含数字;累计融资由公司披露,各轮构成金额来自第三方数据库。员工数来自单一分析师来源,应视为近似值。

[CO001, CO019, CO021, CO022, CO034, CO040]
FO002: 公司快照逻辑

Science Corporation 的身份、产品平台、资本基础和技术来源如何连接成神经工程价值链。

[CO001, CO010, CO011, CO013, CO019]

1.2 PRIMA 技术与产品平台

PRIMA 是一款光伏视网膜下植入物,占地约 2 × 2 mm,厚度仅 30 µm,约为人类头发直径的一半。设备相当于人工感光细胞:这枚微型无线芯片植入视网膜下、位于感光细胞层,能把投射到芯片上的近红外光转化为对下方双极细胞的局部电刺激,在感光细胞已被地图样萎缩破坏的患者体内重新引入视网膜信号通路。PRIMA 系统由三部分组成:视网膜下植入物、内置摄像头和红外投影仪的专用眼镜,以及由 AI 驱动的口袋处理器。处理器捕捉视觉场景、识别重要物体,并把图案化红外光导向植入物。植入物由投射光无线供电,不需要内置电池——这是关键的安全性和寿命优势。光伏视网膜刺激技术最初由 Stanford University 的 Daniel Palanker 教授实验室开发。2024 年 4 月,Pixium Vision SA 在 Paris Commercial Court 进入清算程序后,Science 收购了全部 PRIMA 知识产权和三项进行中的临床试验。收购的试验包包括 PRIMAvera 关键研究(NCT04676854)和两项可行性试验。Science 现在正把 PRIMA 临床项目扩展到 Stargardt 病和视网膜色素变性,这两类疾病是年轻成人遗传性视网膜视力丧失的主要原因。目前 PRIMA 假体视觉为黑白、分辨率相对较低;Science 已披露,下一代设备将靠数字图像处理优化视觉表现。 [CO006, CO007, CO008, CO009, CO010, CO011]

1.3 领导团队与治理

Science 的高管团队由具备 Neuralink 背景的联合创始人领衔,并在 PRIMA 监管终局前补入商业化人才。Max Hodak 是 CEO 兼联合创始人,曾联合创立 Neuralink 并担任总裁,2021 年 5 月离开后创立 Science Corporation。Alan Mardinly 博士是联合创始人兼首席科学官(2025 年 12 月任命),拥有 Harvard 神经生物学博士学位和超过 15 年神经科学研究经验,此前曾在 Neuralink 担任领导职务。Darius Shahida 于 2025 年 12 月获任首席战略官,此前曾帮助 Butterfly Network 融资超过 $1B,带来可直接迁移的大型医疗科技商业化经验。Frank Brodie 医学博士于 2025 年 12 月获任 Vision 与 PRIMA 医学总监,负责美国临床和监管路径。Ralf Hornig 博士负责临床事务,曾有欧洲 BCI 项目经验。Matthias Wagner 负责生物混合设备开发。董事会构成尚未公开;CEO 职位的正式继任计划或关键人物保险也未获公开确认。关键人物集中度最高的是 Max Hodak,他同时掌握科学愿景、投资人关系、监管策略和主要媒体发言人角色。 [CO002, CO003, CO023, CO036, CO043, CO044]

领导层与创始人表
姓名职位背景创始人 / 外部关键人物说明
Max HodakCEO、联合创始人Neuralink 联合创始人兼总裁(2016-2021);Stanford SEAS 背景创始人关键:主要公众面孔、投资人关系、战略与科学愿景
Alan Mardinly, PhDCSO、联合创始人(2025 年 12 月任命)Harvard 神经生物学博士;前 Neuralink 神经科学研究负责人;University of Chicago 学士创始人高:负责生物混合平台和长期 BCI 路线图的科学愿景
Darius Shahida首席战略官(2025 年 12 月任命)在 Butterfly Network 融资 >$1B;拥有深厚 medtech 商业化经验外部高:负责 PRIMA 上市的商业化战略和投资人沟通
Frank Brodie, MD医疗总监,Vision & PRIMA(2025 年 12 月任命)眼科 / 视网膜临床开发领导经验外部中:负责 PRIMA 的美国临床和 FDA 监管路径
Ralf Hornig, PhD临床事务总监欧洲 BCI 临床项目经验;此前从事视网膜植入体工作外部中:负责 CE mark 流程和欧洲关键试验运营
Matthias Wagner生物混合器械总监下一代生物混合神经接口平台的工程领导经验外部中:负责生物混合架构管线开发

董事会成员未公开命名;表格仅反映截至 2026 年 6 月已公开披露的高管。2025 年 12 月任命日期来自分析师报道。

[CO002, CO003, CO023, CO044, CO045]

1.4 融资历史、投资者与估值

Science Corporation 自 2021 年创立以来,已通过四次已知融资事件累计融资约 $490M。公司于 2021 年 7 月完成约 $47.3M Series A,随后在 2022 年 11 月完成约 $113M Series B。2025 年 4 月前后,公司完成 Khosla Ventures 领投的 $104M 可转债,时间点很可能是为衔接 NEJM 发表和 CE 认证提交里程碑。旗舰融资是 2026 年 3 月完成的 $230M Series C,参投方包括 Khosla Ventures、Lightspeed Venture Partners、Y Combinator、IQT(美国情报界的风险投资机构)和 Quiet Capital——均为既有投资者;该轮超额认购,需求超过资金需求。Series C 对应约 $1.5B 投后估值,使 Science 成为全球同阶段资本最充足的私有 BCI 公司之一。资金用途包括 PRIMA 欧洲商业化、制造放量,以及推进生物混合神经接口和 Vessel 灌注平台等管线项目。公司公开表示,其核心任务是“成为第一家规模化并实现盈利的 BCI 公司”。Series A 和 B 的完整投资者名单仍未披露。 [CO016, CO017, CO018, CO019, CO020, CO021]

利益相关方或投资人地图
投资人类型轮次已知金额(USD M)尽调问题
Khosla VenturesTechBio / Deep Tech VC(Menlo Park,投资方)Series A、可转债(约 2025 年 4 月,领投)、Series C~$104M(可转债领投);Series A 和 C 金额未披露确认董事席位;核验总持股比例和 pro-rata 权利
Lightspeed Venture PartnersGrowth / TechBio VCSeries C(新进入者)未披露首次参与 C 轮;确认治理权和任何董事会观察员角色
Y Combinator加速器 / 早期 VCSeries C 参与方未披露厘清持续顾问关系与仅持股关系;确认首笔投资规模
IQT (In-Q-Tel)美国情报界风险投资机构Series C未披露厘清任何 IP 许可、数据共享或国家安全义务
Quiet CapitalVC 基金Series C未披露识别管理合伙人;核验基金策略和持股经济性
未披露早期轮投资人多类Series A ($47.3M)、Series B ($113M) 融资~$160M 合计(A + B);个别投资人未公开命名获取完整股权结构;确认 a16z Bio 或其他已点名基金是否参与

所有金额均按新闻稿或分析师数据库报道;精确轮次经济条款为公司机密。可转债日期基于 Crunchbase 报道的近似表述(相对于 2026 年 3 月 Series C 交割的「last April」)。

[CO016, CO017, CO018, CO019, CO020, CO021]
FO003: 快照 KPI

截至 2026 年中,Science Corporation 的关键表现和成熟度指标快照,覆盖资本结构、临床准备度、监管阶段和技术规模。

估值和员工人数为分析师估计;不良事件率采用 FierceBiotech 报道中的可评估患者分母(32)。

[CO019, CO022, CO027, CO030, CO031, CO034]

1.5 临床证据、监管状态与风险画像

PRIMAvera 关键试验(NCT04676854)在法国、德国、意大利、荷兰、西班牙和英国的 17 个中心入组 38 名患者。所有受试者均患有晚期地图样萎缩,且无残余中心视力。2025 年 10 月 20 日发表于 The New England Journal of Medicine 的结果由 Frank Holz(University Hospital of Bonn,牵头)、José-Alain Sahel(University of Pittsburgh)及 14 名共同作者撰写,显示:80% 受试者在 12 个月时取得有意义的视力改善;84.4% 恢复了阅读字母、数字和单词的能力;ETDRS 视力表平均提升约 25.5 个字母(超过五行)。数据安全监测委员会认为,收益超过手术风险。研究未观察到自然周边视力的显著损失。19 名患者出现不良事件,包括眼压升高、视网膜撕裂和出血;多数在两个月内缓解。 Science 于 2025 年 6 月向欧盟提交 CE 认证申请,并于同月申请美国 FDA 监管批准;公司还持有 FDA 突破性设备认定,该认定最初于 2023 年 4 月授予 Pixium,并通过 2024 年 4 月收购承继。欧洲商业化预计在 2026 年启动,若成行,Science 将成为第一家拥有在售视觉修复产品的 BCI 公司。截至 2026 年中,美国 FDA 决定仍待定。 重大风险包括:PRIMA 目前只能提供黑白、相对低分辨率的假体视觉,而非自然彩色视觉;价格和报销条款未披露;38 人试验规模对于更广泛的适应症标签扩展可能显得有限;12 个月以后的长期安全数据仍在收集;公司仍完全没有收入,在监管和商业化启动过程中依赖投资者持续供血。 [CO024, CO025, CO026, CO027, CO028, CO029]

里程碑表
日期里程碑类别金额 / 状态重要性
2021-07Science Corporation 由 Max Hodak 和 Alan Mardinly 在 Alameda, CA 创立创立Neuralink 之后启动;最初聚焦 BCI 和视觉恢复平台
2021-07Series A:融资 ~$47.3M融资$47.3M种子阶段资本,用于启动 PRIMA 和 Science Eye 研发
2022-11Series B:融资 ~$113M融资$113M扩大研发;推出 Science Eye 平台;为 PRIMA 收购打基础
2023-04FDA 为 PRIMA 授予 Breakthrough Device designation(经 Pixium Vision)监管已授予认定FDA 加速审评路径;2024 年 4 月收购后状态转移至 Science
2024-04Paris Commercial Court 批准收购 Pixium Vision PRIMA 资产产品未披露转移 PRIMA IP 和三项临床试验,包括 PRIMAvera (NCT04676854)
2025-04完成 $104M 可转债,由 Khosla Ventures 领投融资$104M商业化前过桥融资;时点在 CE Mark 递交和 NEJM 发表前
2025-06向欧盟提交 CE Mark 申请;在美国提交 FDA 申请监管已提交申请PRIMA 迈向欧洲和美国商业市场许可的第一步
2025-10-20PRIMAvera 试验结果发表于 The New England Journal of Medicine产品 / 临床80% 患者改善;84.4% 能读字母BCI 视觉恢复试验首次发表于 NEJM;主要作者将其定义为范式转变
2025-12高管任命:Mardinly(CSO)、Brodie(医疗总监)、Shahida(CSO)治理三名关键招聘首款产品上市前搭建商业阶段领导层
2026-03完成 Series C $230M;累计融资 ~$490M;隐含估值 ~$1.5B融资$230M迄今 BCI 视觉恢复领域最大融资;支持 PRIMA 欧盟上市和管线

里程碑日期按公开报道。2025 年 4 月可转债日期为近似值,来自 Crunchbase 在 2026 年 3 月报道中提到的「last April」。未公开披露的董事会层面治理事件和内部里程碑已排除。

[CO011, CO016, CO017, CO018, CO019, CO024]
FO001: 公司里程碑时间线

Science Corporation 从 2021 年创立到 2026 年 3 月完成 Series C 的里程碑时间线,覆盖融资、产品、监管和治理事件。

里程碑日期来自官方新闻稿;2025 年 4 月可转债日期依据第三方报道,属近似值。

[CO001, CO011, CO019, CO024, CO026, CO031]
Chapter 02

02市场分析

2.1 市场边界与定义

Science Corporation 的 PRIMA 系统对应的是一个边界清晰的市场,区别于更宽泛的眼科、AMD 药物治疗和通用 BCI 版图。纳入测算的支出是用于恢复终末期退行性视网膜疾病视觉功能的外科设备系统,具体包括光伏视网膜下植入物、配套外部硬件和植入手术。不纳入的是 AMD 药物治疗市场(抗 VEGF 注射、补体抑制剂、AREDS 补充剂)、光学矫正、低视力辅助工具、基因疗法,以及完全绕过视网膜的皮层视觉假体。 晚期地图样萎缩的现状对照组是观察随访、没有修复选项:目前没有药物疗法能恢复 GA 失去的中心视力,因此 PRIMA 的临床对照是自然病程,而不是竞争药物。湿性 AMD 药物治疗(抗 VEGF)相邻但正交——抗 VEGF 注射每年在美国产生超过 600 万例操作,但面向的是根本不同的疾病阶段和机制;进展到终末期 GA 的患者已经错过抗 VEGF 的获益窗口。PRIMA 最接近的相邻领域是其他面向 RP 和 Stargardt 的视网膜植入路径、针对遗传性视网膜疾病的新兴基因疗法,以及为视网膜—视神经通路受损患者设计的皮层视觉假体。更大的 BCI 神经假体 TAM(Grand View Research 估算侵入式 BCI TAM 为 $168.27B)会抬高可触达市场叙事,PRIMA 专属测算应剔除,因为它覆盖运动功能 BCI、通信设备、神经调控和其他非眼科应用。AMD 药物市场(2025 年 $12.64B)同样应作为背景,而非 PRIMA 的市场,因为它针对的是另一个疾病阶段。 [CM025, CM043, CM024, CM038, CM040]

市场定义——纳入与排除支出
细分 / 类别纳入支出排除支出主要买方 / 支付方与 PRIMA 的相关性
视网膜假体(手术植入体)光伏 / 视网膜上植入器械、外部硬件、植入手术、术后编程Anti-VEGF 注射、激光光凝、光学辅助器具视网膜外科医生;Medicare / 商业保险核心市场——PRIMA 直接在此竞争
AMD 药物治疗Anti-VEGF 生物药(Eylea、Lucentis、Avastin、生物类似药)、补体抑制剂(SYFOVRE、IZERVAY)视网膜植入体、手术干预视网膜专科医生(门诊);Medicare Part B背景 / 可比——适应症正交(湿性 AMD vs. 终末期干性 AMD)
遗传性视网膜疾病(IRD)治疗基因治疗(FDA 批准的 RPE65 治疗)、低视力康复、视网膜假体黄斑变性药物IRD 专科医生、学术中心;Medicare / 商业保险相邻市场——RP 和 Stargardt 是 PRIMA 扩展适应症
视觉皮层假体(BCI)绕过视网膜的皮层植入体(如 Neuralink Blindsight、Orion)视网膜植入体神经外科医生;高度专科化中心;报销待定远端相邻——面向视神经受损患者,不是 PRIMA 适应症
更广义眼科市场白内障手术、屈光手术、诊断、青光眼器械、角膜移植仅视网膜假体普通眼科医生 / 外科医生;广泛支付方组合过宽——抬高市场叙事;与 PRIMA 规模测算无关

市场边界由作者按临床适应症和器械类别划定。「纳入支出」反映每个细分市场的相关商业机会。AMD 药物和更广义眼科市场仅作为背景参考纳入,并非 PRIMA 可服务市场。

[CM025, CM043, CM024, CM038, CM040]

2.2 疾病负担与可触达患者群

PRIMA 的主要目标适应症是年龄相关性黄斑变性(AMD)继发的晚期地图样萎缩(GA)。美国约有 100 万名 GA 患者,全球约 500–800 万名;美国每年新增 160,000 例。GA 对老年人影响尤其大:60–64 岁患病率为 0.28%,70–74 岁升至 0.98%;随着人口老龄化,全球 AMD 患病人数预计到 2050 年达到 2.88 亿。仅欧洲四大市场——德国(400,000–450,000)、法国(265,000)、意大利(170,000)和西班牙(155,000)——GA 就影响超过 100 万名额外患者;若能获得报销,将构成显著的欧洲商业机会。 除 GA 外,Science 已把 PRIMA 临床开发扩展到遗传性视网膜疾病(IRDs)。视网膜色素变性全球发病率约为 1/4,000,美国患者 80,000–110,000 人;约 12% 会发展为重度视力丧失(只能数手指或感知手动)的晚期疾病,进入植入候选层。Stargardt 病是最常见的青少年黄斑营养不良,全球约每 8,000–10,000 人中有 1 人受影响;发病更早,形成了不同且更年轻的手术候选人群。三大主要适应症合计带来约 120 万美国终末期视网膜退化患者池——但手术候选条件,包括残余视网膜功能、完整双极细胞层和手术健康阈值,会把可植入人群压到显著低于这些流行病学总量。GA 约占所有 AMD 相关法定失明的 20%,凸显 PRIMA 所切入病情层级的严重性。 [CM001, CM002, CM003, CM004, CM005, CM006]

FM004: 患者采用漏斗 — 从 AMD 人群到植入患者

分步漏斗说明庞大的 AMD 流行病学人群如何经过疾病严重程度、手术适应性、中心可及性和支付方覆盖逐层收窄, 最终抵达 PRIMA 近期商业窗口中的可植入人群。

第 3–5 步为作者估算,依据已发表的手术候选门槛、眼科手术数据中的中心容量代理指标, 以及 Argus II 采用中的报销先例。所有中间数值均属推测,需用 Science Corporation 的临床开发数据验证。

[CM028, CM030, CM016]

2.3 市场规模:多重口径与相互冲突的估计

视网膜假体的市场规模估计会随边界、方法和分析师显著变化,单一 TAM 数字容易误导。Mordor Intelligence 使用聚焦活跃设备单元的窄口径,估算全球视网膜植入物市场 2025 年为 $56.53M,2026 年增至 $61.28M,2031 年达到 $91.72M,CAGR 为 8.4%。Insight Partners 使用更宽的视网膜假体口径、纳入传统设备,报告 2025 年为 $392.53M,2034 年增至 $922.5M,CAGR 为 11.27%——同一市场和年份表面上出现 7× 差异。两家机构均未公开精确口径差异,这种不一致应视为所有已发布视网膜假体市场估计的局限。 作为背景,AMD 药物市场(抗 VEGF 药物、补体抑制剂)2025 年为 $12.64B,2026 年预计为 $13.58B,哪怕相比最宽口径的视网膜植入物估计也大约高 200×。更广义的整体 BCI 市场(Grand View Research:非侵入式 SOM 为 $397.59M,侵入式 TAM 为 $168.27B)远不止视觉应用,不应拿来替代 PRIMA。基于约 100 万美国 GA 患者、近期手术采用率 2–5%、每套系统平均售价 $50,000–$100,000 的粗略自下而上患者人口 SAM,推得美国潜在天花板 SAM 为 $1B–$5B——但这只是远期天花板;考虑到报销环境仍处早期,且获批后需要 3–5 年爬坡,近期商业相关性不高。已发布估计互相冲突,且缺乏经过验证的 PRIMA 专属测算,这是尽调应明确指出的重大证据缺口。 [CM009, CM010, CM011, CM012, CM013, CM036]

市场规模视角——按来源估算 TAM / SAM
发布方 / 来源基准年份地理范围市场价值(USD M)CAGR边界 / 方法置信度关键局限
Mordor Intelligence2025全球$56.5M (2025) → $61.3M (2026) → $91.7M (2031)8.4% (2026–2031)窄口径:仅主动式视网膜植入器械;专有估算框架低–中定义窄;排除皮层和更广义假体;无同行评审
Insight Partners2025全球$392.5M (2025) → $922.5M (2034)11.27% (2025–2034)宽口径:视网膜假体,包括传统器械(Argus II、Retina Implant Alpha)同一年比 Mordor 高 7×;范围不兼容;包含已停用器械线
Grand View Research(BCI 市场)2025全球非侵入式 SOM $397.6M;侵入式 TAM $168.3B8.73%(非侵入式,2026–2033)完整 BCI 市场,包括运动、通信、神经调控、视觉低(用于 PRIMA)绝大多数为非视网膜应用;不适合作 PRIMA 代理;抬高叙事
Polaris Market Research(AMD 药物市场)2025全球$12.64B (2025) → $13.58B (2026) → $25.18B (2034)8.0% (2026–2034)AMD 药物治疗(anti-VEGF、补体抑制剂、补充剂)高(用于 AMD 药物)与 PRIMA 适应症不同(湿性 AMD vs. 终末期 GA);不是视网膜植入体估算
作者自下而上 SAM 估算(PRIMA)2026美国 + 欧盟近期 $60M–$200M;完全渗透上限 $1B–$5BN/A患者口径:约 1M 美国 GA + 约 190K IRD 患者 × 手术适格率 × ASP $50K–$100K低(概念性)无独立验证;手术适格率和 ASP 均为未确认假设

所有已发布分析师数字均为付费报告后的摘要 / 免费层摘录;方法细节有限。若分析师不披露范围,Mordor 与 Insight Partners 的估算无法调和。作者 SAM 仅作说明。

[CM009, CM010, CM011, CM012, CM013, CM036]
FM001: PRIMA 市场规模金字塔——TAM / SAM / SOM

三层市场规模层级,展示从广义视网膜假体 TAM,到 PRIMA 专属 SAM,再到受报销和外科中心产能约束的现实近期 SOM。

所有层级均为作者估计,综合 Mordor Intelligence、Insight Partners 和患者人群数据。没有公开、经验证的 PRIMA 专属规模测算。 TAM 区间反映分析师口径不一致(7× 差距),不代表真实市场不确定性。

[CM009, CM011, CM041]
FM002: 视网膜假体市场估算区间——分析师观点 vs. 作者 SAM

低到高的估算区间,展示分析师和作者对全球视网膜假体 / 植入物市场的看法分歧,单位为百万美元;所有数值均为百万美元。

Mordor 和 Insight Partners 的 2025 年数值为披露值;2031 年和 2034 年为预测值。作者 SAM 仅作示意; 未经独立验证。窄口径与宽口径之间 7× 的分析师差异,反映的是覆盖范围不兼容, 不是市场波动。

[CM009, CM010, CM037]

2.4 买方、用户与支付方分层

Science 的 PRIMA 系统涉及三个不同市场角色。买方——按手术流程意义——是选择技术、接受植入培训并把它纳入临床工作流的视网膜外科医生或专科眼科中心。用户是晚期 GA、RP 或 Stargardt 患者。支付方是 Medicare、商业保险或国家卫生服务体系。初始资本开支的预算归属医院或外科中心;一旦建立报销,单患者报销资金则来自支付方。 在美国,晚期 GA 主要影响符合 Medicare 资格的人群(65 岁及以上),因此 Medicare 是主导支付方。RP 和 Stargardt 患者更年轻,依赖商业保险,某些情况下依赖 Medicaid;IRD 适应症还可能涉及部分罕见病路径。2025 年,医院占全球视网膜植入物终端用户市场的 54.5%(Mordor Intelligence),专科眼科诊所 CAGR 为 11.42%。北美占全球市场份额 39.75%。RP 占当前视网膜植入物市场疾病适应症的 46.95%——这说明当前商业可行的方向,不一定代表天花板。手术能力的地理集中很重要:视网膜下植入需要受过专门训练的视网膜外科医生,早期采用会集中在学术医学中心,从而限制商业化早期的量。 [CM020, CM021, CM022, CM023, CM042]

细分与买方地图——PRIMA 适应症
患者细分主要买方主导支付方(美国)支付方(欧洲)预算负责人采用触发因素关键瓶颈
晚期 GA / 干性 AMD(美国,65 岁以上)学术 / 高容量中心的视网膜外科医生Medicare Part B(器械 + 手术流程)国家卫生服务;目前无覆盖医院 / 门诊手术中心FDA 批准 + 正向 CMS 全国覆盖决定报销;GA 尚无现有 NCD;外科医生培训
终末期 RP(美国,混合年龄)视网膜 / IRD 专科医生Medicare(65 岁以上)或商业保险;部分 Medicaid带 LCD 类似机制的国家卫生服务专科 IRD / 视网膜中心FDA 批准 + 相关 MAC 中的 LCD(部分 RP 视网膜假体已有 LCD)人群极小(约 9,600 例晚期病例);基因治疗竞争
Stargardt 病(美国,年轻成人)视网膜专科医生;学术中心商业保险;部分州罕见病项目通常为国家卫生服务;覆盖极少学术医疗中心FDA 批准 + 商业保险覆盖决定人群很小;无现有覆盖路径;儿童 / 年轻成人照护路径
晚期 GA / 干性 AMD(欧盟,65 岁以上)医院眼科医生 → 视网膜外科医生(转诊)N/A(美国 Medicare 不适用)国家 HTA + 国家级覆盖决定(NHS/GKV/CNAM/SSN)医院;国家卫生系统CE Mark + 正向国家 HTA 意见 + 各国预算谈判CE Mark 待定;EU HTAR JCA 增加延迟;主要欧盟市场无国家覆盖

美国支付方分配为指示性判断,依据各适应症主导年龄分布和已发布 CMS 覆盖指引。欧盟支付方列反映国家卫生服务结构;报销决定由各国在 CE Mark 后作出。「晚期」RP 估算为美国 80K–110K 名 RP 患者中的约 12% = 约 9,600–13,200 名重症患者。

[CM020, CM021, CM022, CM023, CM042]
FM003: 买方与支付方细分矩阵 — PRIMA 适应症覆盖

矩阵展示 PRIMA 在美国和欧盟主要目标适应症中的买方、支付方、报销状态和关键瓶颈。

美国支付方归属为指示性判断,依据各适应症典型患者年龄分布。欧盟列反映国家医疗体系结构; 实际覆盖决定由各国在 CE Mark 之后作出。「部分 MAC 有 LCD」针对 RP 的说法基于历史 Argus II LCD 背景; 是否适用于 PRIMA 仍需 CMS 确认。

[CM020, CM023, CM039]

2.5 增长驱动与采用约束

PRIMA 的主要市场驱动因素,是晚期 GA 没有任何获批的视觉修复疗法:临床对照是自然病程,因此任何程度的功能恢复都代表对患者有意义的进步。人口老龄化会结构性扩大合格人群,因为 AMD 与年龄高度相关。GA 进展不可逆——多数确诊患者会在 2–3 年内丧失大量视力——如果存在安全且可报销的选择,临床紧迫性会推动患者意愿。监管顺风包括 FDA 的突破性设备认定路径缩短视觉假体审评周期,以及遗传性视网膜疾病在北美每年 $13.5B–$32B 疾病负担的成本分析,可支撑支付方经济性论证。远程眼科正在把视网膜植入候选筛查的地理触达从大型中心向外扩展。 但重大采用约束同样真实。视网膜下植入是复杂的专科手术,需要拥有受训视网膜外科医生的中心,手术会集中在学术场景,社区铺开速度受限。Second Sight Argus II 的失败——Second Sight 财务崩溃后,超过 350 名患者被留在过时且无人支持的植入物上——为视网膜假体市场立下了负面先例。Argus II 单台设备定价 $115,000–$150,000,单患者总治疗成本接近 $500,000;保险覆盖有限,是其商业失败的重要原因。患者和支付方评估一款新型视网膜植入物时,仍会高度关注创业公司上市后设备支持不足的风险。外科场景资本强度高、手术学习曲线陡,以及 GA 适应症植入物缺乏既有报销代码,都会叠加限制近期采用。 [CM026, CM027, CM028, CM029, CM030, CM031]

增长驱动因素与采用约束
驱动因素 / 约束方向时点对 PRIMA 的影响尽调问题
晚期 GA 尚无获批视觉恢复疗法强正向驱动当前临床可比是自然病程;竞争位置独特且有利确认 PRIMA 获批前没有 GA 管线疗法跨过疗效阈值
美国 / 欧盟人口老龄化扩大 GA 患病率正向驱动当前和长期可服务人群结构性增长;65 岁后患病率急剧上升用 GA 发病模型跟踪 Census / Eurostat 老龄化队列预测
GA 在诊断后 2–3 年内持续进展正向驱动当前临床紧迫性和患者动机都强;一旦纳入报销,市场拉力会更足核实转诊路径速度;从 GA 诊断到具备植入候选资格要多久
美国尚无针对 GA 适应症视网膜假体的全国覆盖决定硬约束近期(0–3 年)CMS 发布 NCD 之前,商业化只能靠自费或试点项目确认与 CMS 沟通的时间表;评估 MAC LCD 作为过渡路径的可行性
EU HTAR 要求 III 类器械接受联合临床评估(2026 年起)约束中期(1–4 年)EU HTA 证据负担加重;JCA 流程会拖慢各国覆盖决定尽早接触 EU HTA 机构;规划达到 JCA 标准的比较证据生成
Second Sight / Argus II 遗弃患者的前车之鉴负向约束历史 / 持续存在削弱患者和支付方信任;Science 必须拿出明确的上市后设备支持计划确认 Science 对设备长期支持的承诺,以及支撑该支持的财务储备
视网膜下植入需要专科外科医生培训;落点集中在学术中心约束近期至中期限制铺开速度;学术中心容量有限,向社区扩张会慢盘点美国和欧盟专科手术中心容量;评估培训项目能否规模化
抗 VEGF 已固化在视网膜专科医生工作流里轻度约束持续存在视网膜专科医生熟悉注射治疗;把 GA 患者转诊做手术是新行为评估转诊模式研究;确认视网膜专科医生是否愿意为 GA 转外科
视网膜植入手术中心资本开支高约束近期影像、手术室、AI 处理硬件的初始投入,会挡住低量中心测算单中心初始资本需求,以及站点采用达到盈亏平衡所需时间

时间分类:当前 = 今天已适用;近期 = 0–3 年;中期 = 2–5 年。方向判断是作者基于已收集证据作出的定性判断。尽调问题面向未来,现有证据尚不能回答。

[CM026, CM027, CM028, CM029, CM030, CM031]

2.6 报销格局:美国与欧洲

报销是 PRIMA 在美国和欧洲的核心商业风险。在美国,CPT 代码 67027 覆盖眼内视网膜假体植入,Medicare 也已针对终末期视网膜色素变性的视网膜假体发布地方覆盖决定。但地图样萎缩作为视网膜假体适应症——PRIMA 的主要目标——没有全国覆盖决定,意味着初始覆盖将取决于各 MAC 的 LCD 决策和逐案事前授权申请。CMS 2026 年医师费用表将非 APM 参与者转换因子设为 $33.40/RVU,但对门诊手术中心场地程序的支付削减,会削弱依赖医院基础设施的植入项目的程序经济性。 Argus II 先例——单台设备 $115,000–$150,000、单患者总成本接近 $500,000、保险覆盖高度不稳定——说明支付路径未解决的视网膜植入物商业上有多脆弱。Science 必须证明一个设备加手术的可持续经济模型,支付方能够承受,广泛商业采用才可能发生。在欧洲,CE 认证是进入市场的前提,但不带来报销。欧盟卫生技术评估法规(EU 2021/2282)自 2025 年 1 月分阶段实施,要求包括视网膜假体在内的 IIb 和 III 类器械从 2026 年起接受联合临床评估;在各国 HTA 机构作出覆盖决定前,这会增加一项泛欧洲证据要求。每个欧盟成员国独立决定覆盖;法国 HAS、德国 G-BA 和英国 NICE 历来对长期结果数据有限的新型植入器械较为保守。PRIMA 由 Science 提交的 CE 认证申请(约 2024 年末 / 2025 年初)是欧洲商业成功的必要条件,但远远不够。 [CM014, CM015, CM016, CM017, CM018, CM019]

2.7 图表

Chapter 03

03竞争格局

3.1 竞争格局与层级结构

Science Corporation 的 PRIMA 植入物面对两个结构性层级的竞争,它们在临床重叠、时间线和战略相关性上不同。第一层是直接视觉假体竞争,包括为 PRIMA 目标患者群体——AMD 导致的晚期地图样萎缩和终末期遗传性视网膜疾病——开发视网膜或皮层设备的公司。截至 2026 年中,这一层出奇地稀薄:最初开发 PRIMA 的法国公司 Pixium Vision SA 已被清算,资产于 2024 年 4 月被 Science Corp 收购;Retina Implant AG 于 2019 年解散;Second Sight 的 Argus II 到 2020 年已退出商业市场;仍在运营的直接同业只剩澳大利亚的 Bionic Vision Technologies(BVT)和 Cortigent(Vivani Medical 子公司)。第二层是 BCI 资本和注意力竞争,包括 Neuralink、Synchron 和 Precision Neuroscience。它们不争夺同一批患者,却吸收同一池神经科技投资资本和临床中心培训能力。BCI 板块仅 2026 年 Q1 就融资超过 $960M,资本正向行业观察者所称的侵入式 BCI “Big 6”集中。Science Corp 作为首个进入市场的视觉修复 BCI 公司的叙事,要同 Neuralink 更宏大的技术野心、Synchron 的微创优势争夺有限投资人注意力和专科神经外科经验。必须同时理解两层竞争:直接竞争者设定临床证据门槛和商业失败先例,资本竞争者定义融资环境,并可能把医院和支付方注意力从视觉修复上带走。Pixium Vision 的历史很有启示意义——它开发 PRIMA、开展早期临床试验,并搭建了 Science Corp 如今运营的 CE 认证申请基础设施。收购不只是一次 IP 交易:它移除了一个潜在近期竞争者,并让 Science Corp 独占目前临床进展最成熟的视网膜下光伏平台。截至 2026 年 6 月,任何永久植入式 BCI 设备都没有 Medicare 全国覆盖决定或 FDA 上市前批准,这意味着两个层级的所有竞争者都共同面对未解决的报销风险。 [CP001, CP002, CP026, CP027, CP028]

竞争对手画像表
公司类别累计融资目标适应症模式 / 路径监管阶段关键差异化关键限制
Science Corporation (PRIMA)直接视网膜同业(研究对象)~$490MAMD 导致的 GA、RP、Stargardt视网膜下光伏(378 像素)CE Mark 申请已提交;FDA IDE 有效NEJM 证据;无线;欧盟先发美国报销未解决;前代产品失败
Bionic Vision Technologies (BVT)直接视网膜同业~$18M视网膜色素变性脉络膜上电极阵列(44 通道)FDA 突破性设备;关键试验待启动手术风险更低;2.7 年电极保留率 97%电极数量低;资金不足;尚无商业批准
Cortigent / Vivani Medical皮层视觉同业未披露(Vivani 支持)任何原因导致的法定 / 完全失明视觉皮层假体(ECoG 阵列)6 年可行性完成;关键试验规划中绕过视网膜;适应症广;安全性已显示需要开颅;阶段早;母公司转向药物递送
Second Sight — Argus II历史直接同业(已停产)生命周期约 $300M视网膜色素变性视网膜前电极阵列(60 通道)2019–2020 年停产首个获 FDA 批准的仿生眼(2013 年 HDE)商业化失败;约 350 名患者被遗留且缺乏支持
Retina Implant AG — Alpha AMS(前代产品)历史直接同业(已解散)未披露视网膜色素变性视网膜下电极阵列(1,600 通道)2019 年 3 月解散2016 年获 CE 认证;当时电极数量最高商业化失败;未被收购即解散
NeuralinkBCI 资本竞争者~$1.3B四肢瘫痪(运动 / 沟通)皮层内穿刺阵列(N1,约 1,024 通道)FDA IDE 有效(PRIME 试验,已入组 21 人)资本化最高;芯片到机器人控制全栈无视觉适应症;电极回缩问题;需要开颅
SynchronBCI 资本竞争者~$345MALS、中风、脊髓损伤(沟通)血管内 ECoG(Stentrode,约 32 个触点)突破性设备;关键试验计划 2026 年启动微创(无需开颅);医院更容易接受无视觉适应症;信号分辨率有限;无 PMA
Precision NeuroscienceBCI 资本竞争者~$290MALS、运动障碍硬膜下 ECoG 薄膜(Layer 7)510(k) 获批可使用 30 天;IDE 申请待提交唯一获得 FDA 商业许可的下一代 BCI;85+ 名患者无视觉适应症;仅获短期 FDA 许可;PMA 路径很长

融资数字为截至 2026 年 6 月的披露值或估计值;历史公司(Second Sight、Retina Implant AG、Pixium)仅作背景。所有监管阶段截至 2026 年中。"n/a" = 信息未公开。

[CP001, CP003, CP006, CP011, CP013, CP018]

3.2 直接视觉假体同业——传统失败与存活玩家

直接视网膜假体竞争史很大程度上是一部商业失败史,这也让 Science Corp 的 PRIMA 所在赛道中活跃同业少于技术版图表面显示的数量。Second Sight Medical Products 是第一家获得 FDA 视网膜假体批准的公司,其 Argus II 视网膜上设备于 2013 年通过 Humanitarian Device Exemption(HDE)路径获批;但在 2019–2020 年撤回产品前,公司用约 20 年累计投入约 $300M,只从全球约 350 名患者身上产生约 $32M 总收入。关键失败原因包括:患者群体太小,无法支撑商业经济性;总手术成本过高(含手术和康复约 $500,000);无法获得广泛 Medicare Administrative Contractor(MAC)报销覆盖;医生对新型手术流程不熟悉。产品撤回后,约 350 名 Argus II 患者被留在无人支持的植入物上,形成了具有长期伦理和监管影响的患者遗弃先例。Second Sight 随后于 2022 年与 Nano Precision Medical 合并为 Vivani Medical,并转向药物递送;神经刺激业务被剥离为 Cortigent。Cortigent 的 Orion 视觉皮层假体系统于 2025 年 3 月完成 6 年早期可行性研究,全部 6 名参与者均显示安全性、可靠性和可测量的视觉评估改善,电极功能损失低于 4%。Orion 刺激视觉皮层,而非视网膜,绕过整个受损视觉通路——这是一项重要技术差异,因为它可服务同时失去感光细胞和视网膜神经节细胞功能的患者,适应症比 PRIMA 更广。然而,Cortigent 需要开颅,商业化道路上至少落后 PRIMA 一个完整开发阶段。德国 Alpha AMS 视网膜下电极阵列开发商 Retina Implant AG(2016 年获 CE 认证,1,600 个电极)于 2019 年 3 月决定解散,理由是欧洲报销不足、监管门槛高、临床结果未达到患者预期;公司没有收购方。澳大利亚公司 Bionic Vision Technologies(BVT)是唯一仍活跃、且拥有可信临床数据的直接视网膜假体竞争者。其第二代 44 通道脉络膜上设备植入在视网膜层外、位于巩膜和脉络膜之间;2024 年发表结果显示,4 名临床试验参与者 2.7 年时电极保留率为 97%,并在导航和物体探测上获得功能改善,FDA Breakthrough Device Designation 支持加速监管路径。不过,BVT 披露总融资约 $18M,约为 Science Corp 的 1/27;要开展全球关键试验并进入商业化,还需要大幅增加资本。 [CP003, CP004, CP005, CP006, CP007, CP008]

FP001: 竞争定位图

截至 2026 年中,Science Corp 直接和间接竞争对手在侵入性与监管 / 商业成熟度上的序数定位。 X 轴:手术侵入性(1 = 最低,10 = 最高)。Y 轴:监管 / 商业成熟度(1 = 临床前, 10 = 已上市商业产品)。分数为基于公开证据的方向性分析师判断;并非经审计指标。

轴向分数为有证据支撑的序数估算。侵入性来自手术路径(血管内 < 脉络膜上腔 < 视网膜下 < 硬膜下 < 穿透 / 开颅)。监管成熟度来自已达到的最高里程碑(商业上市=10;CE Mark 已提交=8; Breakthrough Device + 活跃关键试验=6;活跃早期试验=5;可行性完成=4;仅可行性=3)。 Second Sight 作为历史参照展示。

[CP003, CP009, CP014, CP019, CP023, CP029]

3.3 BCI 资本和注意力竞争——Neuralink、Synchron、Precision Neuroscience

Neuralink、Synchron 和 Precision Neuroscience 不做视觉修复,也不争夺 Science Corp 的患者,但它们是同一投资资本池、临床中心培训资源和公众 BCI 叙事的主导索取者;这些资源决定估值倍数、媒体关注和医院合作入口。Neuralink 是资本最充足的 BCI 公司,2025 年 6 月完成 $650M Series E,据报累计融资约 $1.3B。截至 2026 年初,其 PRIME 研究(NCT06429735)已在美国、加拿大和英国入组 21 名参与者,目标是四肢瘫痪患者的光标控制应用。尽管入组势头强劲,Neuralink 第一名人体参与者在植入后数周内约 85% 电极线从脑组织回缩,显著削弱设备信号解码能力;软件更新部分补偿了问题,但未解决硬件不稳定。美国 SEC 于 2024 年末重启对 Neuralink 潜在误导性安全声明的调查。Elon Musk 的公众形象驱动 Neuralink 获得异常强的媒体存在感,形成 BCI 心智份额,可能把投资人注意力和媒体覆盖从 Science Corp 更接近商业化、临床故事也可说更成熟的方向上带走。Synchron 于 2025 年 11 月完成 $200M Series D,使累计融资达到约 $345M;该资金明确投向 2026 年 Stentrode 血管内设备的 FDA 关键临床试验。Synchron 的路径通过颈静脉把植入物送入运动皮层血供,避免开颅;相比穿透式电极方案,这种微创优势扩大了医院采用漏斗。Synchron 面向 ALS、瘫痪和中风患者;它同 Science Corp 竞争的是资本和临床中心注意力,不是同一批患者。如果 Synchron 关键试验成功并获得 FDA PMA,它正在争夺“美国首个商业获批 BCI”的头衔,这一里程碑会在叙事上与 Science Corp 潜在的欧盟率先商业化故事竞争。Precision Neuroscience 于 2026 年 6 月完成 $120M Series D,使累计融资超过 $290M,由 General Catalyst 领投,Microsoft M12 和 Coatue Management 参与。其 Layer 7 Cortical Interface 是硬膜下薄膜电极阵列,已获得 FDA 510(k) 用于最长 30 天神经生理应用的许可,使其成为截至 2026 年中唯一拥有任何商业 FDA 许可的新一代皮层 BCI 设备。公司已在超过 85 名患者体内植入设备,合作医院超过 15 家,并正准备为关键 ALS 试验提交 IDE。和 Neuralink、Synchron 一样,Precision Neuroscience 不面向视觉修复;其竞争相关性在资本和监管注意力。到 2026 年中,这三家公司合计吸收超过 $1.5B 累计资本,形成竞争性融资环境,Science Corp 必须围绕监管接近度和临床证据特异性讲出差异化叙事。 [CP013, CP014, CP015, CP016, CP017, CP018]

功能 / 能力矩阵
标准Science PRIMABVT Gen-2Cortigent OrionNeuralink N1Synchron StentrodePrecision Layer 7
主要临床适应症GA/AMD + RP + Stargardt视网膜色素变性任何失明(皮层旁路)四肢瘫痪(光标控制)瘫痪(ALS、SCI、中风)瘫痪(ALS、运动)
植入解剖位置视网膜下(RPE 下方)脉络膜上(视网膜外)视觉皮层(ECoG,开颅)皮层内运动皮层(穿刺)运动皮层血管内(血管内)硬膜下运动皮层(薄膜)
手术侵入性中等(玻璃体视网膜)低至中等(脉络膜上)高(开颅)高(开颅 + 机器人植入)低(血管内 / 颈静脉)中等(微型开颅)
无线 / 免电池运行是 — 光伏(无电池)否 — 需要外部供电否 — 需要外部单元否 — 需要外部感应充电器否 — 需要外部集线器否 — 需要外部设备
监管状态(最先进)CE Mark 申请已提交(2024 年)突破性设备;尚未获批6 年可行性已完成;关键试验待定FDA IDE 有效(PRIME 试验,21 人)突破性设备;关键试验计划 2026 年启动FDA 510(k) 获批用于 30 天脑图谱绘制
已发表的同行评议临床证据NEJM 2025 年 10 月(N=38–47;80% 改善)2024 年试点研究(N=4;97% 电极保留)NANS 2026 摘要(N=6;6 年可行性)无同行评议发表;公司博客更新早期 FIH 结果有限;无主要期刊发表技术报告(脑图谱绘制);无长期数据
目标商业发布节奏欧盟 2026 年下半年(待 CE Mark)2026 年后(关键试验待启动)2028 年后(关键试验规划中)未说明商业发布时间表美国 2027 年后(若关键试验成功)美国 2027 年后(IDE/PMA 路径)

单元格反映截至 2026 年 6 月的状态。"未知" 表示公开证据不足,无法确认或否定该标准。"n/a" 表示该标准不适用于设备预期用途。

[CP003, CP004, CP009, CP014, CP018, CP023]
FP002: 功能广度 / 能力图谱

截至 2026 年中,Science Corp 与五家活跃同行的能力覆盖对比。每一行是关键差异化标准; 单元格汇总每个竞争者有证据支撑的位置。

[CP005, CP010, CP016, CP020, CP025, CP030]

3.4 比较技术差异——模态、侵入性与证据成熟度

Science Corp 的 PRIMA 在三个维度上区别于所有活跃竞争者:证据成熟度、商业化接近度和无线优先架构。在所有视网膜和皮层假体公司中,PRIMA 是唯一在 CE 认证提交时已拥有 New England Journal of Medicine 发表的 II 期数据、并证明功能性视觉修复的设备——PRIMAVERA 试验中 38 至 47 名参与者里,80% 患者显示有意义的视觉改善。无论是直接视觉假体还是更广义的 BCI 层级,没有竞争者同时达到这种期刊层级、入组规模和监管申报组合。PRIMA 的光伏无线设计在架构上不同:视网膜下芯片中的 378 个光伏像素把患者眼镜传来的近红外光转化为对双极细胞的电刺激,完全取消外部电源连接和经皮导线;相比有线或电池充电方案,这降低感染风险,也简化系统。BVT 的脉络膜上路径把植入物放在视网膜层外,相比视网膜上和视网膜下路径,手术复杂度和视网膜创伤风险更低,但 44 电极设计的理论分辨率显著低于 PRIMA 的 378 像素,且只在 4 名受试者中测试。Cortigent 的 Orion 完全绕过视网膜电路,直接刺激视觉皮层,这是真实的技术差异,可能服务更广泛的失明人群,但需要开颅——手术负担比 PRIMA 的玻璃体视网膜手术高出数量级。Science Corp 收购了 Pixium Vision 全部资产,包括临床试验基础设施、IP 和 CE 认证申请,使其独占 PRIMA 平台且不存在竞争权利主张。Argus II 获得 CMS $150,000 报销、但总手术成本约 $500,000 的先例,留下了负面商业标杆;PRIMA 必须用支付方证据和差异化结果数据明确跨过去。PRIMA 在视网膜下空间刺激双极细胞,保留包括侧抑制在内的内层视网膜处理;相比绕过同一电路的 Argus II 视网膜上神经节细胞刺激,这带来理论分辨率优势。在更广义的 BCI 资本竞争中,Synchron 较低的手术负担是真实竞争优势,因为医院部署成本和操作者认证时间更偏向血管内系统;但这一优势不适用于视网膜手术,后者是成熟术式专科,已有既定外科医生培训路径。Precision Neuroscience 获 510(k) 许可的 Layer 7 电极,为薄膜皮层接口建立了商业监管先例,这一数据点关系到 Science Corp 未来在视网膜适应症之外扩展皮层 BCI 平台。总体看,Science Corp 在监管接近度、证据成熟度和系统优雅性上领先,但仍面对同样的商业化开销——医生采用、支付方沟通和长期支持义务——这些曾摧毁其直接前辈。 [CP029, CP030, CP031, CP032, CP033, CP041]

定价 / 包装对比
产品 / 公司估计系统价格总手术费用报销状态对 Science Corp 的含义
Science PRIMA未披露(商业化前)未披露欧盟 CE Mark 后(待定);美国:仅限研究Argus II 有 $150K CMS 先例;Science Corp 必须证明每美元带来的结果更优
Second Sight Argus II(历史)设备约 $150,000总计约 $500,000CMS HDE $150K;MAC 逐案覆盖;已停产直接负面先例 — CMS 报销不足以支撑商业可行性
Bionic Vision Technologies未披露(商业化前)未披露无商业支付方路径(未获批)定价风险未知;资金不足;没有 M&A 很难做出商业规模
Neuralink未披露(研究阶段)仅限研究无商业定价;不适用只是资本竞争者;没有与 PRIMA 商业论证相关的定价数据
Synchron Stentrode未披露(商业化前)未披露尚无商业支付方路径;关键试验 2026 年启动会建立独立于视觉假体的支付方先例

所有定价均为估计或历史数据;截至 2026 年 6 月,除 Second Sight 外,没有公司披露商业标价。"总手术费用" 包含设备、手术和初始康复。CMS = 美国医疗保险和医疗补助服务中心。

[CP007, CP032, CP027]

3.5 护城河耐久性、替代风险与反证

Science Corp 的竞争护城河真实但脆弱,这一精确竞争细分赛道的历史失败率要求严苛的反方核算。直接视网膜假体的三个前辈——Second Sight、Retina Implant AG 和 Pixium Vision——要么商业失败,要么在困境中被收购。Argus II 的 CE 标志和 FDA HDE 批准没有转化为商业可行性;Retina Implant AG 于 2016 年获得 CE 标志,三年后解散;Pixium Vision 用 PRIMA 走到 CE 认证提交阶段,却在批准前耗尽资本。每一次失败都由报销不足、手术成本高、患者量低和医生采用惯性中的若干因素驱动——Science Corp 尚未证明自己能在规模上克服任何一项。GAO 2025 年关于脑机接口的报告明确指出,缺乏专用报销编码、隐私框架和上市后监测标准是系统性障碍,需要政策行动后,BCI 才可能广泛商业化;Medicare 目前没有针对永久植入式 BCI 的全国覆盖决定或专用编码路径,这意味着 PRIMA 在美国会面临逐案 MAC 覆盖谈判,类似拖垮 Argus II 的路径。如果 Science Corp 陷入财务困境,继续支持植入患者的伦理和监管义务也没有解决;Argus II 遗弃先例(约 350 名患者被留在无人支持的植入物上)制造了推动政策变化的压力,但变化尚未落地。Neuralink 约 $1.3B 资本基础和定义品类的媒体存在感,若其比 Science Corp 实现欧盟商业放量更快获得 FDA 批准,可能把 BCI 投资资金流引向运动 / 通信应用、远离视觉修复。BVT 虽然融资有限,但持有 FDA Breakthrough Device Designation,并在执行清晰的临床项目;如果 BVT 获得大型医疗科技公司收购或合作,就可能迅速扩展为可信竞争者,获得 Science Corp 需要数年才能复制的制造和分销支持。Science Corp 的植入式设备锁定在患者层面很强——取出和更换的手术风险形成事实上的患者留存——但这不能阻止医院系统与面向运动应用的竞争 BCI 平台建立关系,从而可能限制 Science Corp 未来在院内的扩张。外科医生专业化构成有意义的护城河:PRIMA 植入所需的玻璃体视网膜手术技能在主要学术中心已经成熟,相比皮层竞争者所需的新型开颅技术,培训曲线更低。PMC 发表的 Second Sight 与 Onward Medical 比较分析认为,分阶段商业化、借助先导技术和医生熟悉度是关键成功因素,Second Sight 未能使用这些因素;Science Corp 管理层从临床转向商业模式时必须内化这一点。总结来说,Science Corp 的护城河在狭义上真实:它是唯一拥有 NEJM 发表证据、且有正在审理的 CE 认证申请的视网膜 BCI 视觉修复公司;但在结构意义上,只有报销、医生采用和上市后支持模型在商业规模上得到证明后,护城河才算耐久。 [CP034, CP035, CP036, CP037, CP038, CP039]

护城河耐久性 / 竞争风险登记表
护城河主张竞争威胁严重性缓释 / 尽调问题
视网膜 BCI 首个 CE Mark批准延迟(NB 评估需 12–24 个月);BVT 获得更多资金后缩小差距每月跟踪 CE Mark 时间线;为延迟 6–12 个月准备应急方案
NEJM 临床证据优势BVT 或 Cortigent 发表可比结果;Neuralink 率先达成 FDA 里程碑监测同业发表;在随访队列中守住证据优势
自有芯片制造(成本和迭代速度)竞争对手拿下 TSMC 或 IMEC 合作;芯片密度进展快于 Science Corp核实制造产能、良率和迭代周期,并与外包同业对比
收购 Pixium 资产(独占 IP 控制)学术孵化公司开发替代光伏路径;专利遭挑战低-中委托专利 FTO 审查;监测学术光伏植入申请
相比 Argus II 更强的报销叙事PRIMA 未能证明相对现状(假手术对照)具有 CMS 可接受的成本效果现在就接触支付方关系团队;委托健康经济学模型;尽早争取 NCD 沟通
植入设备带来的患者锁定(转换成本高)下一代平台(基因疗法、光遗传学)让 PRIMA 的电刺激路径过时中-长期Science Corp 自有 Science Eye 光遗传学平台是内部对冲;仍需跟踪时间线

严重性评级是分析师基于截至 2026 年 6 月公开证据作出的判断。高 = 可能实质影响商业化轨迹;中 = 执行得当即可管理;低 = 风险较小或较远。

[CP034, CP036, CP039, CP040]
FP003: 护城河 / 就绪度 KPI

截至 2026 年中,Science Corp 与关键 BCI 同行的竞争耐久性指标快照。条目结合披露数据和基于公开来源的分析师综合判断。

融资总额为已报告或根据公开轮次披露估算的累计值。PRIMAVERA 入组区间反映两个试验队列。 竞争者试验规模来自公司公告和 ClinicalTrials.gov 登记。

[CP013, CP015, CP026, CP029, CP037]

3.6 图表

Chapter 04

04财务

4.1 融资历史与资本结构

Science Corporation 自 2021 年创立以来,已通过至少三次有记录的资本事件融资约 $490M。最近一次是 2026 年 3 月超额认购的 $230M Series C,由 Lightspeed Venture Partners 和 Khosla Ventures 领投,Y Combinator、IQT(In-Q-Tel)和 Quiet Capital 追加参与;五家均为复投投资者,说明信心延续,而非新验证者入场。该轮投后估值约 $1.5B,使 Science Corp 成为仅次于 Neuralink、资本化程度第二高的专注脑植入创业公司。Series C 之前,公司于 2025 年 4 月完成 Khosla Ventures 领投的 $104M 可转债。该票据条款——估值上限、折扣率和转换机制——未公开披露;票据是在 Series C 交割时转换,还是仍作为未偿义务存在,也不得而知。Series A 和 Series B 的单轮规模未公开报告,使完整的逐轮稀释路径在尽调中不透明。超额认购结果,以及从票据(2025 年 4 月)到股权轮(2026 年 3 月)的快速节奏,说明投资者需求强劲;但未披露的股权结构阻止外部独立核验创始人稀释、清算优先权栈深度或投资人清算优先级。Science Corp 在首笔商业收入前已投入 $490M 总资本,明确处于临床阶段医疗科技的资本密集层级;在这里,能融资是商业成功的必要条件,但远远不够。 [CI001, CI002, CI003, CI004, CI005, CI006]

资本充足性表
参数数值 / 估计置信度注释
总融资额(截至 2026 年中累计)~$490M依据 BusinessWire Series C 官方公告和 science.xyz;多家新闻来源确认
Series C(2026 年 3 月)$230M超额认购;Lightspeed + Khosla + YC + IQT + Quiet Capital;所有既有投资者参与
2025 年 4 月可转债~$104MKhosla Ventures 领投;转换条款未披露;可能已在 Series C 转换
投后估值(Series C)~$1.5B多家媒体报道;未独立确认;在专注 BCI 公司中仅次于 Neuralink
估计年烧钱速度$15M–$35M/年临床阶段医疗科技行业基准;公司未披露烧钱速度
隐含资金可支撑年限(仅 Series C)6–15 年基于 $230M 和估计烧钱区间;不含代工厂资本开支提款节奏
未偿债务 / 项目融资UnknownUnknown可转债可能仍未偿还;未公开披露正式债务或授信额度

所有估计均来自行业基准;公司财务为私有。高置信度条目来自官方新闻稿。低置信度条目是分析师基准或推断。

[CI001, CI002, CI003, CI004, CI006, CI020]
FI003: 财务估算区间 — 烧钱、现金跑道、定价和估值输入

截至 2026 年中,Science Corporation 关键财务输入有来源支撑的上下界;所有区间来自行业基准或报道数据, 而非公司披露数据。

所有区间均为估算或基准。Science Corp 未披露烧钱率、收入或利润率。估值边界反映多个来源报道区间; 实际投前估值可能不同。

[CI004, CI016, CI020, CI029]

4.2 资本强度与制造扩建

Science Corporation 的资本强度由两个相互强化的成本中心撑住: 临床阶段 R&D(试验、监管申报和持续扩张的工程团队),以及 专门建设的半导体制造基础设施。Science Foundry 部门位于北卡罗来纳州 Durham 的 Research Triangle Park,一座 MEMS 工厂内;该工厂原属法国公司 MEMSCAP,2022 年被收购。Science Corp 计划投入最高 $65 million,将该工厂扩建 57,000 平方英尺;Durham County 委员会一致批准一项为期十年、金额 $930,000 的 绩效型激励,前提是公司达成就业创造和投资目标。扩建预计新增 50 多个岗位, 其中约三分之二只要求两年制学位或更低学历,说明团队构成更偏制造一线,而非 纯研究。多数 medtech 同业外包 MEMS 生产;Science Corp 自建能力,是为了缩短 迭代周期、压低长期单机成本,并消除单一来源供应风险。该 foundry 也向其他医疗和 BCI 设备公司提供第三方 MEMS 服务,正在形成一条早期的第二收入线。$65 million 的 foundry 资本开支叠加在持续 R&D 和临床费用之上,使年度现金消耗在结构上高于 轻资产或外包制造的同业。这项资本开支承诺于 2024 年中宣布,并与 PRIMA 商业化 推进同步执行,意味着制造放量和临床进展并行——这是有意为之的战略选择,但资本 消耗很重,需要投资者在多个里程碑关口持续保持信心。 [CI010, CI011, CI012, CI013, CI032, CI033]

单位经济性表
指标数值 / 估计置信度重要性尽调问题
设备 ASP(PRIMA)未披露(估算 $85K–$145K)单台手术的主要收入驱动项;决定市场规模能否兑现索取标价和医院净价;核查欧盟与美国定价策略
单台设备 COGS未披露(试点产量下估算 $5K–$12K)设定毛利率下限;驱动盈利路径核查物料清单、代工成本分摊、良率假设
毛利率未披露(规模化后估算 85–92%)规模化后经济性是否可持续;可对标高端医疗科技植入物索取当前和前瞻 COGS 与 ASP 排期
客户获取成本(CAC)未披露Unknown医生培训 + 医院采购周期主导获客成本索取单家医院销售周期时长和培训成本估算
回本周期未披露Unknown多年付款方授权周期拉长现金流回收取决于付款方覆盖时间表;上市前无法计算
净收入留存(NRR)N/A(一次性植入物)N/A标准 SaaS 指标不适用;应改看患者终身支持成本评估植入后支持合同或升级路径经济性
R&D / 收入比尚未产生收入;无法计算Unknown商业化后跟踪资本效率;目前不透明首个完整商业季度后计算
烧钱倍数尚未产生收入;无法计算Unknown跟踪每新增一单位 ARR 消耗的资本;上市后才有意义首个商业季度后计算;与医疗科技可比公司对标

所有估算均由作者基于行业基准和类似设备数据(Argus II、Pixium PRIMA)推导;Science Corp 尚未披露任何单位经济性。低置信度估算仅作方向性参考。

[CI016, CI026, CI027, CI028, CI032]
FI004: 资本强度瀑布 — Series C 部署情景

示意性展示 $230 million Series C 募资如何匹配估算资本需求;所有非 Series C 项目均为基于已宣布承诺和行业基准推导的估算。

除 Series C 募资外,所有项目均为作者基于公开承诺(foundry)和行业基准(临床研发、运营)的估算。 实际部署会不同于该情景。Science Corp 未披露资本部署计划。

[CI007, CI010, CI011, CI020, CI033]

4.3 收入模式与报销格局

截至 2026 年中,Science Corporation 在美国仍未产生收入。PRIMA 的商业发布目标是在 CE Mark 申报之后,于 2026 年晚些时候先在欧洲落地;如果实现,Science Corp 将成为 第一家把视觉恢复产品推向商业市场的 BCI 公司。Science Corp 尚未公开任何标价、 医院采购价或手术费用。最近的历史参照是 Argus II(Second Sight):其单设备标价约 $145,000,拥有专门的 CMS 报销代码(C1849),总治疗周期成本(设备、手术和康复) 为 $200,000–$300,000。PRIMA 在欧洲的参照价则来自 Pixium Vision 破产前的报价: 单设备 €80,000–€100,000。美国报销存在结构性不确定性:Anthem 2026 年 4 月的 覆盖政策明确把所有人工视网膜设备列为试验性,且对所有适应症均非医学必需——这代表 当前主要支付方的默认立场,而不是孤例。CMS 和 FDA 于 2026 年 4 月联合宣布 RAPID 覆盖路径,旨在让获得 Breakthrough Device designation 的产品在 FDA 批准后 60–90 天内 获得 Medicare 覆盖;但截至研究日期,公开记录尚未确认 PRIMA 是否拥有 Breakthrough Device designation。新的植入手术通常需要定制 HCPCS 计费代码;Review of Ophthalmology 的 2026 年编码更新确认,人工视网膜假体的覆盖仍取决于具体支付方。CMS 2026 OPPS 最终规则只给复杂眼内手术带来小幅支付提升,APC 平均报销约 $4,222——远低于预期设备 采购成本,进一步说明美国商业采用放量前必须先补上报销缺口。 [CI007, CI008, CI014, CI015, CI016, CI017]

收入来源表
来源机制单位当前状态证据质量尽调问题
PRIMA 欧盟商业发布按手术向医院销售设备每枚植入体欧盟 CE Mark 已提交;目标 2026 年发布;尚未确认收入公司声称确认 ASP、医院净价、外科医生培训成本、各国支付方覆盖
PRIMA 美国临床(IDE)不适用 — 研究设备N/A商业化前;FDA 监管审查进行中监管申报监测 FDA 批准时间线;厘清 IDE 何时转入商业化
PRIMA Stargardt/RP 扩展未来按手术向医院销售设备每枚植入体(未来)I 期已宣布;无收入公司声称首例商业手术时间线;分适应症的监管路径
Science Foundry 第三方 MEMS 服务合同制造收入每片晶圆 / 每批次早期;未披露收入或具名客户基于新闻收入贡献;客户集中度;相对自有设备的利润率
未来脑植入系统医院设备销售;配置未披露N/A临床前;管线阶段未披露公司声称明确监管路径、定价模型和临床开发成本
政府 / R&D 资助资助收入或 R&D 合同每份合同未披露;IQT 投资者关系可能意味着政府 R&D 工作估计核实是否有 SBIR/STTR、IQT 合同或 NIH 资助;判断归类为收入还是成本抵扣

所有当前状态均来自公司声称或新闻;Science Corp 尚未披露任何收入来源的收入。美国监管状态依据截至 runDate 的公开记录。欧盟和未来收入来源根据公司公告推断。

[CI007, CI008, CI009, CI014, CI015]
定价和变现表
产品 / 情景标价(仅设备)完整治疗事件成本(估算)付款方覆盖状态置信度 / 来源
PRIMA 欧盟(目标 2026 年上市)€80K–€100K(估算,Pixium 破产前参考)€120K–€160K(估算)CE 认证待定;各国付款方覆盖未定低 — Pixium 历史定价;Science Corp 未披露
PRIMA 美国(FDA 批准后预测)未披露$200K–$300K(Argus II 类比)无覆盖;Anthem 将其归为研究性项目(2026)低 — 全部为估算;Science Corp 无公开定价
Argus II 历史基准~$145K(仅设备)$200K–$300KCMS C1849 过渡支付代码(约 2022 年失效);当前无覆盖中 — 公开 CMS/MassDevice 记录
PRIMA Stargardt/RP(未来适应症)未披露未披露无监管批准;覆盖路径未定义N/A — 临床前
Science Foundry MEMS 服务(B2B)按批次;未披露N/A不适用(B2B 合同制造)低 — 无公开定价;仅宣布代工服务

欧盟定价由 Pixium Vision 历史参考和 Science Corp 未披露的商业化计划外推。美国定价由 Argus II 历史数据估算;Science Corp 未披露 PRIMA 定价。Argus II pass-through 代码 C1849 已失效,不是当前有效报销路径。

[CI016, CI017, CI028, CI029, CI036]
FI001: 收入模型桥 — 从 PRIMA 临床活动到毛利润

PRIMA 植入手术如何从患者资格确认,经支付方裁定,转化为 Science Corp 设备收入和毛利润贡献;图中突出报销关口。

所有美元数值均为基于行业基准(Argus II、Pixium)的估算。Science Corp 未披露设备定价、COGS 或利润率。 支付方覆盖状态反映 Anthem 2026 政策;其他支付方可能不同。欧盟收入路径(CE Mark)可能早于美国商业上市。

[CI014, CI015, CI017, CI019, CI027, CI028]

4.4 单位经济框架与私有指标缺口

Science Corporation 未公开任何收入、ARR、毛利率、销货成本、获客成本、回本周期或 净收入留存。公司在美国仍处于商业化前阶段,所有单位经济要么是预先设定的估算, 要么是只有 NDA 下才能看到的专有信息。行业基准显示,视网膜下设备在试点制造量级下 单台 COGS 约为 $5,000–$12,000;随着自动化放量和供应链成熟,成本会明显下降。 Science Corp 的垂直 MEMS 整合在规模达成后可能带来结构性 COGS 优势,但 $65 million foundry 扩建已经是沉没资本,只有依靠尚未达到的设备销量才能回收。ABCMoney 的投资分析 估计,若要用典型 medtech 退出倍数支撑 $1.5 billion 的入场估值,公司需要在 7–10 年内 做到 $500 million 或更高年收入;这意味着 PRIMA 每年要完成数千台手术,单台价格 $50,000+——如果报销跑通且 AMD 可服务人群能规模化触达,这一假设激进但并非不可想象。 PRIMA 的销售周期经济与 SaaS 指标完全不同:医生培训成本、医院资本设备采购周期、 单个支付方覆盖决定,将主导获客成本。标准 SaaS 或 marketplace 单位经济(NRR、 quick ratio、burn multiple)不适用于一次性植入模式。截至 2026 年中,Science Corp 尚未以商业口径公开描述 EU 发布的 GTM 路径,因此现阶段无法独立评估销售效率。 [CI014, CI016, CI025, CI026, CI027, CI028]

公开财务缺口表
缺失指标对投资判断的影响尽调路径
收入(任何期间)无法评估商业牵引、预测准确性或收入质量在 NDA 下索取中期财务报表;如可获取,结合付款方理赔数据核验
账上现金和月度 burn rate考虑代工厂资本开支承诺后,无法验证现金跑道是否充足、资本效率如何向管理层索取季度现金流量表和烧钱桥接表
COGS 和毛利率无法评估单位经济性、规模化路径或毛利美元走势在 NDA 下核查当前 COGS 和 3 年前瞻预测
股权结构和清算瀑布无法评估下行情景回报、创始人稀释,或 $490M 融资带来的清算优先权压力索取完整股权结构、投资人条款摘要和清算瀑布模型
可转债条款和转换状态无法判断 2025 年 4 月票据带来的潜在股权稀释或持续债务负担索取票据协议、转换机制和 Series C 资金流向核对

所有缺口均指研究日以前未公开的信息。Science Corp 是私营公司,没有 SEC 申报义务。尽调路径是 VC/PE 标准信息请求项。

[CI014, CI026, CI030, CI031]
FI002: 单位经济桥 — 从设备成本输入到客户经济性

这组定性节点追踪决定 PRIMA 单位经济性的输入,从制造成本到毛利率,再到客户获取和回本; 所有数值均为估算或未披露。

所有节点均为定性或估算;Science Corp 未披露任何单位经济。CAC 和回本周期高度取决于美国市场尚未拿下的支付方覆盖决定。

[CI025, CI026, CI027, CI032, CI039]

4.5 反向证据与财务风险因素

Science Corp 财务模型最突出的反向证据,来自三家直接前身的商业化记录。Second Sight Medical Products 在 Argus II 视网膜假体上投入超过 $300 million、历时 24 年,累计收入 不到 $32 million,随后在 2019 年突然停止,导致全球数百名植入者失去产品支持。 PRIMA 的原开发方 Pixium Vision 在耗尽资本后于 2024 年初进入司法清算;Science Corp 在法院监督下收购其资产。Retina Implant AG 的 Alpha AMS 未能达到商业规模后,于 2019 年解散。三家前身都有可信的临床科学,却都未能把试验数据跨到规模化商业收入。 报销壁垒是结构性的:Anthem 2026 年政策把所有人工视网膜设备列为试验性且非医学必需; 在大规模临床结局数据和标准化 CPT 代码到位前,这就是主流支付方立场。BCI 领域融资在 2026 年 Q1 创下 $960 million 的纪录,Science Corp 超额认购的 Series C 也确认当前投资者 风险偏好仍强,但 VC 资金具有阶段性;临床挫折、监管延误或宏观收紧,都可能显著压缩 可用 runway。MedTech Impact Partners 警告,CMS 提议在 FY2028 前废止 NTAP 替代支付路径, 将提高医院附加支付的证据门槛,压窄届时尚未获覆盖设备的报销窗口。Science 已通过多轮 优先股融资筹得 $490 million;若退出不利,投资者清算优先权可能大幅压缩普通股和创始人 回报,而不透明的 cap table 使该风险无法独立评估。 [CI017, CI021, CI022, CI023, CI024, CI030]

Chapter 05

05产品与技术

5.1 PRIMA 植入体架构与近红外图像路径

PRIMA 系统面向因 geographic atrophy(GA)导致感光细胞被破坏、但内层视网膜仍能工作的患者, 恢复其中心视觉。该设备利用眼球的光学透明性:外部近红外投影器通过瞳孔向无线视网膜下芯片 同时传输能量和图像数据,去掉了早期视网膜假体中问题频出的电池和经皮线缆。 植入体本身是一块薄型晶体硅光伏阵列,占地 2 × 2 mm,厚度仅 30 µm——约为人类头发直径的一半。 其 378 个独立控制像素按六边形(蜂窝)网格排列;每个像素宽 100 µm,同时充当太阳能电池和电极。 当 880 nm 红外光照射某个像素时,该像素会产生局部电流,刺激视网膜相邻的双极细胞,重新接上 因感光细胞丢失而中断的视觉信号路径。选择 880 nm 波长是有意为之:它位于可见光谱之外, 患者残存的感光细胞看不到,因此不会与天然周边视觉串扰。 外部硬件由三部分组成:(1)配有 sun-lens 的眼镜,前向摄像头捕捉场景,面向眼睛的一侧放置 红外投影器;(2)口袋处理器,在把结果编码成 NIR 图案之前,执行实时变焦(最高 12×)、对比度增强 和 AI 辅助物体识别算法;(3)投影器,将图案化 880 nm 光精确打到视网膜下植入体上。只有被照亮的 像素会触发,因此图像分辨率由芯片上的像素数量决定。光伏芯片从入射光取电,不需要植入电池, 设备完全无线——相较 Argus II 等需要经皮导线的早期系统,这是关键优势。该技术最早由 Stanford University 的 Daniel Palanker, PhD 于 2005 年提出;其实验室完成了首次人体可行性试验 (NCT03333954、NCT03392324)。Science Corporation 后来在 2024 年 4 月从 Pixium Vision SA 收购所有 PRIMA 知识产权和临床试验后,将项目继续推进。 [CE001, CE002, CE003, CE004, CE005, CE006]

产品模块 / 资产矩阵
模块 / 资产目标用户 / 适应症状态 / 成熟度关键差异化尽调缺口
PRIMA Stim 植入物(视网膜下光伏芯片)内层视网膜保留的晚期 GA(AMD)患者CE 认证已提交(欧盟 2025 年 6 月);FDA PMA 审评进行中;预计 2026 年欧盟上市无线(无电池 / 线缆)、378 像素蜂窝阵列,首个可提供形状视觉的设备定价未披露;美国报销路径不清晰;批量制造爬坡尚未确认
PRIMA NIR 眼镜 + 口袋处理器手术后的 PRIMA 植入用户作为临床研究设备可用;商业形态待定880 nm NIR 投影、12× 数字变焦、AI 对比度增强、周边视觉同步保留最终商业设计(外形、耐用性、IP 保护)未公开披露
Science Eye(FlexLED + 光遗传基因疗法)RGC 完整的 RP 和干性 AMD 患者(约 1M 个视神经细胞)临床前;推迟到 PRIMA 上市后;动物安全性数据持续产出microLED 密度约为 iPhone 13 的 8×;随眼球移动,保持像素—细胞映射稳定;避开 PRIMA 所需的黄斑中央凹开孔首例人体时间表未确认;基因疗法监管路径复杂;尚无临床阶段牵引
NYX 神经记录 ASIC 系列(NYX、NYX 512)神经科学和 BCI 研究人员已面向研究商业销售;由 Science Foundry 制造低噪声、可编程,提供 512 通道选项,适合高通道数神经记录外部商业采用有限;主要支持内部 R&D 和研究合作伙伴
LUX microLED 驱动 ASIC 系列(LUX、LUX 2K、LUX 16K)光遗传学研究人员;未来 Science Eye 生产已面向研究商业销售;2K 主动和 16K 被动版本主动 2,048 像素和被动 16,384 像素选项,支持高密度光遗传刺激Science Eye 上市前,LUX 必须在 Science Eye 产量下完成商业资质认证
生物混合神经接口(嵌入神经元的薄膜电极阵列)神经损伤、运动障碍或 ALS 患者(长期)临床前;动物安全性研究(小鼠)2024 年发表;首个人体传感器试验筹备中神经元与宿主大脑形成突触连接;带宽有望达到十亿级突触,而非数百个电极通道活细胞植入物监管路径前所未有;批次神经元质量不确定;无人类安全性数据
Synapse API 和开发者平台神经接口开发者、学术研究人员开源;18 个公开 GitHub 仓库;最后提交 2026 年 6 月 26 日多语言(Python、TypeScript、C++)gRPC API;NDTP 协议;实验室自动化栈(Galago)采用指标未公开;主要用于内部和研究场景

成熟度评估基于截至 2026 年中的公司沟通、监管文件和临床试验登记。任何资产均未确认商业收入。Science Eye、生物混合和 Synapse API 的描述反映公司披露意图及公开研究成果;商业时间表未确认。

[CE001, CE002, CE024, CE025, CE026, CE027]
FE001: PRIMA 系统架构栈

分层展示 PRIMA 系统从外部可穿戴硬件到皮层视觉处理的链条,说明每一层的功能角色、关键组件, 以及 Science Foundry 的制造位置。

层级边界代表概念性的功能划分;物理解剖并不完全相同。光学耦合效率和精确像素电流输出取决于植入深度、 组织光学特性和眼镜对准情况——Science Corp 均未公开披露。

[CE001, CE002, CE004, CE007, CE020, CE039]

5.2 临床证据与手术流程

PRIMAvera 关键研究(NCT04676854)是 Science Corporation 支撑 PRIMA 的决定性临床证据基础。 这项开放标签、多中心、前瞻性、基线对照试验,在法国、德国、意大利、荷兰、西班牙和英国的 17 个中心,纳入 38 名 60 岁以上、晚期 GA 且基线视力差于 20/320(logMAR ≥ 1.2)的患者。 主要终点是从基线到第 12 个月,视力至少改善 0.2 logMAR,达到临床意义。32 名完成 12 个月随访的 参与者中,26 人(81%;95% CI 64–93;p < 0.001)达到主要终点。对 6 名失访者(三例死亡、 一例退出、两例不可用)做多重插补后,估计响应率为 80%(95% CI 66–94)。平均视力增益约为 5 行 ETDRS(25.5 个字母);单名患者最佳改善达到 12 行。32 名完成者中,27 人在第 12 个月能够 阅读字母、数字或单词。部分同时使用眼镜数字变焦和对比度功能的患者,达到相当于 20/42 的视力。 该研究由 Frank G. Holz(University of Bonn)牵头,José-Alain Sahel(University of Pittsburgh) 和 Daniel Palanker(Stanford)担任资深作者;论文于 2025 年 10 月 20 日发表于 New England Journal of Medicine(NEJM 394:232–242;DOI 10.1056/NEJMoa2501396)。Jacque Duncan(UCSF)在同期 NEJM 社论中称 PRIMA 是晚期 GA 中「第一种恢复视觉的治疗」。 植入流程采用标准玻璃体视网膜手术技术。外科医生先做三孔 23 或 25 gauge 睫状体平坦部玻璃体切除, 必要时诱导后玻璃体脱离,在黄斑中心凹颞侧做小视网膜切开口,再用专门定制注射器把 PRIMA 芯片送入 视网膜下空间。芯片居中置于萎缩的中心凹下方;内激光光凝固定视网膜切开口边缘;可执行液-气或气体交换。 没有外部电源线从眼球引出。术后四到五周,患者配戴 PRIMA 眼镜,并在低视力专家指导下开始结构化视觉康复; 通常需要数月训练才能达到峰值表现——类似人工耳蜗用户所需的听觉学习。 试验安全性负担显著但可管理。38 名患者中 19 人发生 26 起严重不良事件(50%)。最常见的是眼压升高 (6 起,涉及 23% 患者),其次是周边视网膜裂孔(5 名患者)、植入期间视网膜下出血(3 名患者)和 全层黄斑裂孔(3 名患者)。关键在于,所有 SAE 都归因于植入手术本身,或手术与设备共同作用; 没有 SAE 被归因于设备单独导致。几乎所有 SAE 都在两个月内消退且无后遗症。Data Safety Monitoring Board 正式认定,PRIMA 的获益超过其手术风险。 [CE008, CE009, CE010, CE011, CE012, CE013]

工作流 / 用例表
用户任务当前工作流(PRIMA 前)Science Corp 方案可量化收益(证据)局限
阅读印刷文字(书籍、标签、标牌)放大镜、盲文、屏幕阅读器、依赖他人PRIMA 眼镜提供数字变焦和对比度;口袋处理器编码大字符84.4% 的 PRIMAvera 参与者在 12 个月时可阅读字母、数字或单词(NEJM 2025)仅黑白;需要高对比度;最小字号受 100 µm 像素间距限制
环境导航(室内 / 室外)白手杖、导盲犬、持续陪同协助假体中央视觉与自然周边视觉融合;周边视觉零干扰所有试验参与者均确认假体视觉 + 自然周边视觉可同步存在仅中央视野(2×2 mm 芯片);缺少精细细节和颜色;障碍物检测有限
人脸和物体识别GA 患者没有功能性中央视觉时无法完成口袋处理器做场景处理和变焦;可部分识别形状部分患者报告日常活动独立性改善,并能阅读地铁标识和食品标签灰度软件尚未交付;人脸识别需要灰度;目前仅黑白
PRIMA 植入手术(视网膜外科医生)晚期 GA 既往没有可行手术治疗标准 PPV + 定制视网膜下注射器;无外部线缆;激光闭合视网膜切口由受训玻璃体视网膜外科医生执行;无经皮导线;保留 MRI 兼容性50% 患者出现 SAE(多为手术相关:IOP、视网膜裂孔、出血);必须由经验丰富的外科医生操作
植入后视觉康复N/A(该适应症此前无假体视觉设备)低视力专家制定结构化项目;在家用眼镜练习;训练变焦 / 对比度12 个月平均提升 5 行;2/3 患者达到中高用户满意度需要数月训练;并非所有患者都能熟练阅读;方案尚未标准化

用例反映 PRIMAvera 试验参与者体验(N=38,NEJM 2025)和公司沟通。真实商业化后的结果、康复成功率和外科医生学习曲线,会不同于受控试验条件。

[CE008, CE010, CE012, CE014, CE015, CE016]
FE002: PRIMA 患者流程 — 从诊断到视觉结果

PRIMA 患者体验的分步流程,从最初识别候选人,到手术植入、康复,再到 PRIMAvera 报告的 12 个月临床结果。 图中突出关键决策关口和植入后训练依赖。

手术步骤反映标准玻璃体视网膜技术与公司对 PRIMA 递送系统的描述。具体时间、填塞选择和术后方案因手术中心而异。 临床结果来自 PRIMAvera(N=38);真实世界结果可能不同于受控试验条件。

[CE008, CE010, CE011, CE012, CE013, CE014]

5.3 技术限制与下一代路线图

PRIMA 当前商业规格有四个结构性技术限制,限定了初始患者体验和可服务人群。第一,视觉只有黑白: 光伏像素输出是二元的,当前软件版本不会从场景中计算灰度。Palanker 表示灰度软件升级正在开发; 患者把人脸识别列为仅次于阅读的第二优先需求,而人脸识别需要灰度层次。第二,植入体 2 × 2 mm 的 面积只覆盖中央黄斑区;周边视觉不受影响,但以当前尺寸,单块芯片无法覆盖更大视网膜表面。 第三,100 µm 的像素间距带来粗颗粒的假体视觉;估计 100 µm pitch 设备在无数字变焦下的理论视力上限约为 20/500,配合 12× 变焦增强后可提升至大约 20/320–20/200 区间。第四,也是患者筛选中最关键的一点, PRIMA 需要内层视网膜仍能工作:PRIMA 刺激的双极细胞层必须完整。视网膜退化已经进展到内层视网膜萎缩、 或存在青光眼性神经节细胞丢失的患者,不符合条件。 目前处于临床前开发、并已在 Stanford 大鼠实验中测试的下一代 PRIMA 芯片,目标是 20 µm 像素间距, 每块芯片约 10,000 个像素。Stanford 的 Daniel Palanker 预计,20 µm pitch 设备可提供 20/80 的基线假体视力; 与眼镜电子变焦结合后,在部分条件下可接近 20/20 功能视力。无线光伏设计也支持未来原位升级: 理论上,植入体可通过再次视网膜下手术、以最小切口替换成更高分辨率芯片,不同于有线牵引设备。 多块芯片也可按拼贴方式植入,覆盖更大的萎缩区域。Science Corporation 还在把 PRIMA 扩展到新适应症: NCT07266584 是澳大利亚 Sydney Eye Hospital 的一项可行性试验,评估 PRIMA 用于 Stargardt disease 和 retinitis pigmentosa——两者都是累及黄斑的遗传性视网膜退化;预计入组 5 名患者,主要完成时间预计为 2029 年 3 月。早期 PRIMAvera 试验 38 名患者的样本量足以支持 CE Mark 申报,但在初始 GA 适应症之外, 若未来申请标签扩展或提交报销材料,可能会被视为有限。 [CE015, CE021, CE022, CE023, CE039, CE040]

路线图 / 发布 / 开发阶段表
日期 / 阶段功能 / 里程碑状态含义来源
2018–2021Stanford Palanker 首例人体可行性试验(NCT03333954 欧洲,NCT03392324 美国)完成;结果发表于 Ophthalmology(2020)、Nature Commun(2022)验证了光伏视网膜刺激机制和手术递送;构成 PRIMAvera 设计基础来源:science.xyz/technologies/prima/、Stanford Medicine
2025-06CE 认证申请已提交欧盟;FDA PMA 申请已递交已提交;监管审评进行中PRIMA 商业上市的首条可行路径;预计 2026 年欧盟获批;美国时间表不确定来源:science.xyz/news/primavera-trial-preliminary-results/
2025-10-20PRIMAvera NEJM 发表(DOI 10.1056/NEJMoa2501396);80% 有意义改善,N=38已发表;完整同行评审完成为监管申报建立临床可信度;触发超额认购的 $230M Series Cbusinesswire.com NEJM 公告、Stanford Medicine 新闻
2026-03NCT07266584 PRIMA Stargardt 病和 RP 试验在澳大利亚 Sydney Eye Hospital 开启招募中;约 5 名患者;预计 2029 年 3 月完成主要终点第二适应症概念验证;扩大 PRIMA 总可触达人群;距离获批仍需多年来源:clinicaltrials.gov/study/NCT07266584
2026(预计)欧盟 CE 认证批准和欧洲初始商业上市尚未获批;公司沟通中预计Science Corp 首个产品收入;进入欧盟市场后开始商业学习曲线insidebci.com Series C 报道;science.xyz Series C 新闻
2027+(TBD)下一代 PRIMA 芯片(20 µm 像素间距,约 10,000 像素);灰度软件临床前;Stanford 已在大鼠中测试 20 µm假体视力最高可达 20/80;灰度支持人脸识别;为 Gen-2 产品刷新铺路Stanford Medicine PRIMA 新闻;med.stanford.edu

时间线信息来自已发表临床试验记录、公司官方新闻和学术报道。未来里程碑(欧盟批准、下一代芯片)由公司预测或分析师推断,仍受监管和开发不确定性影响。

[CE008, CE009, CE021, CE041, CE042, CE043]

5.4 平台技术:Science Eye、LUX 与 NYX 芯片家族

PRIMA 之外,Science Corporation 正在开发更宽的神经接口平台;支撑硬件层与 PRIMA 商业化推进同步建设。 Science Eye 是 Science Corporation 的第二个视觉假体项目。PRIMA 用视网膜下光伏植入体刺激视网膜双极细胞; 与之不同,Science Eye 将一种 optogenetic gene therapy 与 Lightsheet 柔性薄膜 microLED 显示器结合: 前者由 adeno-associated virus(AAV)载体递送,靶向视神经中的 retinal ganglion cells(RGCs); 后者直接插入视网膜上方。基因治疗让 RGCs 对光敏感;microLED 阵列随后输出定制刺激图案。相较基于投影器的方案, 该路线的关键技术优势在于柔性薄膜会随眼球移动,保持稳定的像素到细胞映射;这是学习基础闪光以外更复杂视觉编码的 必要条件。Science Corporation 称 Lightsheet 显示器的像素密度约为 iPhone 13 Super Retina 显示屏的 8 倍。 Science Eye 面向 retinitis pigmentosa 和 dry AMD 患者:他们的感光细胞已经被破坏,但每只眼约 100 万个 RGCs 仍能工作。该项目已正式推迟到 PRIMA 在全球某地实现商业发布之后;动物安全性数据仍在收集,尚未披露首次人体试验日期。 Science Eye 与 PRIMA 被视为互补:PRIMA 解决高分辨率中央(中心凹)视觉;Science Eye 可覆盖周边视觉,未来配置中 两套系统可能由同一副眼镜驱动。 NYX 芯片家族提供平台的神经记录层:NYX 是一款低噪声、低功耗、可编程混合信号 ASIC,用于细胞外神经记录; NYX 512 是 512 通道全差分版本,为高通道数脑图谱实验优化。LUX 家族提供 optogenetic 应用所需的 microLED 驱动能力: LUX 是面向单通道 optogenetic 刺激的小尺寸驱动器;LUX 2K 是主动式 2,048 像素驱动器;LUX 16K 是被动式 16,384 像素驱动器,用于高密度刺激阵列。两个芯片家族均已面向研究用途商业销售,并由北卡 Durham 的 Science Foundry 制造。Synapse headstage 电子系统在神经探针与网络之间处理、存储和传输信号。Axon 家族包括薄膜和硅神经探针, 用于接触组织的电极阵列。 [CE024, CE025, CE026, CE027, CE028, CE029]

FE004: 产品成熟度与能力图谱

截至 2026 年中,Science Corporation 七个主要技术项目的成熟度和就绪度评估快照。 行为产品 / 项目;列评估成熟阶段、临床证据强度、监管状态、制造就绪度和关键差异点。 序数判断基于公开证据;未经公司验证。

成熟阶段为作者基于公开证据的评估。Science Eye 和 biohybrid 细节来自公司官方博客文章和 TechCrunch 报道; 公司尚未发布包含这些项目日期的正式产品路线图。

[CE001, CE021, CE024, CE025, CE028, CE029]

5.5 生物混合神经接口与开发者生态

Science Corporation 时间跨度最长的技术项目,是生物混合神经接口。传统 BCI 使用金属或硅电极,无法与神经元形成突触连接, 且通常会因胶质瘢痕在数年内退化。Science 的生物混合架构把实验室培养的神经元直接嵌入薄膜电极阵列; 这些神经元源自干细胞,并经过基因改造以表达光敏 opsins(optogenetic transduction)。植入后,这些神经元向宿主大脑延伸 轴突和树突,形成突触连接。一块播种 100 万个嵌入神经元的薄膜设备,理论上可与大脑形成 10 亿个突触, 带宽上限比金属电极阵列高出几个数量级。 当前生物混合设备在约豌豆大小的面积内塞入 520 个记录电极。Neuralink 采用穿透式电极;Science 的设备则置于脑表面 (皮层表面放置,不穿透),因此可能减少胶质损伤并延长设备稳定性。2024 年,公司发表 working paper,展示在小鼠中 安全植入并对皮层回路执行功能性 optogenetic 刺激。截至 2026 年 4 月,Yale Medical School 神经外科系主任 Murat Günel 已加入,担任人体植入项目科学顾问。近期计划是在已经接受大型脑手术的患者体内植入该设备的电极传感器 (不含嵌入神经元),例如因脑肿胀管理而接受 craniectomy 的患者,以评估安全性。Science Corp 表示,不计划为这些 初始的仅电极传感器试验寻求 FDA 批准,理由是设备不构成显著风险。含嵌入神经元的完整临床试验预计在 2027 年之后。 生物混合 R&D 团队由 CSO Alan Mardinly 领导,约有 30 名研究人员。 开发者生态方面,Science Corporation 在 github.com/sciencecorp 保持活跃开源存在。主要开发者工具是 Synapse API—— 一组 protocol buffer 定义,让 Python、TypeScript 和 C++ 客户端库能通过 gRPC 发现 Synapse 设备并与其通信。 synapse-python 仓库有 17 个 GitHub stars,并于 2026 年 6 月 26 日更新;更大的组织下有 18 个公开仓库,覆盖 Synapse API、Galago 实验室自动化栈、NDTP(Neural Data Transport Protocol)和实验应用示例。开源生态降低了 Science 芯片家族被研究用户采用的门槛,也在商业产品推出前制造开发者锁定。 [CE030, CE034, CE035, CE036, CE037, CE038]

技术 / 运行架构表
层 / 组件角色关键依赖技术风险
PRIMA 光伏像素阵列(晶体硅)将 880 nm NIR 光转换为电脉冲;刺激相邻视网膜双极细胞Stanford/Palanker IP(经 Pixium 收购);Science Foundry MEMS 制造100 µm 像素间距限制分辨率;患者内层视网膜必须完整;光伏效率约束刺激幅度
PRIMA NIR 眼镜投影仪(880 nm)向视网膜下芯片输送带图案的红外光,同时提供供能和图像数据定制光学和投影仪校准;用户头戴式可穿戴形态投影仪与植入物必须精确对准黄斑中央凹;环境 NIR 光源可能干扰;眼镜形态必须让患者可接受
口袋处理器 + AI 图像算法捕捉视觉场景;做对比度增强、变焦和场景编码;生成 NIR 投影图案软件 IP;电池供应;实时处理延迟要求图像编码质量限制感知视觉分辨率;AI 算法质量决定有效分辨率;电池续航约束使用时长
LUX microLED 驱动 ASIC 系列驱动 microLED 阵列做光遗传刺激(Science Eye 开发和研究用途)LUX 2K/16K ASIC 由 Science Foundry 制造;与 Science Eye Lightsheet 薄膜共同开发16K 通道像素驱动复杂;Science Eye 商业化前,芯片必须在 Science Eye 生产产量下完成认证
NYX 神经记录 ASIC 系列放大并数字化脑组织电极阵列采集的细胞外神经信号外部 ADC 生态;Synapse API 软件栈;探针到 ASIC 接口多种脑机接口应用中的噪声底表现;每个适应症都需要系统级验证
生物混合神经元—电子接口将实验室培养神经元与宿主大脑进行生物整合;形成突触连接,服务高带宽 BCI干细胞培养质量;监管接受活细胞医疗器械;生物神经元规模化制造全新监管路径;批次间神经元活性和行为差异;长期突触稳定性未知
Science Foundry MEMS 工厂(Durham, NC)内部制造光伏 PRIMA 硅阵列、NYX/LUX ASIC 和神经探针结构MEMSCAP 历史设备;$65M 扩建资本;洁净室 ISO/cGMP 资质目前仅试点规模;商业 cGMP 认证推进中;PRIMA 批量制造存在爬坡风险

架构细节来自 Science Corporation 官方产品页、Stanford Medicine 研究论文和 NCBiotech 设施报道。生物混合和 Science Eye 组件仍处 R&D 阶段;这些层的制造准备度尚未确认。

[CE001, CE003, CE004, CE007, CE028, CE029]

5.6 Science Foundry:制造与合规

Science Corporation 拥有并运营 Science Foundry,这是一座位于北卡罗来纳州 Durham 的 Research Triangle Park 的 MEMS 和 ASIC 制造设施。该设施源自 Science Corporation 约在 2022 年收购 MEMSCAP 的美国 MEMS foundry 资产和团队, 让公司获得了临床阶段神经接口公司中少见的内部微制造能力。Foundry 生产 Science 的专有 ASIC 芯片家族 (NYX 神经记录、LUX microLED 驱动器),并计划承担 PRIMA 光伏视网膜下阵列的生产。 2024 年 7 月,Durham County Commissioners 一致批准一项十年期、$930,000 的绩效型经济激励,支持 Science Corporation 计划中的设施扩建。扩建将在 Durham 园区新增约 57,000 平方英尺,预计需要 $65 million 资本投入,并预计创造 50 多个 高薪制造岗位。该设施目标是取得 ISO 9001:2015 质量管理体系认证,并符合 FDA current Good Manufacturing Practice(cGMP)。 NCBiotech 报道称,扩建将支持 Science 产品组合的制造,包括 PRIMA 植入体和先进神经接口技术。 垂直整合 MEMS foundry 给 Science Corporation 带来三项战略优势:(1)掌控 PRIMA 核心光伏硅阵列的供应链——降低对外部 foundry 合作伙伴的依赖,因为对方产能和 IP 容忍度可能不确定;(2)随着产量放大获得单位成本下降杠杆,因为自有 foundry 的增量芯片成本主要由材料和周期时间驱动,而不是叠加利润的外部定价;(3)作为美国本土 MEMS foundry,为精选外部 biotech 和 medtech 伙伴服务,产生非稀释性收入,并用外部用例验证 foundry 的质量体系。Foundry 当前能否以商业 PRIMA 量级投产, 尚未被公开确认;当前状态是试点规模制造。要为一款 CE Mark 批准医疗设备放大到商业级、cGMP 认证生产,仍是重大运营里程碑; 截至 2026 年中,Science 尚未公开达到该节点。 [CE031, CE032, CE033]

信任 / 质量 / 合规表
控制 / 认证 / 质量指标状态范围缺口
FDA 突破性设备认定2023 年 4 月授予(给 Pixium Vision);Science Corp 于 2024 年 4 月承接PRIMA,用于干性 AMD 导致的地理萎缩不等于上市批准;PMA 审评进行中;美国商业上市时间表未确认
欧盟 CE 认证申请2025 年 6 月提交;决定待定PRIMA,用于干性 AMD 导致的晚期 GA(PRIMA Stim 设备 + PRIMA Glasses 系统)截至 2026 年中尚未获批;预计欧盟上市取决于审批决定
独立数据安全监测委员会(DSMB)审查PRIMAvera 关键中期审查已完成;DSMB 认定收益超过手术风险PRIMAvera 试验(NCT04676854),38 名患者,17 个中心12 个月以后的长期安全性数据仍在收集;上市后监测设计未披露
ISO 9001:2015 / FDA cGMP 制造合规Science Foundry(Durham, NC)目标认证;状态为推进中PRIMA 设备制造和 ASIC 芯片生产认证未获公开确认;商业规模达成 cGMP 是美国 PMA 和欧盟商业生产的前提
不良事件报告和上市后监测NEJM 论文报告 19 名患者出现 26 起 SAE;均按方案处理;DSMB 监测PRIMAvera 临床试验阶段无上市后监测数据;取出植入物和长期设备失效率未知;商业药物警戒计划未披露

监管状态反映截至 2026 年中的公开文件和公司公告。Science Foundry 的 ISO 9001 和 FDA cGMP 认证是公司声明的目标;实际认证日期未获公开确认。

[CE008, CE009, CE016, CE018, CE031, CE033]
FE003: 关键依赖图 — PRIMA 商业化路径

有向无环图梳理 Science Corporation 要实现 PRIMA 商业可用、产生收入并让患者获得产品,必须同时收敛的关键依赖。 节点代表参与方、资产或关口;边展示依赖方向。

依赖边代表基于监管常规和公司沟通推导的逻辑前提。边的方向经过简化; 部分节点之间存在双向强化关系(例如临床证据和报销彼此影响、也彼此强化)。欧盟 AMNOG 和美国覆盖时间仅作示意。

[CE031, CE032, CE033, CE008, CE009]
Chapter 06

06客户

6.1 商业化前格局:界定 Science Corporation 的客户宇宙

截至 2026 年 6 月,Science Corporation 仍完全处于商业化前。公司没有传统意义上的付费客户;PRIMA 视网膜植入系统的所有当前用户, 要么是临床试验参与者,要么处在 compassionate-access 或登记背景下的患者,要么是收入前安排中的商业伙伴。这对临床阶段 medtech 公司并不反常,但对任何商业评估都是关键背景:今天的所有需求信号都是前瞻性的,并取决于尚未解决的监管结果。客户格局可拆成四类角色: (1)因 AMD 导致 advanced geographic atrophy、或遗传性视网膜退化且符合植入条件的患者;(2)作为临床守门人并最终成为商业采购方的 植入玻璃体视网膜外科医生和学术医院系统;(3)汇集并导流患者需求的患者倡导组织和登记系统;(4)foundry 生态伙伴——授权使用 Science 神经工程平台的早期 BCI 公司。每个细分都有不同采购路径、不同牵引证据、不同收入时间线。下方客户分层表完整映射这些维度; 客户旅程图则展示患者从疾病出现到长期使用设备依次经历的阶段,并标出 Science 必须先交付哪些商业基础设施,才能把患者需求转成收入。 [CU001, CU009, CU010, CU013, CU025, CU040]

按角色、用例和商业阶段划分的客户分群
分群买方 / 付款方角色用例当前规模收入 / 战略价值商业缺口
试验患者(晚期干性 AMD / GA)患者终端用户(未来通过付款方成为商业购买者)视觉恢复;阅读能力恢复;生活质量改善38 人入组(PRIMAvera);全球符合条件者达数百万人商业化前;今天无直接收入无付费关系;取决于监管和报销批准
植入手术玻璃体视网膜外科医生临床守门人和手术服务提供者手术植入;患者筛选;术后管理17 个欧盟中心;1 个美国中心;合计约 18 个临床合作点间接收入推动者;独立赚取手术费用需要专门视网膜下培训;并非普遍具备认证
学术医院系统 / 临床试验中心机构运营方和未来设备购买者R&D 合作伙伴;潜在商业设备买方18 个活跃临床合作中心无商业合同;仅试点阶段需要商业设备定价、保险代码和 GPO 合同
患者登记参与者(AMD、RP、Stargardt)潜在患者和需求信号未来 PRIMA 植入候选人;临床试验入组候选人入组人数未披露;全球开放收入前意向信号;尚无转化路径无商业定价;无报销代码;无市场准入协议
Foundry 生态合作伙伴(如 Neurosoft Bioelectronics)B2B 技术被许可方和共同开发伙伴面向非 PRIMA 神经适应症的 BCI 平台访问1 个具名合作伙伴(Neurosoft);其他为预计多年合同收入;金额未披露仅 1 个公开具名合作伙伴;管线和收入未披露
未来医院 / 医疗系统购买者(欧盟)设备买方和植入设施商业 PRIMA 设备采购和患者护理CE 认证后以 17 个欧盟中心为初始目标设备 ASP 可能在 $85K–$145K 区间(估算);报销不确定无商业标价;未建立欧盟报销代码
政府 / 国家付款方(NHS、CMS、GKV)报销权威;覆盖决策方PRIMA 手术的覆盖和支付对患者准入有全国性影响放量的关键前提;各地都尚未建立覆盖尚未启动全国覆盖决定、NCD 或 NICE 评估

客户分层和规模数据来自 Science Corp 新闻稿、ClinicalTrials.gov 以及截至 2026 年 6 月的分析师覆盖。公司尚无商业客户交易;所有收入值均为估算或零。

[CU001, CU002, CU009, CU013, CU025, CU040]
FU001: PRIMA 患者客户旅程图 — 从诊断到长期视觉使用

六阶段旅程展示晚期干性 AMD 患者从初始诊断、手术植入到长期设备使用的连续步骤, 以及每个阶段的 Science Corp 触点和采用障碍。

[CU014, CU015, CU016, CU017]

6.2 试验中心与植入医生作为入口客户

参与 PRIMAvera 试验(NCT04676854)的 17 个欧洲临床试验中心,是 Science 早期客户网络中最成熟的一层。这些中心覆盖法国、德国、 意大利、荷兰和英国,共纳入 38 名 60 岁或以上、患有晚期 dry AMD 和 geographic atrophy 的患者。首席研究者是 University Hospital Bonn 的 Dr. Frank Holz,他共同署名了 2025 年 10 月的关键 NEJM 论文。在英国,Moorfields Eye Hospital 是唯一试验中心,在 Moorfields 及 UCL Institute of Ophthalmology 高级玻璃体视网膜顾问 Dr. Mahi Muqit 领导下纳入 5 名患者。Dr. Muqit 称 PRIMA 芯片操作 可由「任何受训玻璃体视网膜外科医生在两小时内」安全完成;这一表述显然是在为广泛临床可扩展性定位。在美国,由 Dr. José-Alain Sahel 领导的 University of Pittsburgh Medical Center(UPMC)Vision Institute 是首个在临床可行性研究中植入 PRIMA 的美国中心,因此成为 任何 FDA 放行发布的美国锚点商业中心。Science Corporation 在患者登记公告中称,截至该公告,已建立 18 个临床合作关系;该数字涵盖 PRIMAvera 欧洲中心和正在扩张的美国可行性网络。监管批准后,这些试验中心很可能成为首批商业客户,从研究者发起的试验角色转为采购设备的 医院科室——前提是发布时已有报销基础设施。 [CU002, CU006, CU007, CU008, CU013, CU017]

客户采用轨迹与关键增长指标
指标数值日期来源类型置信度含义缺失分母
PRIMAvera 试验入组患者382023–2025监管(ClinicalTrials.gov)证明可规模化完成概念验证植入未披露筛查到入组的总比例
12 个月随访完成率32/38 (84.2%)Oct 2025官方新闻稿 + NEJM试验耐久性强;脱落率低未披露 6 名脱落患者的原因
达到主要终点(≥5 个字母增益)26/32 (81.3%)Oct 2025NEJM 同行评审达到主要终点;构成监管申报依据PRIMAvera 设计没有假手术对照组
居家设备功能使用(自报)27/32 (84.4%)Oct 2025NEJM 同行评审实验室外的真实采用得到确认未披露经过验证的客观使用日志数据
已建立临床合作(试验 + 注册登记)182025–2026Science Corp 官方全球临床足迹在为商业化铺路18 家机构并非全部已正式签约支持商业发布
具名 B2B foundry 生态伙伴1(Neurosoft)Feb 2026官方 BusinessWire 新闻稿B2B 收入机会仍处早期;管线未披露未披露 foundry 合同金额或收入
患者注册登记入组数未披露2025–2026Science Corp 官方有意向信号;但没有量化需求管线入组人数未公开报告
新适应症试验启动(NCT07266584,IRD)1 项新 IND 已提交 / 试验已启动2026ClinicalTrials.gov适应症扩展已经推进(RP、Stargardt、其他 IRD)机构数量和入组目标尚未发布

数值数据来自 ClinicalTrials.gov(NCT04676854)、2025 年 10 月 NEJM、Science Corp 新闻稿和 BusinessWire。置信度反映来源层级;中/低表示依赖公司披露,缺少独立验证。所有数值都处于研发和商业化前阶段;尚无商业采用指标。

[CU002, CU004, CU005, CU013, CU039]
具名客户证据表
客户 / 机构分层部署 / 用例状态已记录结果关键限制
Moorfields Eye Hospital(英国伦敦)临床试验机构(唯一英国 PRIMAvera 机构)视网膜下 PRIMA 植入;术后康复试验完成(5 名患者入组);预计后续可继续使用全试验读字率 84%;英国患者 Sheila Irvine 恢复阅读能力和日常独立性英国商业准入需要 MHRA + NICE + NHS England 批准;CE 标志后仍需 1–2 年
University Hospital Bonn(德国)牵头临床试验机构;主要研究者所在机构PRIMAvera 试验协调和入组;NEJM 论文牵头处于试验随访阶段;NEJM 论文发表于 2025 年 10 月Dr. Frank Holz 牵头关键 NEJM 论文;德国被定位为首个欧盟商业市场尚无商业合同;首例商业患者仍在等待 CE 标志批准
UPMC Vision Institute(美国匹兹堡)美国临床可行性机构;首个美国 PRIMA 植入中心PRIMA 临床可行性研究;在 PRIMAvera 分析中担任共同 PI活跃;美国首个为人类患者植入 PRIMA 的中心Dr. José-Alain Sahel 共同牵头 NEJM 分析;UPMC 被定位为美国锚点机构美国商业发布取决于 FDA PMA 批准和 Medicare 报销路径
Neurosoft Bioelectronics(日内瓦,瑞士)Foundry Ecosystem 伙伴(B2B 技术客户)获得 Science BCI 平台栈,用于耳鸣 / 癫痫神经设备多年期合同生效中(2026 年 2 月 20 日宣布)Neurosoft 探针已有 12 名既往人体患者;开发成本从 $75–100M 降至 <$5M只有 1 家具名 foundry 伙伴;未披露收入;不是 PRIMA 适应症
患者 Sheila Irvine(Moorfields 试验患者)单个试验患者(客户证据 / 患者结果)晚期干性 AMD 后视觉恢复;阅读和日常活动恢复已完成 PRIMAvera 试验;继续使用设备报告称能读填字游戏、罐头和处方标签;称植入物「带来很大不同」单个具名个体;不具统计代表性;不是商业交易

列举并不完整,只包括公开具名的机构和伙伴。PRIMAvera 试验覆盖 17 家欧洲机构;只有部分中心公开了研究者或患者证言。UPMC 之外的美国可行性研究机构尚未公开具名。Neurosoft 之外的 foundry 伙伴管线未披露。所有关系都处于商业化前阶段;没有设备采购交易。

[CU006, CU007, CU008, CU025, CU036]

6.3 患者登记与倡导:导流潜在需求

Science Corporation 于 2025–2026 年推出正式患者登记,覆盖确诊 age-related macular degeneration、retinitis pigmentosa 和 Stargardt disease 的个人。这个登记系统同时承担多项商业功能:生成未来 PRIMA 患者的合格管线,为扩展适应症的临床试验设计提供信息, 并在公司与患者社区之间建立直接沟通渠道。参与免费、尊重隐私且可撤回,参与者保留完整数据控制权。登记系统不仅跟踪最佳矫正视力等传统临床指标, 也跟踪 Science 认定最能体现患者获益的功能结局——阅读能力、人脸识别和导航。登记参与者可优先进入未来临床试验和新一代设备。Science Corp 称,全球约有 500 万人受 end-stage atrophic AMD 影响,目前没有能恢复视觉的疗法;这构成主要可服务患者池。资格门槛很严:患者必须 年满 60 岁,患有伴 geographic atrophy 的 advanced dry AMD,研究眼最佳矫正视力为 20/400 或更差,并确认存在中央暗点。这个狭窄的 资格漏斗,再加上保留周边视觉且没有其他混杂眼部疾病的要求,预计会使商业就绪人群远小于 AMD 大盘数字暗示的规模。患者倡导组织已经参与; Macular Society(UK)称 PRIMA 结果「令人鼓舞」,并表示正在密切关注英国监管批准和未来 NHS 可及性。 [CU003, CU009, CU010, CU011, CU012, CU014]

FU002: PRIMA 患者采用漏斗 — 从疾病负担到试验完成

漏斗展示 PRIMAvera 试验中,从全球晚期干性 AMD 患病人群,经过资格标准、试验入组、随访完成, 到达到主要终点的逐步收窄。

全球 AMD 患病率和适格患者估算来自 Science Corp 患者登记库公告和一般人群数据; 正式 PRIMA 特异性流行病学尚未发表。筛查患者数为估算。试验入组和终点数据来自 2025 年 10 月 NEJM。

[CU003, CU004, CU010, CU014]

6.4 商业采用路径:监管关口与报销缺口

Science Corporation 已同步向欧盟提交 PRIMA 的 CE mark 申请,并向 FDA 提交 IDE/PMA-equivalent;公司 2026 年 6 月患者登记公告确认了这一点。 行业分析师和新闻报道一致把德国视为预期首个商业市场,原因是其新设备监管框架以及 neurotechnology 早期准入批准的历史先例。市场普遍预期 CE mark 会在 2026 年中获批,但截至报告日期尚未公开宣布批准。英国方面,Dr. Muqit 确认,CE mark 之后可能还需要一到两年额外审批, 依次通过 MHRA、NICE health technology assessment 和 NHS England commissioning——这是高成本植入设备典型的三阶段守门流程。美国路径更长: PRIMA 拥有 FDA Breakthrough Device designation(继承自 Pixium 收购),可加快审评,但不能绕过完整 PMA 路径。关键在于,截至 2026 年 6 月, 美国没有 PRIMA 的 Medicare 报销代码、全国覆盖决定或地方覆盖政策。American Society of Retina Specialists 对 2026 年 Medicare 支付提案的审阅, 未发现 PRIMA 系统的特定代码;新型植入设备通常需要先发生 FDA clearance 事件,再分配 CPT/HCPCS 代码,并随后进入 CMS 审查,之后才开始任何有覆盖的 临床使用。该顺序可能在美国监管放行与实际可报销商业部署之间制造多年缺口。 [CU019, CU020, CU021, CU022, CU023, CU024]

试验队列留存、满意度与耐久性指标
指标数值分层置信度尽调问题
12 个月试验随访完成32/38 (84.2%)PRIMAvera 试验患者(欧盟,所有机构)评估 6 名脱落患者的原因;区分设备失效与患者退出
居家设备使用(12 个月自报)27/32(84.4%)将设备用于功能性任务PRIMAvera 试验患者(欧盟,所有机构)用客观使用遥测验证;未披露设备使用日志
患者证言满意度正面(1 名具名患者;仅定性)Moorfields 患者(Sheila Irvine);单个案例获取全部 38 名参与者的系统性患者报告结局量表
长期设备耐久性(超过 12 个月)尚未披露所有 PRIMAvera 试验患者试验随访持续到 2029 年;长期数据不可得
临床机构持续参与(注册登记承诺)18 家机构维持注册登记网络所有活跃临床合作机构确认各机构是否已正式承诺签署商业过渡协议

留存和满意度数据来自 PRIMAvera 试验 NEJM 论文(2025 年 10 月)和 Science Corp 患者注册登记公告。尚无商业 NRR、GRR 或队列续约数据;所有指标反映商业化前试验表现。患者满意度证据仅限一条已发布证言;尚未公开报告经过验证的 PROM 工具。

[CU004, CU011, CU016, CU034, CU036]
FU004: PRIMAvera 试验队列在随访节点的留存率

PRIMAvera 临床试验队列(n=38)在关键随访节点的时间序列留存率,以仍留在研究中的入组患者占比表示。

第 3 个月 97% 的留存率为估算值(未公布明确的 3 个月完成人数);12 个月留存率 84%(32/38)已由 NEJM 2025 年 10 月论文确认。NCT07266584(2026 年 IRD 扩展)因尚未报告入组情况,取值为 null。

[CU011, CU012, CU038]

6.5 Foundry 生态:Science BCI 平台的 B2B 客户

Science Corporation 于 2026 年 2 月 20 日宣布,与总部位于 Geneva 的神经数据公司 Neurosoft Bioelectronics 建立多年期 BCI Ecosystem 合作。 Neurosoft Bioelectronics 开发用于 tinnitus、epilepsy 和其他神经疾病的超软皮层探针;合作宣布前,其探针已用于十几名人体患者。在 Science Ecosystem 模式下,合作伙伴可获得 Science 完整的临床级神经记录技术栈——包括记录电子系统、神经探针、软件和应用工具——成本只是独立开发同等基础设施的一小部分。 Science 估计,从零搭建全栈临床级 BCI 平台需要 $75 million 到 $100 million,并耗时多年;Ecosystem 合作使首次人体试验能以不到 $5 million 的成本、 用更短时间启动。Science CEO Max Hodak 称 Neurosoft 交易是预期「许多」合作中的第一笔,把 Science Ecosystem 定位成类似半导体行业 foundry 服务的平台商业模式。 这些生态伙伴是不同于患者 / 医院渠道的 B2B 客户类别:他们购买 Science 工程基础设施的使用权,而不是 PRIMA 设备本身,并且瞄准 Science 内部项目之外的神经适应症。 生态交易的收入模式是多年期、数百万美元级合同,但具体财务条款尚未披露。下方客户证明矩阵对四类客户细分的证据质量作横向比较。 [CU025, CU026, CU027, CU028, CU029]

FU003: 客户证据质量矩阵 — 跨所有细分

截至 2026 年 6 月,对 Science Corp 四类主要客户的临床证据强度、患者结果具体性、监管状态和商业就绪度作跨细分比较。

[CU019, CU021, CU027, CU028, CU029]

6.6 采用壁垒、反向证据与集中度风险

多个结构性壁垒削弱了乐观需求叙事。第一,也是最重要的一点,截至 2026 年 6 月,PRIMA 收入为零:所有「客户」都是临床试验参与者或收入前伙伴, 尚未发生商业交易。第二,设备不良事件情况需要审视:PRIMAvera 试验中,38 名患者里 19 人发生 26 起严重不良事件(患者层面 SAE 发生率 50%), 包括视网膜裂孔、眼压升高、视网膜脱离和增殖性玻璃体视网膜病变。虽然 95% 事件在两个月内消退,但接受治疗的眼睛相比未治疗对照眼,geographic atrophy 面积增幅更大,引发了设备是否加速基础疾病的未决问题。第三,近期商业足迹高度集中在精英学术玻璃体视网膜中心。PRIMAvera 试验需要 17 个欧洲专科中心和 1 个美国中心;并非所有玻璃体视网膜外科医生都具备所需的视网膜下植入技能和术后康复基础设施。因此,广泛地理可及性会被培训和中心认证要求结构性限制, 即使监管批准后,该瓶颈仍会存在。第四,Science 的患者资格标准狭窄:候选者必须年满 60 岁,完全失去中央视觉,并保留残余周边视觉——这一画像会筛掉很多 AMD 患者。第五,美国商业路径在 FDA clearance 和 Medicare 报销同时拿下前都会停滞,而这个顺序可能远长于欧洲 CE-mark 时间线。合起来看,即使在乐观情形下, Science 的商业爬坡也会很慢,依赖少数专科中心,并且初期局限在报销环境最有利的单一欧洲国家市场。 [CU001, CU023, CU030, CU031, CU032, CU033]

扩张驱动因素与客户集中风险
扩张驱动因素集中风险潜在影响尽调路径
CE 标志批准 → 德国发布依赖欧盟单一国家;若 2026 年未获 CE 标志,将出现延迟高 — 若 CE 标志延后,整个欧洲商业化受阻通过监管文件确认 Science Corp 的 CE 标志申请状态
适应症扩展:RP 和 Stargardt(NCT07266584)新适应症仍处于入组前;尚无临床数据中等 — TAM 扩张有实质意义;但证据基础尚未建立跟踪 NCT07266584 入组和中期数据发布
美国 FDA PMA 批准取决于单一 FDA 审评决定;没有替代美国路径高 — 没有 FDA 批准和 CPT 代码分配,美国收入为零确认 Breakthrough Device 状态活动;跟踪 FDA 审评里程碑
Foundry 生态扩张(更多 BCI 伙伴公司)只有 1 家具名伙伴;平台收入高度集中中等 — B2B 收入未分散;第二家伙伴公告待定识别是否正在谈判更多生态交易
NHS / 全国支付方覆盖(英国、欧盟)英国需要 MHRA + NICE + NHS 委托采购;CE 标志后滞后 1–2 年高 — 数年内限制英国患者量和手术机构收入跟踪 NICE 评估启动;对比 iStent 或其他高成本眼科设备先例
手术培训中心认证限于学术玻璃体视网膜中心;社区机构被排除高 — 即便获批后,地理可及性仍受约束评估 Science 是否在搭建正式认证和湿实验室培训项目

扩张驱动和风险评估基于截至 2026 年 6 月的当前临床阶段、监管文件和行业先例,具有前瞻性。上述扩张路径均尚未转化为商业合同。所有竞争对比均为指示性;Science Corp 尚未发布排序或时间表指引。

[CU019, CU020, CU022, CU023, CU033, CU035]

6.7 图表

Chapter 07

07风险

7.1 监管批准风险:FDA PMA 与 EU CE Mark

Science Corporation 的近期商业可行性完全取决于两个并行监管结果:美国市场的 FDA Pre-Market Approval,以及欧洲市场的 CE mark 确认。两项批准都不保证, 且时间存在不确定性,可能使商业化比当前预期推迟一到三年。FDA Breakthrough Device designation(2023 年取得)可加快互动和审评排期,但不会降低安全性和有效性的 证据门槛。Class III 植入设备的 FDA PMA 路径通常在提交后需要 180 天实质审查,Agency 保留发出 Additional Information request 的权利,这可能把周期再延长 12 个月或更久。Science 于 2025 年 6 月向 BSI 提交的 CE mark 申请包含约 60 gigabytes 文档,目前处于 EU MDR 审查中;截至 2026 年 6 月尚未确认决定。 EU Medical Device Regulation(2017/745)要求 Class IIb 植入物开展上市后临床随访,Science 必须在批准前把该要求嵌入商业监测计划。FDA 与 CMS 于 2026 年 4 月联合宣布 RAPID Coverage Pathway,允许 Breakthrough Devices 在 FDA 批准同时获得 Medicare 覆盖,而不必等待通常批准后 6–18 个月的周期。 这显著缓解报销尾部风险,但要求 Science 主动与两个机构同时接触。Humanitarian Use Device designation 申请若获批,将允许 PRIMA 在每年 8,000 名患者上限、 利润限制较低的框架下商业化,为临床试验与完整商业规模之间提供桥接机制。考虑到通过 Pixium 收购取得的 Palanker Lab 专利组合,知识产权风险较低; 但视网膜下光伏像素设计空间的 freedom-to-operate 验证仍是开放尽调项。 [CR001, CR002, CR003, CR004, CR005, CR006]

监管 / 法律风险登记表
风险 / 规则司法辖区当前状态发生可能性严重性缓释措施剩余暴露尽调路径
FDA PMA 未获批或 RTF(Refuse to File)美国PMA 尚未提交;Breakthrough Device 指定有效;RAPID 路径可用(Apr 2026)低–中严重 — 阻断美国商业化Breakthrough Device 优先审评;补充 PRIMAvera 数据;接入 FDA 早期准入路径高 — 整个美国收入论点取决于正面决定确认 PMA 提交日期;审阅 FDA 反馈信和提交前会议纪要
EU MDR 2017/745 下欧盟 CE 标志延迟或被拒欧盟申请于 2025 年 6 月提交(约 60 GB 文件);BSI 审评中;决定待定低–中高 — 首笔商业收入延后 12–24 个月完整技术文件;60 GB 申报材料;提交中包含上市后临床随访计划中 — 欧盟时间表是公司目标;若美国 RAPID 路径推进,延迟可能但并非灾难性获取 BSI 里程碑时间表;确认 PMCF 计划已被纳入并接受
CMS 尚无针对视网膜下植入物的全国覆盖决定美国视网膜下假体尚无 CPT 代码或 NCD;RAPID 路径(Apr 2026)提供并行批准选项高 — 阻断商业支付方报销和医院采用主动加入 RAPID 路径;任命 CMS 联络人;启动覆盖前会议中 — RAPID 路径存在,但尚未在视觉假体上验证确认 RAPID 路径入组;审阅 CMS 提交前会议申请状态
专利自由实施:视网膜下光伏像素设计全球(美国、欧盟、日本、中国)核心 Palanker Lab 专利(如 US9592396B2)随 Pixium 收购进入 Science 组合;FTO 尚未公开确认中 — 挑战可能要求绕开设计或支付专利费通过 Pixium 获得 Palanker Lab IP;广泛专利组合覆盖光伏像素架构低 — 基础专利已持有;增量设计变更仍有风险委托正式 FTO 意见,覆盖像素几何、无线供电和芯片封装权利要求
Humanitarian Use Device(HUD)指定:利润上限约束美国HUD 申请已提交;截至 2026 年 6 月指定尚未确认;年度 8,000 名患者上限中 — 若获批,HUD 上限会限制规模和定价灵活性;FDA 也可能拒绝将 PRIMA 指定为 HUD 作为过渡;商业规模下仍按完整 PMA 利润规则规划低 — HUD 为可选路径;完整 PMA 无论如何仍是主路径向 FDA Office of Orphan Products Development 确认 HUD 申请状态

风险按严重性排序。状态反映截至 2026 年 6 月的公开证据,可能未覆盖保密的 FDA 或 BSI 往来。发生可能性和严重性是基于行业先例、监管申报历史和独立分析的定性评估。本登记表覆盖通过 FDA、GAO、EU MDR 和专利记录研究识别出的重大监管与 IP 风险;不包括次要司法辖区申报和出口管制事项。

[CR001, CR004, CR005, CR006, CR007]
FR003: 依赖关系图

有向无环图梳理 Science Corporation 的关键外部依赖;PRIMA 要商业化上市,必须维持或解决这些依赖。节点代表参与方、资产或关口; 边表示依赖方向,终点指向商业化患者可及。

[CR001, CR004, CR006, CR024, CR035]

7.2 临床安全性与疗效持久性风险

PRIMAvera 关键试验报告,38 名患者中 19 人发生 26 起严重不良事件——发生率 50%——其中四起被归为重度:黄斑裂孔、视网膜脱离、增殖性玻璃体视网膜病变和 眼压升高。26 起 SAE 中有 21 起发生在植入后前两个月内,更像手术阶段事件谱,而不是慢性设备失效模式。Science Corp Data Safety Monitoring Board 审阅 24 个月数据后建议试验继续且不修改方案,这是 DSMB 认为获益-风险平衡可接受的信号。不过,任何 FDA 审评都会重点看两个具体信号。第一,24 个月内,PRIMA 植入眼的 geographic atrophy 面积增加 8.5 mm²,同侧对照眼为 2.5 mm²,差异具有统计学意义;Science 将其归因于进入视网膜下空间的手术创伤,而非光伏芯片本身。 尽管视力和读字母收益在这一解剖进展下仍然维持,该机制仍需在 PMA 申报中解释。第二,任何商业视网膜下光伏设备都没有超过 36 个月的长期植入耐久性数据。 PRIMA 第二代芯片设计提升了像素密度和生物相容性,可能实际上把耐久性时钟相对于临床前或更早队列数据重新归零。相邻 BCI 开发中也能看到类似风险: Neuralink 首次人体试验在 2024 年 5 月的初始植入中记录到导线从脑组织回缩,说明长期的生物-非生物界面稳定性仍是植入式神经设备共同面对的开放挑战。 Science 的视网膜下放置方式在机械上不同于皮层阵列,但数十年维持稳定、生物相容界面的根本挑战是相同的。任何 FDA 针对 GA 进展或植入耐久性的 Additional Information request,都可能给审评时间线增加 12–24 个月,并要求补充试验数据。 [CR008, CR009, CR010, CR011, CR012, CR013]

FR001: 风险热力图

风险热力图按发生可能性和影响严重度评估 Science Corporation 的 12 项关键风险。右上象限风险(高可能性 + 高影响)可能打破投资论点;左下象限风险属于监测项。每项风险均标出缓释成熟度。

[CR008, CR009, CR016, CR017, CR027, CR031]

7.3 报销与商业采用风险

截至 2026 年 6 月,美国 Centers for Medicare and Medicaid Services 没有针对植入式视网膜下假体的专门 CPT 代码、覆盖决定或支付框架。私人支付方通常跟随 CMS:Anthem Blue Cross 的 DME.00052 政策已经把 brain-computer interface 康复设备列为试验性且非医学必需,无论 FDA 审评状态如何。GAO 2025 BCI 技术评估 (GAO-25-106952)发现,BCI 开发者历史上很难与 CMS 就其设备接触;临床试验中接受植入的患者在试验资金结束后面临设备 explantation——这是保险方和医院决定是否采用 新型植入平台时会权衡的结构性风险。2026 年 4 月 RAPID Coverage Pathway 直接回应这一问题,允许 Breakthrough Devices 同时获得 FDA 批准和 Medicare 覆盖决定; 但该路径很新,尚未在任何植入式视觉假体上测试过,并要求 Science 在预提交阶段主动与 FDA 和 CMS 双线接触。任何公开文件中都没有 PRIMA 商业定价,使外部无法判断 Science Corp 的目标价位是否能被医院、HTA 机构或国家支付方接受。人工耳蜗类比提示设备成本可能为 $30,000-$100,000;视网膜特定支付方的怀疑可能把可接受上限压得更低。 在欧洲,CE mark 之后,Science 必须逐国走 AMNOG 和 HTA 流程取得报销——对于新型植入设备,这一过程通常需要 12–36 个月。CMS 不行动造成的下游私人支付方真空会放大该风险: 没有美国覆盖信号,欧洲支付方在设定报销费率时没有先例可参考,Science 的商业发布时间线将取决于哪个市场的支付基础设施先成熟。 [CR016, CR017, CR018, CR019, CR020, CR021]

7.4 制造、运营与质量体系风险

Science Corporation 正在北卡罗来纳州 Durham 建设制造工厂,2024 年宣布投入 $65 million,目标是在生命科学制造领域新增 150+ 个岗位。PRIMA 芯片是一种光伏 CMOS/MEMS 器件,制造依赖半导体晶圆代工流程;Science 的自有晶圆代工业务(向外部 BCI 客户提供定制芯片制造)承担双重角色:既分散收入来源,也是 PRIMA 阵列的核心制造平台。21 CFR Part 820 质量体系法规及其国际对应标准 ISO 13485:2016 要求植入式器械制造商具备完整的设计控制、工艺验证记录和纠正措施体系。据称,Science 的 CE 认证提交材料约包含 60GB 质量体系文件,说明质量管理已推进到较成熟阶段——但本章研究未发现 Durham 工厂已获得 ISO 13485 认证的公开确认。若上市前检查发现制造不符合项,可能触发停产或要求提交 PMA 补充材料,商业化上市或被推迟 12-24 个月。CMOS/MEMS 制造供应链集中,尤其是让 PRIMA 区别于标准 CMOS 的专用光伏像素工艺,构成另一项运营风险;自有晶圆代工能力尚未完全化解该风险。在 PRIMA 商业爬坡的关键阶段,硬件代工业务的外部客户也可能与 PRIMA 争夺产能和工程人才。 [CR023, CR024, CR025, CR026, CR027, CR028]

运营 / 质量 / 安全风险登记表
失效模式发生可能性严重性缓释成熟度剩余暴露未解决缺口
Durham 设施 ISO 13485 不符合项高 — 可能触发停产或 PMA 补充部分 — 60 GB CE 文件存在;ISO 13485 认证未确认高 — 未公开确认认证Durham 设施 ISO 13485 认证状态未公开确认
CMOS/MEMS foundry 供应链中断低–中高 — 光伏像素阵列依赖单一来源制造部分 — 内部 foundry 降低但未消除第三方依赖中 — 内部产能有帮助;专用 CMOS 工艺可能仍需外部晶圆厂未识别专用光伏工艺的第二 foundry 选项
手术培训规模瓶颈中–高中 — 即便监管获批,也会限制商业发布速度低 — 未公开宣布正式外科医生培训项目高 — 流程复杂性限制机构快速认证外科医生培训课程和认证路径尚未公开建立
上市后监测失败:商业使用中未报告 SAE高 — 监管后果;患者伤害;报销损失部分 — 患者注册登记已启动;试验中 DSMB 活跃中 — 注册登记基础设施存在,但商业监测框架未验证商业规模 PMS 方案尚未发布或经过审查

缓释成熟度评估分为高(完整系统已到位)、部分(已有部分措施)或低(无已确认措施)。所有评估均基于公开证据;内部质量文件可能改善图景。

7.5 战略集中、关键人物与平台分心风险

截至 2026 年 3 月,Science Corporation 约有 150 名员工,却同时运营四条不同项目线:PRIMA 视网膜植入体、生物混合神经科学、硬件晶圆代工,以及 Vessel(可穿戴 BCI 中枢)。PRIMA 是唯一已有临床数据且具备近期收入路径的项目;其他项目仍处于临床前或收入前阶段。这样的集中度意味着,一旦 PRIMA 出现监管延误、临床挫折或商业失败,公司的主要收入论点就会消失,且没有近期替代支点。Max Hodak 是 Neuralink 联合创始人,也是 Science Corp CEO,是公司最主要的公众面孔和首席战略设计者。TIME Magazine 将 Hodak 列入 2026 TIME100 Health 榜单,凸显其个人影响力——但公司尚未披露继任计划和关键人物保险安排。若 Hodak 或首席科学顾问 Jacques-Alain Sahel 离职(Sahel 是全球领先的视网膜假体研究者之一,也是关键 PRIMA 临床论文的共同作者),很可能削弱监管可信度、临床网络获取能力和投资人信心。Yale 合作者 Murat Günel 的评论显示,生物混合脑传感器项目目标是在 2027 年前完成人体首次植入;该项目科学上很有吸引力,但需要专门的工程和临床试验基础设施,会与 PRIMA 争夺人才和管理层注意力。Science Corp 累计融资 $490M,估算年烧钱 $20-30M,意味着 7-8 年 runway——足以覆盖 3 年监管周期——但商业化所需资本强度,包括外科医生培训、上市后监测基础设施和美国设施建设,可能压缩这一估计。Vessel 平台又增加了第四条监管和商业路径,在 PRIMA 迫切需要集中执行的阶段进一步稀释组织焦点。 [CR029, CR030, CR031, CR032, CR033, CR034]

伙伴 / 依赖风险登记表
依赖项交易对手角色集中度失效情景严重性缓释措施剩余暴露
FDA 监管批准决定US Food and Drug Administration(美国监管机构)美国市场商业授权关口单一 — FDA PMA 没有替代不可批准信或 RTF;要求补充数据严重Breakthrough Device 指定;RAPID 路径;提交前会议高 — FDA 是唯一批准方;无法绕过
BSI / 欧盟公告机构 CE 标志BSI(British Standards Institution)欧盟市场商业授权关口单一 — 申请已提交给 BSI被拒或要求补充临床数据60 GB 文件;PMCF 计划;参与 CE 标志流程中 — 若 CE 标志早于 FDA 到位,欧盟将成为首个市场
Jacques-Alain Sahel(首席科学顾问)个人临床可信度;视网膜假体专业能力;监管通道高 — 专长独特;PRIMA NEJM 论文重要共同作者离任会削弱监管网络和临床可信度未公开该顾问角色的继任计划高 — 未识别替代人选
Neurosoft Bioelectronics(BCI 生态伙伴)Neurosoft BioelectronicsFoundry 收入;生态验证低 — 只有具名伙伴;但 foundry 还有其他潜在客户伙伴破产或技术转向低 — foundry 已多元化多年期伙伴协议;并非只依赖 Neurosoft低 — Science foundry 服务多个客户

交易对手依赖按集中度评级(高 = 单一来源,低 = 可替换)。FDA 和 BSI 是结构性垄断;个人依赖为定性判断。

人员 / 执行风险登记表
角色 / 职能依赖或缺口发生可能性严重性缓释措施尽调路径
Max Hodak(CEO、联合创始人)唯一具名高管领导者;主要公众面孔;战略架构;投资者关系低 — 近期离任可能性低但未披露严重 — 若流失,可能冲击投资者信心并带来融资风险未公开继任计划或 CEO 保险索取继任计划、雇佣合同条款和关键人保险文件
Jacques-Alain Sahel(首席科学顾问)世界级视网膜假体研究者;NEJM 共同作者;监管可信度高 — 流失会削弱监管申报质量和临床试验网络准入未识别公开继任者;顾问关系可能并非排他确认顾问安排的合同条款;明确备用临床科学领导层
外科培训和商业化团队缺口——未公开宣布商业规模的外科医生培训计划高——没有受训外科医生,不管监管结果如何,商业发布都会卡住17 个 PRIMAvera 欧盟站点提供首批受训骨干;美国 UPMC 站点提供锚点确认商业培训路线图,以及 Year 1 发布目标外科医生认证产能
监管事务和市场准入团队缺口——FDA PMA 和 CMS RAPID 沟通所需团队规模与构成尚未公开确认高——监管失误可能让 PMA 推迟 12–24 个月假设已有 Breakthrough Device 顾问和法律顾问,但未确认审查监管团队人数、FDA 预提交会议记录和 CMC 过往记录

人才风险天然不确定;评估基于公开披露以及未见证据的缺口。所有严重性判断均假设 2026–2028 年近期商业化时间线; 若时间线拉长,紧迫性下降,但接班风险上升。

FR002: 风险传导图

有向无环图展示 Science Corporation 的主要风险如何传导至商业、财务和战略结果。每条边代表一条因果或后果传导路径。

[CR031, CR032, CR033, CR034, CR027]

7.6 行业先例、结构性风险与终止标准

视网膜假体行业即便完成临床概念验证,也持续出现商业失败。Second Sight Medical Products 在 2019 年放弃 Argus II 仿生眼产品线,留下 350 多名患者使用缺乏支持的植入体;Barbara Campbell 的案例成为典型警示——她的 Argus II 在纽约地铁站内失效,使她在本已危险的环境中突然失明,说明医疗科技公司若退出却没有器械连续性计划,会给患者带来什么后果。Pixium Vision 在 2024 年进入 Paris Commercial Court 清算程序,并被 Science 以约 €4 million 收购,说明即便已有 CE 认证的器械,如果缺乏报销支持,也无法产生足够商业收入支撑独立公司。Retina Implant AG 在开发 16 年后停止运营,理由是监管和报销环境不利于创新。Science Corp 的 PRIMA 设计与这些前代产品有根本差异——它是视网膜下光伏刺激,面向仍有部分视力的患者,而非完全失明者;没有外部导线,手术技术也显著更简单——但支付方怀疑、长期耐久性证据有限、可服务患者规模小等结构性风险仍然相同。GAO 将 BCI 技术中的伦理和隐私风险列为该行业 12 项最紧迫政策挑战之一,包括数据安全、认知自由和「神经监控」风险;PRIMA 连接到可穿戴 AI 处理单元并持续捕捉视野数据,使这些担忧更强。机构支付方的两难进一步放大这些结构性风险:商业收入有限会压低正向覆盖决定所需的临床证据基础,而覆盖缺口又阻止产生这些证据所需的商业化。可监测的终止标准包括:(1) FDA 拒绝受理 PMA 或发出不可批准函;(2) RAPID pathway、CE 认证或 CMS 覆盖决定延迟到 2029 年之后;(3) PRIMAvera 延长期数据中出现第三类严重 SAE;(4) Hodak 离职且未披露继任计划;(5) Series C 后 24 个月内无法以持平或上轮估值融资完成 Series D。 [CR037, CR038, CR039, CR040, CR041, CR042]

缓释措施和终止标准表
风险可监测触发器阈值 / 事件行动含义
FDA PMA 被拒FDA 发出不可批准函或拒绝受理 PMA任一阶段收到 RTF 或不可批准函论点破裂——重估美国收入时间线;判断缺陷是否可补救;考虑重新接入 RAPID
报销失败FDA 批准后 24 个月,CMS 拒绝启动 NCD 或 RAPID 路径批准后 24 个月仍无 CMS 覆盖决定,也无商业保险采纳重大下行——美国商业收入被堵住;重估欧盟优先的商业化策略
临床安全升级第三类严重 SAE 出现,或 DSMB 建议暂停试验;36 个月 GA 加速超过对照组 3 倍新的严重 SAE 类别或 DSMB 暂停;36 个月数据中 GA 面积比 >3×论点破裂——立即通知投资人;PMA 可能延迟;评估设计变更
资本:无法完成 Series D 融资Science Corp 在 2026 年 3 月 Series C 后 24 个月内,无法以平轮或上轮估值完成 Series D到 2028 年 3 月仍无 Series D 条款清单,且投后估值 ≥$1.5B严重——显示投资人信心流失;现金消耗见底加速;考虑过桥或合作
关键人物离职Max Hodak 宣布离职,或在未披露接班计划的情况下被替换 CEOHodak 离职公告重大——重估领导层连续性;短期估值大概率受冲击;审查接班安排

触发器可监测、基于事件,而不是概率预测。阈值设在投资论点必须被明确重估的水平。终止标准彼此独立: 任一触发器出现,都足以启动重估。

7.7 附录

Chapter 08

08估值

8.1 估值背景与轮次条款

Science Corporation 于 2026 年 3 月以 $1.5 billion 投后估值完成 Series C,融资 $230 million。公司称该轮超额认购,这一需求信号意味着公司有能力融更多钱,或可以拒绝部分资本。投资方包括前几轮全部五家重复参与者:Lightspeed Venture Partners、Khosla Ventures、Y Combinator、IQT (In-Q-Tel) 和 Quiet Capital。投资人全为老股东,削弱了新领投方可能带来的独立验证,但也说明一批成熟 deeptech 投资人持续看好公司;他们已近距离观察公司的进展。自 2021 年成立以来,Science 累计融资约 $490 million,在全球私有 BCI 公司中资本化程度仅次于 Neuralink。上一轮融资事件是 2025 年 4 月由 Khosla 领投的 $104 million 可转债;该票据的转股上限、折扣率和转股机制均未公开披露。票据是在 Series C 交割时转股,还是仍作为未偿义务影响有效投后估值计算,目前未知。完整股权结构——包括各投资人持股比例、优先清算权层级、参与权和清算瀑布——尚未公开披露;没有 NDA 保护下的访问权限,就无法独立核验 $1.5B 对每类利益相关方的经济含义。在整体 medtech VC 交易量从 2025 年峰值降温的季度里完成超额认购,是一个正向信号;但缺少独立定价纪律(无新领投方、无披露估值方法)限制了价格标记的信息含量。 [CV001, CV002, CV006, CV007, CV008, CV009]

建议摘要表
维度评估投资含义
建议跟踪 / 继续研究监测 CE Mark 结果和欧盟商业发布;在 $1.0–$1.2B 入场估值或 CE Mark 后数据出现时重估
信心临床证据强;商业执行不确定;前代赛道失败率仍未解决
风险评级监管结果二元;收入为零;赛道基准中 3 个商业成功案例为 0
估值立场偏高$1.5B 已计入部分牛市情形兑现;按当前情景概率,期望价值约 $1.0–$1.25B
上行触发器CE Mark + 德国发布 + RAPID若 2026 年下半年获得 CE Mark、欧盟爬坡、2028 年前获 FDA Breakthrough,牛市情形为 $2.5–$3.5B

建议仅基于公开证据;未披露的股权结构和可转债条款可能显著影响有效入场经济性。估值立场采用按阶段调整的 medtech 逻辑, 而非 SaaS 倍数。

[CV001, CV030, CV037]
FV001: 推荐逻辑

从临床验证、监管势能到先发位置,证据链最终指向 TRACK 建议;同时展示负面因素(前代公司失败、报销缺口、估值溢价)如何约束最终判断。

[CV001, CV005, CV012, CV022, CV030, CV037]

8.2 投资论点与反论点

Science Corporation 的多头逻辑建立在临床、监管和战略因素的连贯对齐上。PRIMA 关键试验结果发表于 NEJM(17 个欧盟和美国站点、47 名患者中,80–84% 实现有意义的视力改善),是植入式视觉恢复器械迄今公开发表的最高质量临床证据。公司于 2025 年 6 月向 BSI 提交 CE Mark 申请,文件约 60GB;若 2026 年中获批,Science Corp 将成为全球第一家让产品上市的 BCI 公司。在竞争激烈的平台市场中,这种先发溢价可以支撑高于纯 DCF 基本面的估值。FDA Breakthrough Device designation 让公司获得更强 FDA 互动;叠加 2026 年 4 月 RAPID Coverage Pathway,形成 FDA 批准与 Medicare 覆盖同步推进的机制,移除了历史上获批后 12–18 个月的报销缺口。自有 MEMS 晶圆代工厂带来制造差异化,也提供其他视网膜假体公司从未具备的平台可选性。反论点同样扎实。Science Corp 在全球仍处于收入前阶段。三家获得 CE 认证的前代视网膜假体公司——Second Sight (Argus II)、Pixium Vision (Prima Alpha/PRIMA) 和 Retina Implant AG (Alpha IMS)——均获得监管许可却商业失败,整个行业合计消耗了数亿美元投资资本。Anthem 2026 年政策将所有人工视网膜器械归为研究性且非医学必要,代表 Science Corp 建立商业存在前,美国私营支付方的当前默认立场。CE Mark 后,欧盟报销仍需逐国 HTA 流程,关键市场通常耗时 12–36 个月。生物混合神经接口、Vessel 和晶圆代工业务带来的平台分心风险,在需要集中执行 PRIMA 的阶段分散管理层和资本。2025 年 VC 周期中的私募市场估值膨胀,加上全由老股东超额认购,也让人质疑 $1.5B 标记反映的是基本价值,还是周期高点的乐观定价。 [CV013, CV014, CV015, CV016, CV017, CV018]

论点 / 反论点表
支柱论点(牛市)反论点(熊市)什么会改变判断
临床证据NEJM 论文发表,47 名患者、17 个站点中 80–84% 患者达到成功标准;这是同等规模下首个经同行评审的 BCI 视觉恢复数据24 个月 GA 进展(植入眼 8.5 mm²,对侧眼 2.5 mm²);SAE 率 50%;36 个月之后没有长期耐久性数据长期疗效数据(48+ 个月);商业规模下 SAE 率低;DSMB 继续建议推进
监管路径CE Mark 已于 2025 年 6 月提交;FDA Breakthrough Device;2026 年 4 月 RAPID 路径可用;首个 BCI 可能在 2026 年中上市EU MDR 下 CE Mark 决策不可预测;FDA PMA 180 天 + 可能的 AI 延期;Breakthrough 不降低证据门槛CE Mark 批准确认;FDA 预提交会议纪要显示没有额外试验要求
市场机会全球 5M+ GA/AMD 患者;可扩展至 Stargardt/RP;BCI 市场从 $3.2B(2025)增长到 $12.9B(2034)可及市场受报销限制;仅限适合手术的患者;基因疗法在适应症上竞争CMS 对视网膜下假体发布 NCD;CE Mark 后 12 个月内,≥2 个欧盟市场确认报销
竞争位置首个商业化 BCI 植入物;市场上没有可比视觉恢复产品;MEMS 代工厂形成差异化Neuralink 通过 Blindsight 项目用皮层植入物瞄准同一 AMD;非侵入式光遗传疗法可能削弱手术意愿Neuralink Blindsight 在任何共同适应症上早于 PRIMA 商业发布
资本效率为 CE Mark 阶段临床资产融资 $490M;自有代工厂带来垂直整合优势在融资 $490M 后仍处收入前;Second Sight 花费 $300M+,峰值年收入低于 $5MCE Mark 批准后 18 个月内,年化收入 >$20M
估值纪律$1.5B vs Synchron 约 $1B,且临床证据更少;CE Mark 临近可部分支撑溢价$1.5B 已把部分牛市执行计入价格;按情景概率,期望价值约 $1.0B;存在私募市场标记膨胀风险以 $1.0–$1.2B 入场(二级或下一轮),可把期望价值推到正区间

所有评估基于截至 2026 年 6 月的公开证据。公司内部财务、股权结构和管线预测可能显著改变各支柱判断。 本表仅供参考,不构成投资建议。

[CV005, CV012, CV015, CV017, CV018, CV019]

8.3 可比估值分析

Science Corp 应采用按阶段调整的 medtech 逻辑,而不是 SaaS 倍数。2026 年商业化 medtech 硬件公司 EV/Revenue 倍数为 3.5–5.5x,EV/EBITDA 倍数为 11–14x(据 healthcare.digital 的 2026 M&A 矩阵和 Equidam 的 2026 行业 EBITDA 调查)。若假设 2027–2028 年首个完整年度欧盟收入为 $50M–$80M、毛利率 35–40%,这些倍数对应的企业价值约为 $175M–$440M,远低于 $1.5B。估值溢价由三类非倍数因素支撑:(1) BCI 视觉恢复中的先发和赢家多拿动态;(2) 生物混合、晶圆代工和 Science Eye 项目带来的平台可选性;(3) Bain 跟踪的 2025 年 medtech M&A 活动达到 $97.6 billion,战略买家为品类领先器械支付 15–25x EBITDA。BCI 同业集提供最直接相关的可比:Neuralink 在 Series E 后估值 $9.6B(2025 年 5 月,融资 $656.5M,为 Science Corp 估值的 6x),Synchron 在 Series D 后约 $1B(2025 年 11 月,累计融资 $345M),Precision Neuroscience 在 Series C 后约 $500M(2024 年 12 月,累计融资 $155M)。Science Corp 累计融资 $490M、估值 $1.5B,在资本效率上有利:Synchron 融资 $345M 后隐含约 $1B 投后,约为累计融资的 2.9x;Science 的 $1.5B 对 $490M 为 3.1x——几乎相当,但 Science Corp 的临床证据明显更强,近期商业化路径也更清晰。眼科 M&A 可比的参考价值较低,因为它们涉及已有收入的平台(Cencora 以 $4.6B 收购 Retina Consultants of America,对应 15–23x EBITDA),但它们确认,视网膜领域战略买家愿意为差异化平台支付高倍数。历史视网膜假体失败案例(Second Sight 峰值市值约 $100M;Pixium 约 €4M 被收购;Retina Implant AG 停止运营)定义的是该行业的下行情景,而不是退出倍数。 [CV003, CV004, CV005, CV023, CV024, CV025]

可比估值表
公司阶段(2026 年 6 月)已融资本报道 / 隐含估值依据对 Science Corp 的相关性关键局限
Neuralink临床阶段(3+ 人体试验,20+ 植入,商业化前)生命周期累计 ~$1.3B$9.6B(Series E,2025 年 5 月)一级市场:披露投后估值的 $656.5M Series E;二级市场隐含 $14–40B最接近的侵入式 BCI 同业;为平台型 BCI 估值设定天花板;Science Corp 约为 Neuralink 估值的 16%适应症完全不同(运动 / 光标);Neuralink 估值由 Musk 品牌溢价和二级市场投机驱动
Synchron关键试验阶段;血管内方案;商业准备推进中总计 ~$345M约 $1B(Series D 后,2025 年 11 月,隐含)Forbes:"about a billion dollars" post-$200M Series D;未正式披露最直接的阶段可比(CE Mark 阶段、融资规模相近);Science 临床数据更强适应症不同(瘫痪 / 运动,不是视觉);公开临床证据少于 Science
Precision Neuroscience早期人体测试;硬膜下 ECoG;IDE 批准待定总计 ~$155M约 $500M(Series C 后,2024 年 12 月,隐含)多个分析师估计;未披露;基于 Series C 金额和阶段可比公司显示 $155M 总融资可隐含 $500M;Science 融资高 3 倍,证据强度也相应更高关键试验前;适应症不同;估值估计未确认
Second Sight Medical Products(Argus II,前代公司)商业阶段失败的视网膜假体(历史案例)累计投入 >$300M峰值市值约 $100M(NASDAQ: EYES);实际清算价值 $0上市公司文件;股价数据;Fierce Biotech 负面报道定义赛道下行:获 FDA 批准、获 CE Mark 的视网膜植入物,有专用 CMS 代码,收入仍失败历史案例;Science PRIMA 的临床结果显著优于 Argus II
Pixium Vision(PRIMA 前身)破产(清算,2024 年 4 月);被 Science Corp 收购投入 €50M+Science Corp 收购价约 €4MScience Corp BusinessWire 收购公告;MarketScreener 记录最直接的技术先例;倒闭时 CE Mark 待批;失败点在商业化,不在监管Science Corp 资本更多、临床数据更好,上市路径强于 Pixium 当年
Retina Implant AG(Alpha IMS,前代公司)停止运营(2019);设备曾获 CE Mark投入约 €40M+终值约 €0;IP 逐步收尾bioregio-stern.de;bionic-vision.org 历史记录Pixium 之前唯一进入商业化的 CE Mark 视网膜假体;同样败在报销设备公司更小、资本更少;Science Corp 在规模上完全不同
Blackrock Neurotech临床研究工具;Tether 投资 $200M 后转向商业化~$200M+(投资后)未披露;学术 / 实验室工具历史让商业估值不透明Neurofounders Big 6 分析;Tether $200M 控股投资(2024 年 4 月)BCI 生态中的平台 / 基础设施玩家;作为纯视觉恢复公司可比性较弱混合商业模式(研究工具 + 临床);估值方法不清晰
EdgeMedical(IPO 基准)商业阶段外科机器人;已有收入N/A(IPO 公司)$2.2B IPO 估值(2026 年 1 月,NASDAQ)PitchBook 2026 年 Q1 Medtech VC and PE Trends;Private Markets Insights 报道为品类领先 medtech 设定公开市场 IPO 基准;Science Corp 需要 FDA 批准和收入,才可能接近这条路径产品类别完全不同;商业阶段且有收入;Science Corp 仍处收入前

除非明确确认,私营公司报道估值均为分析师隐含或二级市场估计。Science Corp 的 $1.5B 是 Series C BusinessWire 新闻稿报道的投后数。 历史可比公司反映终局结果,不代表运营活跃期的峰值估值。除另有说明,所有数字均为美元。

[CV003, CV004, CV013, CV019, CV020, CV021]
FV004: 投资 KPI

从八个维度给出可供 IC 使用的评分,评估 Science Corporation 在 2026 年 3 月 $1.5B 入场估值下的投资理由。评分仅反映截至 2026 年 6 月的公开证据。

[CV003, CV004, CV005, CV006, CV027, CV028]

8.4 情景分析与概率加权估值

截至 2026 年 6 月,Science Corp 的估值区间可由三种情景框定。多头情景假设:2026 年 Q3 获得 CE Mark;2026 年 Q4 在德国商业化上市并获得早期准入报销;以 €80K–€90K 器械 ASP,每季度爬坡到 10–15 例手术;RAPID pathway 使美国 Medicare 在 FDA 批准后 90 天内覆盖(2027–2028 年);晶圆代工 / 生物混合平台可选性贡献终值。在该情景下,以 2028 年预计收入 $120M–$200M 的 7–10x 远期收入倍数计算,隐含企业价值为 $2.5B–$3.5B,相比 $1.5B 进入价可带来扎实的风险调整回报。基准情景假设 H2 2026 获得 CE Mark,但德国和法国在形成有意义报销前需要 12–18 个月 EU HTA 流程,美国 FDA 在 2029 年作出决定,医院采用也因行业过往失败史而谨慎。基准情景首年收入 $30M–$50M,按 4–6x 远期倍数计算,隐含 EV 为 $700M–$1.1B,低于 2026 年 3 月进入估值。空头情景假设 CE Mark 延迟到 2026 年之后,欧盟报销在 2 年内无法建立(沿用行业先例),FDA 进一步延后到 2030 年之后,并需要降价轮或困境战略出售。空头情景隐含价值为 $150M–$400M,Series C 投资人将面临显著资本损失。按 30% 多头、50% 基准、20% 空头加权,期望价值约为 $1.0B–$1.25B,较 2026 年 3 月 $1.5B 进入价低 17–33%。这一负期望价值缺口部分被 DCF 情景未捕捉的可选性价值抵消:生物混合项目、晶圆代工业务和 Science Eye 平台创造了收入型估值未建模的上行情景。对投资期 7–10 年、组合能承受二元 FDA 风险的基金而言,$1.5B 估值可以辩护但并不诱人;若以 $1.0B–$1.2B 进入,吸引力会明显提升。 [CV027, CV028, CV029, CV030, CV031, CV037]

牛市 / 基准 / 熊市情景表
情景关键假设隐含估值逻辑概率信号论点风险
牛市(30% 权重)2026 年 Q3 获 CE Mark;2026 年 Q4 德国发布;12 个月内欧盟报销落地;2027–2028 年 FDA RAPID 路径获批; 2028 年收入 $120M–$200M;代工厂 / 生物混合平台增加终值以 2028 年收入中点 $150M 计,7–10x 远期收入倍数 = $1.05B–$1.5B;叠加 $1.0B–$2.0B 平台期权价值 = $2.5B–$3.5B 区间>30%:需要 CE Mark、没有 HTA 延迟、RAPID 路径按设计运转,且商业爬坡高于赛道先例FDA 延迟、CE Mark RFI 或欧盟 HTA 滑坡会把情景压缩至基准;平台价值需要生物混合 IND 获批
基准(50% 权重)2026 年下半年获 CE Mark;德国 HTA 流程 12–18 个月;美国 FDA 2029 年批准;2028 年收入 $30M–$60M; MEMS 代工厂第三方收入 <$5M以 2028 年收入中点 $45M 计,4–6x 远期收入 = $180M–$270M;加上欧盟先发战略溢价 = $700M–$1.1B 区间50%:最符合历史 BCI 监管和报销时间线报销延迟、Anthem 维持负面政策,或商业使用中 SAE 重新出现,都可能压缩至熊市
熊市(20% 权重)CE Mark 延迟至 2026 年之后,或 RFI 要求补充数据;欧盟报销 3 年内失败(赛道先例);FDA 不可批准函; 下轮融资或困境退出以总投入 1–2x 困境退出($490M–$980M 区间),再扣除优先权负担;低于优先权阈值时,股权净值更可能 <$400M20%:符合赛道基准(此前 3 家视网膜假体公司均商业化失败)以下任意两项:CE Mark 延迟、FDA 拒绝、报销失败、关键高管离职——触发该情景

情景概率为分析师估计,基于截至 2026 年 6 月的监管先例、赛道可比公司和公开证据;不反映 Science Corp 内部预测。 收入估计采用同业设备 ASP 基准($80K–$145K)和保守渗透假设。为与 2026 年 3 月入场价做现值比较, 对 2028 年价值采用 25% 的 DCF 折现率。

[CV027, CV028, CV029, CV037]
FV002: 估值敏感性

五组情景组合下的隐含企业价值,变量包括 CE 标志时点、欧盟报销速度和 FDA 路径假设。2026 年 3 月 $1.5B 入场估值作为参照基线。数值单位为百万美元。

[CV027, CV028, CV029, CV037]
FV003: 估值 / 回报区间

牛市、基准和熊市情景下的企业价值区间,同时给出概率加权预期价值和入场估值参照点。上下限反映各情景的内部假设。所有数字单位为百万美元。

[CV027, CV028, CV029, CV037]

8.5 退出成熟度与战略路径

按标准公开市场标准,截至 2026 年中,Science Corporation 尚未具备 IPO 条件:公司没有收入,未披露单位经济,且所有市场均处于商业化前。2026 年 1 月 EdgeMedical 以 $2.2B 估值上市,让 medtech IPO 窗口重新打开,但至少要求已有商业收入和可见盈利路径。Science Corp 更现实的 IPO 窗口是 2028–2031 年,前提是欧盟商业上市执行到位、获得美国 FDA 批准,并在获批后至少 12–18 个月形成商业牵引,年化收入达到 $50M+。近期更可能的退出路径,是被大型 medtech 或 pharma 平台战略收购。战略契合度最强的潜在收购方包括 Medtronic(神经科产品组合)、Boston Scientific(2026 年 1 月以 $14.5B 收购 Penumbra 以建设神经 / 血管能力)、Johnson & Johnson MedTech(视网膜和器械平台)以及 Abbott(诊断和神经调控)。战略收购很可能由 FDA 批准和至少一年商业收入触发,对应 2029–2031 年交易窗口。缺少二级市场流动性数据、已披露 SPAC 讨论或投行参与,使外部无法评估公司当前退出状态。任何收购方尽调都需要厘清优先清算权层级和可转债处理方式,才能判断各类利益相关方的实际收益分配。 [CV032, CV033, CV034, CV035, CV036]

论点破裂和终止触发器表
触发器阈值 / 事件对论点的传导行动含义
CE Mark 被拒或 RFI 延迟BSI 在 2026 年发出 RFI 或正式拒绝;批准推迟到 2027 年 Q2 之后抹掉先发溢价,并让整个欧盟商业论点延迟 12–24 个月;转向基准或熊市重估持仓;要求查看预提交会议纪要;考虑在延迟后以更低估值入场
欧盟商业收入低于目标德国 CE Mark 批准后 6 个月,每季度手术量 <5 例显示医生采纳门槛、报销摩擦或患者管线问题;基准情形恶化为熊市要求商业细节(站点数、报销确认、待处理病例);暂停追加承诺
FDA 不可批准函或要求新试验FDA 发出完整回应函,要求新的关键试验或重大数据补充美国收入论点延迟 3–5 年;现金消耗率上升;可能需要紧急过桥融资以有利价格退出或减仓;5 年持有期内,美国收入不再支撑退出估值
负面报销决定CMS 对视网膜下假体发布正式不覆盖决定(NCD),或 RAPID 路径被拒美国报销路径变成多年倡议过程;牛市情形坍缩到基准或熊市按纯欧盟业务重构估值,估值相应下调;重估入场纪律
关键高管离职Max Hodak(CEO)或 Alan Mardinly(CSO)在 12 个月内离职,且没有清晰接任者削弱监管可信度、投资人信心和临床网络准入;提高下轮融资概率加快尽调接班计划和董事会构成;列入观察名单

触发器定义为可观察的外部事件,而不是内部预测。阈值采用公开可得的赛道先例(Second Sight、Pixium)和 medtech FDA 审评统计。 并非穷尽;临床安全信号、制造不合规或重大 IP 挑战也可能打破论点。

[CV005, CV022, CV029, CV035]

8.6 建议与最终尽调问题

建议:跟踪(继续研究姿态)。投资论点连贯,并由 NEJM 级别临床证据、可信的先发商业路径和强投资人组合部分去风险。反论点是结构性的:行业商业失败基本率高、收入披露为零、股权结构优先权层级未披露,以及 $1.5B 进入价已计入部分多头情景执行。风险评级为高——二元监管结果主导估值,而前代行业失败率(三家获得 CE 认证的视网膜假体公司中,0 家实现商业可持续)不能被忽视。置信度为中——临床证据强,监管流程推进较深,但商业执行证据尚不存在。以 $1.5B 进入,考虑期望价值缺口,风险 / 回报边际不足。论点破裂触发条件包括:CE Mark 被拒或 RFI 将批准推迟到 2027 年 Q2 之后;欧盟上市 6 个月后收入低于每季度 5 例手术;FDA 发出不可批准函或要求新关键试验;或任意两名创始高管在 12 个月内离职。作出任何投资承诺前,尽调应优先厘清股权结构、优先清算权层级和可转债处理,以评估有效进入经济性。CE Mark 批准结果和德国首个定价 / 报销数据,是近期信息含量最高的催化剂。应独立核验 FDA/CMS 的 RAPID pathway 互动时间线,以及来自 FDA 一手记录的任何已确认 Breakthrough Device designation。若下一轮或二级交易的修正进入估值为 $1.0B–$1.2B,将显著改善期望价值计算。 [CV039, CV040, CV041, CV042]

最终尽调问题表
主题缺失证据重要性负责人 / 尽调路径
股权结构和优先权堆栈各投资人持股、清算优先权、参与权、反稀释条款和可转债转换条款Series C 投资人的有效入场经济性取决于优先权深度;$1.5B 投后估值未必反映 Series C 经济所有权Science Corp CFO / 法律顾问;索取完整股本表和 Series C 条款清单
CE Mark 结果和时间线BSI 审查状态;2025 年 6 月提交后,Science Corp 是否收到任何 RFI 或补充信息请求CE Mark 批准或拒绝是 2026 年信息量最高的单一事件;整个欧盟商业论点都押在上面Science Corp IR;BSI(通过 Science Corp);独立 EU MDR 律师
FDA 一手确认FDA 官方记录,确认 Breakthrough Device 指定;预提交会议纪要;PMA 递交日期Breakthrough Device 指定会加速审评;截至研究日期,PMA 尚未递交;FDA 时间线未知Science Corp 监管事务;FDA CDRH 预提交数据库;直接向 FDA 查询
德国商业发布定价和报销PRIMA 设备 ASP、首批医院合同、AMNOG 早期准入申请,以及德国 NUB(新检查和治疗方法)状态首个市场定价会确立全球 ASP 先例;德国 AMNOG HTA 提供约 12 个月早期准入桥梁,通向完整覆盖Science Corp 商业团队;Spitzenverband Bund der Krankenkassen(GKV-SV)数据库
现金消耗和通向盈利的跑道实际年度现金消耗(估计 $20–$30M)、Series C 后月度跑道,以及 $65M 代工厂 capex 是否计入 $490M 总融资跑道决定美国 FDA 批准(约 2028–2029 年)前是否需要 Series D;稀释风险影响 Series C 回报Science Corp CFO;若 NDA 下可获得,审查经审计财务报表

截至 2026 年 6 月,所有事项均未获公开确认。尽调路径假设可在 NDA 下接触 Science Corp 管理层。 在任何投资承诺前,股权结构尽调优先级最高,因为它决定有效入场经济性。

[CV001, CV010, CV039, CV040, CV041, CV042]

8.7 附录

免责声明

本报告由 AI 辅助研究流程生成,仅使用公开可得来源,不构成投资建议。所有估值、财务数字和前瞻性表述均反映截至 2026 年 6 月可获得的公开信息;由于公司披露私密,内容可能存在重大不完整。读者在作出任何投资决策前,应自行完成独立尽调。

证据索引

结论
编号陈述可信度来源
CO001 Science Corporation was incorporated in 2021 and is headquartered in Alameda, California, USA. SO001, SO002, SO007
CO002 Max Hodak co-founded Science Corporation and serves as CEO; he previously co-founded and was president of Neuralink before departing in May 2021. SO001, SO002, SO006, SO007
CO003 Alan Mardinly, PhD, co-founded Science Corporation and was appointed Chief Scientific Officer in December 2025. SO017, SO022
CO004 Science Corporation describes itself as a neural engineering company developing brain-computer interface (BCI) technologies with PRIMA as its primary near-term commercial product. SO001, SO004, SO028
CO005 Beyond PRIMA, Science Corporation is developing a Science Eye optogenetic platform combining gene therapy and a micro-LED display, targeting retinitis pigmentosa and dry AMD patients. SO006, SO027, SO028
CO006 The PRIMA implant is approximately 2 × 2 mm in footprint and 30 µm thick — about half the diameter of a human hair — and is placed subretinally. SO003, SO004, SO018
CO007 The PRIMA system comprises three components — a wireless subretinal implant, specialized camera glasses with infrared projector, and an AI-powered pocket processor — working together to restore prosthetic central vision. SO003, SO004, SO026
CO008 The PRIMA glasses capture the visual scene and project patterned near-infrared light onto the implant; the implant converts those light patterns into localized electrical stimulation of surviving bipolar retinal neurons. SO003, SO004, SO013
CO009 The PRIMA implant requires no internal battery; it is powered wirelessly by near-infrared light projected from the glasses. SO004, SO013
CO010 PRIMA technology was originally developed at Stanford University in Professor Daniel Palanker's laboratory; Pixium Vision of France subsequently licensed and clinically developed it. SO006, SO027, SO029
CO011 Science Corporation acquired the PRIMA intellectual property and related assets from Pixium Vision SA in April 2024, after Pixium entered liquidation proceedings approved by the Paris Commercial Court. SO006, SO008, SO021
CO012 The April 2024 Pixium acquisition transferred three ongoing clinical trials to Science Corporation, including the PRIMAvera pivotal study (NCT04676854) and two feasibility trials in France and the U.S. SO006, SO008
CO013 PRIMA targets patients with advanced geographic atrophy due to age-related macular degeneration — a condition for which no vision-restoring treatment previously existed. SO004, SO006, SO009
CO014 Science Corporation is expanding the PRIMA clinical trial program to include inherited retinal diseases — Stargardt disease and retinitis pigmentosa — which affect approximately 1 in 10,000 and 1 in 4,000 people respectively. SO001, SO002
CO015 Science Corporation operates a MEMS semiconductor foundry in Research Triangle Park, North Carolina, providing in-house manufacturing capability for its neural interface devices. SO006, SO022
CO016 Science Corporation closed a Series A financing of approximately $47.3 million in July 2021. SO007, SO019
CO017 Science Corporation closed a Series B financing of approximately $113 million in November 2022. SO007, SO019
CO018 Science Corporation closed a $104 million convertible note led by Khosla Ventures in approximately April 2025, bridging ahead of the NEJM publication and CE Mark submission. SO007
CO019 Science Corporation closed a $230 million Series C financing in March 2026, with participation from Khosla Ventures, Lightspeed Venture Partners, Y Combinator, IQT, and Quiet Capital — all pre-existing investors. SO001, SO002, SO007, SO011
CO020 The Series C was oversubscribed, with investor demand exceeding the company's stated capital needs. SO001, SO002
CO021 Total capital raised by Science Corporation is approximately $490 million since its 2021 founding, as disclosed in the Series C announcement. SO001, SO002, SO007
CO022 The Series C implies a post-money valuation of approximately $1.5 billion for Science Corporation, according to analyst and press reporting. SO007, SO022, SO011
CO023 Darius Shahida was appointed Chief Strategy Officer in December 2025; he previously helped raise over $1 billion at Butterfly Network, a medtech company. SO022
CO024 FDA Breakthrough Device designation was originally granted to Pixium Vision for the PRIMA System in April 2023; the designation was inherited by Science Corporation via the April 2024 acquisition. SO002, SO010
CO025 The PRIMAvera pivotal trial (NCT04676854) enrolled 38 patients with geographic atrophy at 17 sites across France, Germany, Italy, the Netherlands, Spain, and the United Kingdom. SO003, SO004, SO016
CO026 The PRIMAvera trial results were published in The New England Journal of Medicine on October 20, 2025, authored by Frank Holz (University Hospital of Bonn) and co-investigators. SO003, SO004, SO009
CO027 Eighty percent of PRIMAvera participants demonstrated meaningful improvement in visual acuity at 12 months post-implantation. SO001, SO003, SO004
CO028 84.4% of PRIMAvera trial participants regained the ability to read letters, numbers, and words using the PRIMA system. SO003, SO004, SO026
CO029 The mean visual acuity improvement in the PRIMAvera trial was approximately 25.5 letters — more than five lines — on the ETDRS chart. SO003, SO004, SO013, SO016
CO030 Nineteen of the PRIMAvera patients experienced adverse events including elevated intraocular pressure, retinal tears, and bleeding; the majority of these events resolved within two months of implantation. SO029, SO013, SO004
CO031 Science Corporation submitted a CE Mark application to the European Union in June 2025; FDA regulatory approval for the U.S. was also filed in June 2025. SO010, SO009
CO032 A U.S. FDA regulatory decision remains pending as of mid-2026; Science holds FDA Breakthrough Device designation to support expedited review. SO001, SO005, SO002
CO033 Science anticipates PRIMA's first European commercial launch in 2026, which would make it the first BCI company with a vision-restoring product on the market. SO001, SO002, SO011
CO034 Science Corporation's headcount grew to approximately 150 employees as of early 2026, according to analyst reporting. SO022
CO035 PRIMA currently delivers black-and-white prosthetic central vision at relatively low resolution, not natural color vision; Science has stated the next device generation will optimize visual performance. SO029, SO013
CO036 The Data Safety Monitoring Board for the PRIMAvera trial concluded that PRIMA's benefits outweigh the surgical risks. SO013, SO016, SO004
CO037 No significant loss of natural peripheral vision was observed in PRIMAvera trial participants after PRIMA implantation. SO004, SO013
CO038 Frank Holz, MD, lead author of the NEJM paper, described PRIMA as a "paradigm shift in treating late-stage AMD" because it actively restores lost vision rather than slowing disease progression. SO009, SO027, SO029
CO039 Pricing and reimbursement terms for PRIMA have not been publicly disclosed by Science Corporation as of mid-2026; the Macular Society noted pricing remains undefined. SO026, SO028
CO040 Science Corporation remains entirely pre-commercial and pre-revenue as of mid-2026; no regulatory approval has been received and no product is yet on the market. SO001, SO002, SO005
CO041 The adverse event rate in the PRIMAvera trial was approximately 59% (19 of 32 evaluable patients), raising questions about broader population safety and the need for long-term follow-up data beyond 12 months. SO029, SO004
CO042 PRIMA's clinical validation to date covers only AMD with geographic atrophy; trials for Stargardt disease and retinitis pigmentosa are in early planning and have not yet reached pivotal stage. SO001, SO014, SO026
CO043 Max Hodak departed Neuralink in May 2021 and founded Science Corporation shortly thereafter; he had been a co-founder and president at Neuralink since 2016. SO002, SO007, SO019
CO044 Alan Mardinly holds a BA from the University of Chicago and a PhD in neurobiology from Harvard University, and had prior leadership experience in neuroscience research at Neuralink. SO017
CO045 Darius Shahida's prior role at Butterfly Network involved raising over $1 billion; this experience is directly relevant to Science Corporation's commercialization and investor engagement objectives. SO022
CM001 Geographic atrophy affects approximately 1 million patients in the United States and 5 to 8 million globally, with approximately 160,000 new US cases diagnosed annually. SM001, SM002, SM012, SM009
CM002 AMD (all stages) was estimated to affect 19.8 million Americans aged 40 and older in 2019, with this number projected to rise as the US population ages. SM002, SM001
CM003 Global AMD prevalence is projected to reach 288 million people by 2050, and GA prevalence quadruples every decade from age 50, reflecting the disease's strong age-dependence. SM012
CM004 Retinitis pigmentosa affects approximately 1 in 4,000 individuals worldwide, with an estimated 80,000 to 110,000 patients in the United States; approximately 12% develop advanced disease with severe vision loss limited to counting fingers or detecting hand motion. SM005, SM012
CM005 Stargardt disease, the most common form of juvenile macular dystrophy, affects approximately 1 in 8,000 to 10,000 individuals globally, typically manifesting in childhood or adolescence and causing progressive central vision loss. SM012
CM006 Geographic atrophy accounts for approximately 20% of all AMD-related legal blindness, identifying it as a disproportionate driver of severe visual disability within the AMD spectrum. SM009
CM007 GA affects approximately 400,000 to 450,000 patients in Germany, 265,000 in France, 170,000 in Italy, and 155,000 in Spain, representing a combined European patient population exceeding 1 million in these four countries alone. SM012
CM008 Advanced RP patients who can only count fingers or detect hand motion — approximately 12% of the 80,000 to 110,000 US RP population — represent the clinical severity tier with the strongest rationale for retinal prosthetic intervention. SM005
CM009 Mordor Intelligence values the global retinal implants market at $56.53 million in 2025 and $61.28 million in 2026, projecting $91.72 million by 2031 at an 8.4% CAGR, using a narrow active-device-units definition. SM006
CM010 Insight Partners reports the global retinal prosthesis market at $392.53 million in 2025 growing to $922.5 million by 2034 at an 11.27% CAGR, using a broader perimeter that includes legacy and pipeline device lines — approximately 7× larger than Mordor's 2025 estimate for ostensibly the same market. SM014
CM011 Grand View Research estimates the global non-invasive BCI serviceable obtainable market at $397.59 million in 2025 with 8.73% CAGR through 2033, and the invasive BCI total addressable market at $168.27 billion in 2025 — a figure that encompasses far beyond visual applications including motor BCIs, ALS communication devices, and neuromodulation. SM007
CM012 The global AMD pharmacological treatment market was valued at $12.64 billion in 2025 and projected at $13.58 billion in 2026 at an 8% CAGR, reflecting the massive scale of anti-VEGF biologics and complement inhibitor drug spending — approximately 200 times larger than the narrow retinal implants market. SM011
CM013 Published analyst estimates for the 2025 global retinal prosthetics or implants market range from $56.5 million to $392.5 million — a 7-fold variance — reflecting incompatible market perimeter definitions rather than genuine uncertainty about the same underlying commercial opportunity. SM006, SM014
CM014 In the United States, CPT code 67027 covers surgical implantation of an intraocular retinal prosthesis, and Medicare has issued local coverage determinations for retinal prosthetics in end-stage retinitis pigmentosa; no national coverage determination currently exists for geographic atrophy as a retinal prosthetics indication. SM010, SM015
CM015 The CMS 2026 Physician Fee Schedule set the non-APM participant conversion factor at $33.40 per RVU with modest increases for office-based care, but imposed significant payment cuts for facility-based procedures at ambulatory surgical centers — a structure that disadvantages retinal implant programs relying on hospital-based surgical infrastructure. SM010
CM016 CE marking in Europe indicates that a device meets safety and performance standards but does NOT guarantee reimbursement; each EU member state independently determines coverage through its own national HTA and reimbursement process, and reimbursement can lag CE Mark by multiple years. SM008
CM017 The EU Health Technology Assessment Regulation (EU 2021/2282) entered phased implementation in January 2025 and mandates joint clinical assessments for Class IIb and Class III medical devices — including retinal prosthetics — beginning in 2026, adding a pan-European evidence requirement before national HTA bodies can issue coverage decisions. SM008
CM018 The Second Sight Argus II was priced at approximately $115,000 to $150,000 per device, with total per-patient treatment costs estimated near $500,000, and received limited and inconsistent insurance coverage that contributed to the device's commercial failure and eventual abandonment of the product line. SM003
CM019 The AAO reported in 2026 that CMS payment cuts for facility-based procedures in ambulatory surgical centers create a reimbursement structure that disadvantages specialized surgical settings, affecting the economics of procedure-intensive implantable devices like retinal prosthetics that require hospital-level infrastructure. SM010
CM020 Hospitals accounted for 54.5% of the global retinal implants market by end-user setting in 2025, with specialty ophthalmic clinics growing at the fastest rate at 11.42% CAGR through 2031, per Mordor Intelligence. SM006
CM021 Retinitis pigmentosa accounted for 46.95% of the global retinal implants market by disease indication in 2025, making it the largest current commercial indication, with Stargardt disease projected to be the fastest-growing indication at 10.21% CAGR through 2031. SM006
CM022 North America held 39.75% of the global retinal implants market share in 2025, followed by Europe and Asia-Pacific; Asia-Pacific is projected to expand at 9.41% CAGR through 2031, reflecting rapid aging demographics in Japan and South Korea. SM006
CM023 Advanced GA in the United States predominantly affects the Medicare-eligible population aged 65 and older, making Medicare the dominant anticipated payer for PRIMA procedures; RP and Stargardt patients are younger and rely primarily on commercial insurance and in some cases Medicaid. SM006, SM004, SM002
CM024 The US anti-VEGF injection standard of care for wet AMD generates over 6 million procedures annually, representing the dominant retinal disease treatment spend; this is a wholly different treatment paradigm from retinal implants, which target end-stage dry AMD patients whose central photoreceptors are permanently lost. SM004
CM025 For patients with advanced geographic atrophy, no pharmacological therapy currently exists that restores central vision once photoreceptors are lost; the clinical comparator for a retinal implant in GA is natural history rather than a competing restorative treatment. SM009, SM001, SM020
CM026 Second Sight Medical Products discontinued support for its Argus II epiretinal prosthesis around 2019 due to financial difficulties, leaving more than 350 patients with effectively obsolete implanted devices; these patients received no upgrade, servicing, or spare parts, creating a patient-abandonment precedent that damaged trust in retinal implant companies. SM003, SM013
CM027 The Argus II's high device price ($115,000–$150,000), limited and inconsistent insurance coverage, total per-patient costs near $500,000, and financial unsustainability of Second Sight represent the primary adverse market precedent that new retinal implant entrants including Science Corporation must explicitly address. SM003
CM028 GA progression is relentless — a majority of patients diagnosed with GA lose substantial vision within 2 to 3 years of initial diagnosis — creating clinical urgency that strengthens patient motivation to seek restorative intervention once a safe and reimbursed option is available. SM009
CM029 Cost-of-illness analyses estimate that inherited retinal diseases drain $13.5 billion to $32 billion annually in North America, providing an economic rationale for payers and policymakers to favor reimbursement of device-based vision restoration that could reduce long-term disability costs. SM006
CM030 Subretinal implantation is a specialized surgical procedure requiring training beyond standard vitreoretinal surgery; the AAO's preferred practice patterns emphasize referral to specialized retinal centers for advanced AMD management, concentrating eligible PRIMA implant procedures at academic medical centers and high-volume specialty practices. SM004
CM031 The FDA's Breakthrough Device Designation pathway has reduced median review timelines for novel retinal and visual neuroprosthetic devices; Neuralink's Blindsight cortical visual prosthesis received BTD in 2024, reflecting continued FDA engagement with visual restoration technologies. SM006
CM032 Tele-ophthalmology programs are expanding geographic reach for retinal implant candidate identification beyond major academic centers, enabling initial screening and clinical triage in communities that lack direct access to specialized retinal surgical centers. SM006
CM033 Science Corporation's PRIMAvera pivotal trial (NCT04676854) achieved its primary endpoint with results published in the New England Journal of Medicine in October 2025; successful commercialization now requires sequential regulatory approval (CE Mark, FDA) and independent payer coverage decisions — each a separate and potentially multi-year hurdle. SM015, SM017
CM034 Science Corporation filed for CE Mark approval for PRIMA in late 2024 or early 2025 per its public communications; the timeline to a potential CE Mark decision is consistent with typical 12 to 18 month review cycles for Class III medical devices, suggesting earliest European market entry in 2026 or 2027. SM025, SM019
CM035 Science Corporation acquired all PRIMA intellectual property and ongoing clinical trials from Pixium Vision SA in April 2024 following Pixium's commercial court liquidation in France, giving Science full ownership of the PRIMA market-authorization asset base and accumulated clinical data. SM018, SM019
CM036 BCI market TAM figures — such as Grand View Research's $168.27 billion invasive BCI TAM — are inappropriate proxies for the retinal prosthetics addressable market because they encompass motor-function BCIs, ALS communication devices, neuromodulation, and other applications with no relevance to vision restoration. SM007, SM006
CM037 BCI market analysts consistently report TAM figures that overstate the addressable market for photovoltaic retinal implants specifically; PRIMA's realistic SAM is orders of magnitude smaller than headline BCI TAM figures suggest, and investor materials using broad BCI TAM as a proxy should be scrutinized for perimeter inflation. SM007, SM006, SM003
CM038 Advanced GA accounts for approximately 20% of all AMD-related legal blindness; the eligible surgical implant pool is a subset of the approximately 1 million US GA patients, further narrowed by residual retinal function requirements, surgical health thresholds, and reimbursement constraints. SM009, SM001
CM039 The FDA's Breakthrough Device Designation pathway for retinal and visual prosthetic devices has shortened expected review timelines and signals the agency's receptiveness to novel visual restoration technologies, reducing some of the FDA regulatory timing risk for PRIMA's US approval application. SM006, SM019
CM040 Globally, approximately 596 million people live with moderate-to-severe vision impairment and 43 million are blind, but the majority have conditions not addressable by current retinal implant technology — including damaged optic nerves, glaucoma, diabetic retinopathy complications, and corneal disease. SM006
CM041 A bottom-up SAM estimate for PRIMA based on approximately 1 million US GA patients at a 2 to 5 percent near-term surgical adoption rate and a $50,000 to $100,000 estimated per-system average selling price suggests a potential US ceiling SAM of $1 billion to $5 billion at full theoretical penetration — but this is a far-future ceiling with no near- term commercial relevance given the nascent reimbursement environment. SM001, SM006
CM042 Science Corporation's PRIMA involves distinct buyer and user roles — retina surgeons in specialized surgical centers serve as procedural buyers while AMD patients (predominantly Medicare-eligible) and younger IRD patients (commercial insurance) are end-users — requiring separate adoption strategies for clinician uptake versus payer coverage. SM006, SM004
CM043 PRIMA does not compete with the dominant anti-VEGF ophthalmology market (over $12.6 billion in 2025); anti-VEGF injections target wet AMD patients who retain functional photoreceptors, while PRIMA targets end-stage dry AMD patients whose photoreceptors are permanently destroyed — a categorically different patient population at a later disease stage. SM004, SM011, SM009
CP001 Science Corporation's competitive landscape divides into two structural tiers: direct visual prosthetics peers competing for the same patient population, and broader BCI competitors competing for investor capital and hospital attention. SP010, SP015
CP002 Pixium Vision SA was liquidated through Paris court proceedings in 2024 and Science Corporation acquired its clinical assets, IP, and ongoing CE Mark application in April 2024, removing Pixium as an independent competitive entity. SP013, SP017, SP018
CP003 Bionic Vision Technologies' second-generation suprachoroidal retinal prosthesis received FDA Breakthrough Device Designation, providing an accelerated regulatory pathway toward commercialization. SP001, SP014
CP004 BVT's second-generation 44-channel suprachoroidal device places the implant between the sclera and choroid, outside the retinal layers, offering a lower surgical-trauma risk than epiretinal or subretinal approaches. SP001, SP014
CP005 BVT's clinical trial (2018–2021, N=4 participants) published in 2024 showed 97% electrode retention at 2.7 years after implantation and functional gains in navigation, mobility, and object detection across all subjects. SP001, SP014
CP006 Second Sight Medical Products ceased production and marketing of the Argus II retinal prosthesis in 2019 and suspended all technical support for existing patients in 2020, leaving approximately 350 implanted patients without device support. SP019, SP023, SP027
CP007 Second Sight's Argus II generated approximately $32 million in total revenue from roughly 350 patients globally against approximately $300 million in cumulative investment and development costs over two decades before commercial withdrawal. SP023, SP019
CP008 Second Sight merged with Nano Precision Medical in 2022 to form Vivani Medical, which pivoted to drug-delivery implants; the neurostimulation business was subsequently spun off as Cortigent, a Vivani subsidiary. SP002, SP016
CP009 Cortigent's Orion visual cortical prosthesis system completed a 6-year early feasibility study in March 2025, with all 6 participants demonstrating safety, reliability, and measurable functional gains in visual assessments; electrode function loss was under 4% across the full study duration. SP003, SP026
CP010 Cortigent's Orion stimulates the visual cortex, not the retina, bypassing the entire damaged visual pathway from photoreceptors through retinal ganglion cells—enabling it to serve patients who have lost both photoreceptor and retinal ganglion cell function, a broader indication than PRIMA's subretinal approach. SP016, SP003
CP011 Retina Implant AG, the German developer of the Alpha AMS subretinal electrode array, resolved to dissolve in March 2019, citing inadequate reimbursement across Europe, high regulatory barriers in the EU, and clinical outcomes that did not meet patient expectations. SP020, SP019
CP012 Retina Implant AG's Alpha AMS achieved CE marking in 2016 with 1,600 electrodes but never achieved visual acuity better than 0.1 in clinical practice; the company dissolved without an acquirer, leaving no active successor in the subretinal electrode array segment. SP019, SP020
CP013 Neuralink raised $650 million in a Series E round in June 2025, with total lifetime funding reported at approximately $1.3 billion as of mid-2026, making it the most capitalized BCI company and Science Corp's primary competitor for investor narrative and BCI category leadership. SP010, SP015
CP014 Neuralink's PRIME study (NCT06429735) enrolled 21 participants across the United States, Canada, and the United Kingdom as of early 2026, targeting individuals with tetraplegia for cursor-control and communication applications. SP009, SP015
CP015 In Neuralink's first human participant, approximately 85% of electrode threads retracted from brain tissue within weeks of implantation, substantially reducing signal decoding capacity; Neuralink responded with algorithm updates but did not resolve the underlying hardware instability. SP022, SP015
CP016 Neuralink targets motor restoration and communication in patients with tetraplegia, not visual restoration; it is a BCI capital and attention competitor to Science Corp rather than a direct clinical competitor for the same patient population. SP008, SP010
CP017 The US Securities and Exchange Commission reopened an investigation into Neuralink and Elon Musk in late 2024 regarding potential misleading safety claims related to animal and human trial results. SP022, SP027
CP018 Synchron closed a $200 million Series D funding round in November 2025, bringing total funding to approximately $345 million; the capital is explicitly targeted at a 2026 FDA pivotal clinical trial for the Stentrode endovascular BCI device. SP004, SP005, SP028
CP019 Synchron's Stentrode device is implanted via the jugular vein, routing the electrode array into the superior sagittal sinus adjacent to the motor cortex, avoiding open-skull surgery entirely and representing a fundamentally different surgical model than craniotomy-based BCIs. SP004, SP021
CP020 Synchron targets ALS, spinal cord injury, and stroke paralysis patients, not visual restoration; its competition with Science Corp is principally for BCI investor capital, clinical-center training capacity, and the 'first commercially approved BCI' regulatory milestone. SP004, SP010
CP021 Synchron has stated its goal of being the first company to file for FDA premarket approval (PMA) of a permanently implanted BCI if its 2026 pivotal trial succeeds, creating a direct narrative competition with Science Corp's CE Mark first-to-market positioning. SP005, SP028
CP022 Precision Neuroscience closed a $120 million Series D in June 2026, bringing total funding to over $290 million; General Catalyst led the round alongside Microsoft M12 and Coatue Management. SP006, SP007
CP023 Precision Neuroscience's Layer 7 Cortical Interface received FDA 510(k) clearance for up to 30-day temporary implantation for neurophysiology and brain mapping, making it the only next-generation cortical BCI device with any commercial FDA clearance as of mid-2026. SP006, SP007
CP024 Precision Neuroscience implanted its Layer 7 device in over 85 patients across more than 15 hospital partners as of mid-2026 and was preparing an IDE submission for a pivotal ALS motor restoration trial. SP007, SP006
CP025 Precision Neuroscience's non-penetrating subdural Layer 7 form factor does not target visual restoration; its primary competitive overlap with Science Corp is for BCI investor capital and clinical-center resource allocation, not for patient population. SP007, SP010
CP026 The BCI sector raised over $960 million in Q1 2026 alone, with capital concentrating in what NeurFounders calls the 'Big 6' invasive BCI companies: Neuralink, Synchron, Precision Neuroscience, Paradromics, Blackrock Neurotech, and Science Corp. SP011, SP010
CP027 Medicare has no national coverage determination (NCD) or dedicated reimbursement coding pathway for permanently implanted BCIs as of March 2026, representing a systemic commercialization risk shared by all BCI companies including Science Corp. SP024, SP025
CP028 Blackrock Neurotech received a $200 million majority investment from Tether in April 2024, positioning it as the incumbent research-infrastructure BCI provider most likely to benefit broadly if BCIs commercialize, rather than a direct Science Corp competitor. SP010, SP015
CP029 Science Corp's PRIMA is the only active retinal BCI with data published in the New England Journal of Medicine demonstrating functional vision restoration—80% of patients showing meaningful visual improvement—at the time of its CE Mark submission in 2024. SP012, SP029
CP030 PRIMA's wireless, battery-free photovoltaic design converts near-infrared light from the patient's glasses directly into electrical stimulation of bipolar cells via 378 photovoltaic pixels, eliminating transcutaneous leads and external batteries and reducing infection risk versus wired devices. SP017, SP018
CP031 Science Corp's acquisition of Pixium Vision assets gave it sole operational control of PRIMA's IP, clinical trial infrastructure, and CE Mark application, removing Pixium's predecessor technology as a competitive threat and consolidating the subretinal photovoltaic segment. SP013, SP017
CP032 The Argus II achieved CMS reimbursement of $150,000 per device under the HDE pathway, with total procedure costs of approximately $500,000 including surgery and rehabilitation, and still failed to secure broad Medicare Administrative Contractor (MAC) coverage approval, which undermined the commercial rollout. SP023, SP019
CP033 PRIMA stimulates bipolar cells in the subretinal space, preserving downstream inner retinal processing including lateral inhibition, whereas the Argus II stimulated retinal ganglion cells epiretinally, bypassing that processing layer—a theoretical resolution and perception-quality advantage for PRIMA. SP017, SP018
CP034 Retina Implant AG achieved CE marking in 2016 and dissolved three years later in 2019, establishing a direct precedent that CE Mark approval alone does not guarantee commercial viability in retinal prosthetics; inadequate reimbursement was the primary cited cause. SP019, SP020
CP035 The abandonment of approximately 350 Argus II patients without technical support, software updates, or a replacement supply chain after Second Sight's 2020 withdrawal created ethical, regulatory, and legal precedents around device lifecycle obligations for implanted neurotechnology companies. SP027, SP023
CP036 BVT's total disclosed funding of approximately $18 million is approximately 27 times smaller than Science Corp's approximately $490 million, suggesting BVT is unlikely to reach commercialization scale independently without additional capital from a major medtech acquirer or strategic partner. SP001, SP012
CP037 Neuralink's approximately $1.3 billion capital base and Elon Musk's media profile generate outsized BCI public and investor mindshare that can divert attention from Science Corp's more near-term and arguably more clinically mature vision-restoration story. SP010, SP015
CP038 The US GAO's 2025 report on brain-computer interfaces identified the lack of dedicated reimbursement coding, privacy frameworks, and post-market surveillance standards as systemic barriers that require coordinated policy action before widespread BCI commercialization is feasible. SP025, SP024
CP039 Implanted BCI devices inherently create strong patient-level switching costs: once a device is surgically implanted and the patient has completed the rehabilitation and training program, the procedural risk of explanation and re-implantation creates de facto patient retention. SP023, SP027
CP040 Vitreoretinal surgical expertise required for PRIMA implantation is a well-established competency at major academic eye centers worldwide, reducing the physician-adoption friction that hampered Argus II and making PRIMA's surgical training curve shorter than for novel craniotomy-based BCI procedures. SP017, SP023
CP041 BVT's suprachoroidal implant placement outside all retinal layers offers a potentially safer implantation plane with reduced risk of retinal damage compared to both epiretinal and subretinal approaches, but also results in greater distance from the target neural cells, potentially reducing stimulation efficiency. SP001, SP014
CP042 Science Corp's PRIMA and BVT's suprachoroidal device both target residual retinal neural circuitry in degenerated retinas but use fundamentally different stimulation mechanisms: PRIMA uses photovoltaic pixel stimulation; BVT uses conventional external-powered electrode array stimulation. SP017, SP001
CP043 Synchron's $345 million total funding and endovascular minimally-invasive positioning are genuine competitive advantages in the broader BCI capital narrative, but Synchron's commercial timeline still extends beyond PRIMA's CE Mark target of 2026 and addresses an entirely different patient population. SP005, SP012
CP044 A peer-reviewed comparative analysis published in PubMed Central (2025) concluded that phased commercialization, predicate-technology leverage, and physician familiarity are the critical success factors Second Sight failed to employ—and that future BCI companies, including Science Corp, must internalize as they transition from clinical to commercial mode. SP023, SP019
CP045 Precision Neuroscience's FDA 510(k)-cleared Layer 7 electrode establishes a regulatory precedent for a commercially cleared thin-film cortical interface, relevant to Science Corp's potential future expansion of its BCI platform beyond retinal indications. SP006, SP007
CI001 Science Corporation closed an oversubscribed $230 million Series C in March 2026 led by Lightspeed Venture Partners and Khosla Ventures, with Y Combinator, IQT, and Quiet Capital also participating; all five were repeat investors. SI001, SI025
CI002 Science Corporation's total capital raised reached approximately $490 million since its 2021 founding, as reported at the Series C close in March 2026. SI001, SI025
CI003 Science Corp raised a $104 million convertible note in April 2025 led by Khosla Ventures to advance its retina and brain implant development programs. SI005, SI006, SI007
CI004 Science Corp's post-money valuation after the March 2026 Series C is reported at approximately $1.5 billion, making it the second most valuable dedicated brain-implant company after Neuralink. SI007, SI017, SI018
CI005 Science Corporation was founded in 2021 by Max Hodak and Alan Mardinly, both former Neuralink executives; Hodak serves as CEO. SI007, SI017
CI006 The March 2026 Series C was described as oversubscribed, indicating investor demand exceeded the $230 million offering size. SI001, SI025
CI007 Science Corp stated the Series C proceeds will accelerate PRIMA commercialization, support expansion of clinical trials to Stargardt disease and retinitis pigmentosa, and scale research, manufacturing, and operational infrastructure. SI001, SI025
CI008 Science Corp's Chief Strategy Officer stated the company's imperative is to become the first BCI company to scale and achieve profitability. SI001
CI009 Science Corp submitted a CE Mark application for PRIMA to the European Union and applied separately to the FDA for US regulatory approval, following Phase III NEJM publication. SI020, SI025, SI026
CI010 Science Corp is investing up to $65 million to expand its MEMS foundry in Durham, NC, adding 57,000 square feet to support PRIMA manufacturing and third-party MEMS services. SI002, SI003, SI008, SI013
CI011 Durham County commissioners unanimously approved a $930,000 ten-year performance-based incentive award conditioned on Science Corp meeting job-creation and investment targets for the foundry expansion. SI002, SI013
CI012 Science Corp's Science Foundry division operates the MEMS facility in Research Triangle Park, Durham, acquired from French company MEMSCAP; it also offers third-party MEMS fabrication services to other medical and BCI device companies. SI002, SI003
CI013 The Durham foundry expansion is expected to create more than 50 new jobs, with approximately two-thirds requiring a two-year degree or less and nearly 90% requiring five years of experience or less. SI002, SI003
CI014 Science Corp has not publicly disclosed any revenue, ARR, gross margin, or financial performance metric; the company is pre-commercial in the United States as of mid-2026. SI007, SI017, SI022
CI015 Science Corp targeted a commercial PRIMA launch in Europe in 2026 following CE Mark submission, which would make it the first BCI company to place a vision-restoration product on a commercial market. SI001, SI025
CI016 No list price for PRIMA has been disclosed by Science Corp; the Argus II historical benchmark was approximately $145,000 device-only; Pixium Vision's pre-insolvency EU pricing was cited at €80,000–€100,000. SI011, SI016
CI017 Anthem Blue Cross Blue Shield's April 2026 coverage policy (SURG.00113) classifies the use of artificial retinal devices as investigational and not medically necessary for all indications. SI009
CI018 CMS and FDA jointly announced the RAPID coverage pathway in April 2026, designed to accelerate Medicare coverage of FDA Breakthrough Device-designated products to within 60–90 days of approval. SI004, SI014
CI019 PRIMA's FDA Breakthrough Device designation status is unconfirmed in public records as of mid-2026; RAPID pathway eligibility requires such designation. SI004, SI022
CI020 At the clinical-stage medtech industry-benchmarked burn rate of $15M–$35M per year, Science Corp's $230 million Series C implies approximately 6–15 years of standalone runway before the next raise is needed, excluding the foundry capex commitment. SI007, SI017
CI021 The BCI sector raised a record $960 million in Q1 2026 across multiple companies, indicating continued capital availability for credentialed medtech firms in the near term. SI023
CI022 Second Sight Medical Products invested more than $300 million in its Argus II retinal prosthetic over 24 years but generated less than $32 million in cumulative revenue before its commercial failure in 2019, leaving hundreds of implant recipients without product support. SI015, SI021
CI023 Pixium Vision, the original PRIMA developer, entered judicial liquidation in early 2024 after running out of cash; Science Corp acquired its intellectual property and clinical assets under court supervision. SI010, SI012
CI024 Retina Implant AG dissolved in 2019 after its Alpha AMS retinal implant failed to reach commercial scale, representing a third predecessor failure in the subretinal prosthetics sector. SI021, SI019
CI025 The ABCMoney investment analysis estimates that Science Corp must achieve $500 million or more in annual revenue within 7–10 years to justify a $1.5 billion valuation at typical medtech exit multiples, implying thousands of PRIMA procedures annually at $50,000+ per procedure. SI007
CI026 Science Corp has not disclosed COGS, gross margin, device ASP, CAC, payback period, or NRR; all unit economics remain private or pre-commercial as of mid-2026. SI007, SI017, SI014
CI027 Industry benchmarks for subretinal devices in pilot manufacturing volumes suggest per-unit COGS of $5,000–$12,000, with material cost decline expected as volume automation and supply-chain maturation are achieved. SI007, SI014, SI016
CI028 The Argus II device-only list price of approximately $145,000 and the associated CMS pass-through code C1849 represent the nearest historical pricing and reimbursement benchmark for retinal prosthetic devices in the United States. SI011, SI016
CI029 Total episode cost for a retinal prosthetic procedure—including device, surgery, and rehabilitation—historically ran $200,000–$300,000 per patient based on Argus II and analogous device experience. SI016, SI011
CI030 With $490 million raised across multiple preferred-share tranches, Science Corp's cap stack reflects substantial investor liquidation preferences; founder equity stakes and the full preference waterfall are not publicly disclosed. SI007, SI017
CI031 The April 2025 convertible note terms—including cap, discount rate, and conversion mechanics—were not publicly disclosed; whether it converted at the Series C or remains as outstanding debt is unknown. SI005, SI006
CI032 Science Corp's vertical integration of MEMS manufacturing through Science Foundry reduces outsourcing risk, accelerates design iteration cycles, and is expected to lower long-run unit costs versus medtech peers that rely on external chip fabrication. SI002, SI003, SI007
CI033 Science Corp acquired the MEMSCAP Durham MEMS foundry in 2022, two years before the April 2024 Pixium acquisition, demonstrating an early strategic commitment to in-house chip fabrication independent of the PRIMA program. SI002, SI003
CI034 MedTech Impact Partners warns that the proposed CMS repeal of the NTAP alternative payment pathway by FY2028 would raise the evidence bar for hospital add-on payments, narrowing the reimbursement window for new implant devices not yet covered before that date. SI014
CI035 PRIMA's Breakthrough Device designation status with FDA is unconfirmed in public records; RAPID pathway eligibility—accelerating Medicare coverage to 60–90 days post-approval—requires that designation. SI004, SI022
CI036 The CMS 2026 OPPS final rule brings modest payment increases for complex retinal surgery, with APC-average reimbursement for complex intraocular procedures approximately $4,222—far below the expected PRIMA device purchase cost. SI016
CI037 The Review of Ophthalmology's 2026 coding update confirms that artificial retinal prosthesis coverage remains payer-specific and that new implant procedures still typically require custom HCPCS codes rather than established CPT coverage. SI011
CI038 BCI sector VC funding is episodic; a clinical setback, regulatory delay, or macro-tightening could compress available runway, given that Q1 2026 record fundraising ($960M) may reflect peak-cycle conditions rather than a sustainable run rate. SI007, SI023, SI024
CI039 Science Corp has not publicly described its GTM approach for the EU PRIMA launch in commercial terms; physician training costs, hospital procurement cycles, and payer pre-authorization are expected to dominate customer acquisition cost for a surgical implant. SI007, SI017, SI014
CE001 The PRIMA implant is a 2 × 2 mm, 30 µm thick crystalline silicon photovoltaic array designed for subretinal implantation. SE001, SE009
CE002 The PRIMA implant contains 378 independently controlled photovoltaic pixels arranged in a hexagonal (honeycomb) grid. SE001, SE009, SE008
CE003 Each PRIMA pixel is 100 µm wide; the pixel array comprises the entire 2 × 2 mm implant surface. SE008, SE009
CE004 PRIMA operates fully wirelessly using 880 nm near-infrared light to deliver both power and image data to the implant; no internal battery or external cable is required. SE001, SE008, SE009
CE005 The PRIMA external system has three components: (1) glasses with front-facing camera and eye-facing NIR projector, (2) a pocket processor, and (3) sun lenses. SE001, SE008
CE006 The PRIMA pocket processor enables real-time zoom adjustment up to 12× and brightness/contrast enhancement. SE001, SE008
CE007 PRIMA bypasses destroyed photoreceptors and directly stimulates the surviving bipolar cells, reintroducing the retinal signal pathway. SE001, SE009, SE015
CE008 The PRIMAvera pivotal trial NCT04676854 enrolled 38 patients at 17 sites across France, Germany, Italy, the Netherlands, Spain, and the United Kingdom. SE016, SE017, SE009
CE009 The primary endpoint of PRIMAvera was a clinically meaningful improvement of ≥ 0.2 logMAR from baseline to 12 months post-implantation. SE009, SE017
CE010 Among 32 PRIMAvera completers, 26 (81%; 95% CI 64–93; p < 0.001) achieved the primary endpoint of ≥ 0.2 logMAR improvement at 12 months. SE009, SE017, SE008
CE011 After multiple imputation for six dropouts, the estimated responder rate was 80% (95% CI 66–94; p < 0.001) across all 38 enrolled participants. SE009, SE017
CE012 The mean visual acuity gain in the PRIMAvera trial was approximately 5 ETDRS lines (25.5 letters); the best individual improvement was 12 lines. SE008, SE017, SE009
CE013 With zoom and contrast enhancement enabled, some PRIMAvera participants achieved visual acuity equivalent to 20/42. SE008, SE009
CE014 Twenty-seven of 32 PRIMAvera completers (84.4%) could read letters, numbers, or words at 12 months post-implantation. SE017, SE021, SE009
CE015 Current PRIMA provides only black-and-white prosthetic vision; no color or grayscale is supported in the commercial device to date. SE008, SE009
CE016 Twenty-six serious adverse events occurred in 19 of 38 PRIMAvera participants (50%); all were attributed to the surgical procedure, not the device alone. SE009, SE017
CE017 The most common serious adverse event in PRIMAvera was ocular hypertension, occurring in 6 events across 23% of participants. SE009, SE017
CE018 No serious adverse event in PRIMAvera was attributed to the PRIMA device alone; all were related to the implantation procedure or the procedure combined with the device. SE009, SE017
CE019 Almost all serious adverse events in PRIMAvera resolved within two months without significant sequelae; the Data Safety Monitoring Board concluded benefits outweigh surgical risks. SE009, SE017, SE015
CE020 Post-implant visual training takes several months before patients reach peak performance, analogous to the auditory learning period required for cochlear implant users. SE008, SE009
CE021 The next-generation PRIMA chip targets pixel pitch of 20 µm with approximately 10,000 pixels per chip; the 20 µm version has been successfully tested in rats at Stanford. SE008, SE009
CE022 A 20 µm pixel chip could deliver prosthetic visual acuity of approximately 20/80 baseline, potentially approaching 20/20 with electronic zoom enhancement. SE008
CE023 PRIMA's wireless photovoltaic design allows future in-situ upgrade by replacing the implant with a higher-resolution chip; multiple chips can also be tiled to cover larger atrophic zones. SE009
CE024 The Science Eye combines optogenetic gene therapy (targeting retinal ganglion cells via AAV vector) with a Lightsheet flexible thin-film microLED display inserted over the retina. SE005, SE006
CE025 The Science Eye Lightsheet microLED display achieves approximately 8× greater pixel density than an iPhone 13 Super Retina display, and moves with the eye to maintain constant pixel-to-cell mapping. SE005
CE026 The Science Eye targets the approximately one million retinal ganglion cells per eye that remain intact after photoreceptor loss from RP or AMD. SE005, SE006
CE027 Science Corporation formally deferred first-in-human Science Eye trials until after PRIMA achieves a commercial product launch in at least one country. SE006, SE023
CE028 The NYX chip family provides low-noise, programmable mixed-signal ASICs for neural recording; NYX 512 is a fully differential 512-channel variant designed for 2024. SE003, SE021
CE029 The LUX chip family provides microLED drivers for optogenetic stimulation; variants include LUX, LUX 2K (active 2,048-pixel), and LUX 16K (passive 16,384-pixel). SE003, SE021
CE030 Science Corporation's GitHub organization (github.com/sciencecorp) hosts 18 public repositories including the Synapse API (Python, TypeScript, C++ clients) with the most recent commit on June 26, 2026. SE007, SE011
CE031 Science Corporation acquired US MEMS foundry assets from MEMSCAP in Research Triangle Park, Durham, North Carolina approximately in 2022. SE024, SE004
CE032 The Science Foundry expansion involves a $65 million capital investment adding approximately 57,000 square feet and more than 50 high-wage manufacturing jobs in Durham County. SE024, SE023
CE033 The Science Foundry is targeted for ISO 9001:2015 quality management system certification and FDA current Good Manufacturing Practice (cGMP) compliance; certification status is not yet publicly confirmed as of mid-2026. SE024, SE004
CE034 The biohybrid neural interface embeds lab-grown stem-cell-derived neurons into a thin-film electrode array containing 520 recording electrodes in an area approximately the size of a pea. SE011, SE010
CE035 In 2024, Science Corporation published a working paper demonstrating safe biohybrid device implantation and optogenetic cortical stimulation in mice. SE010, SE011
CE036 As of April 2026, Yale neurosurgeon Murat Günel (chair of Yale Department of Neurosurgery) has engaged as scientific adviser for Science Corporation's first human biohybrid sensor trial. SE011, SE010
CE037 The initial human biohybrid trial will implant the electrode sensor (without embedded neurons) in craniotomy patients, resting on the cortical surface; Science Corp does not plan to seek FDA approval for this initial sensor trial. SE011
CE038 The biohybrid R&D program is led by CSO Alan Mardinly with a dedicated team of approximately 30 researchers. SE011
CE039 PRIMA requires an intact inner retina (functional bipolar cells); patients whose inner retinal layers have atrophied or who have glaucomatous ganglion cell loss are not candidates for PRIMA. SE001, SE008
CE040 PRIMA's prosthetic field of view is limited to the 2 × 2 mm footprint of the chip; peripheral vision is unaffected but prosthetic coverage cannot extend beyond the chip boundary without tiling. SE001, SE009
CE041 Clinical trial NCT07266584 is evaluating the PRIMA Stim implant in approximately 5 patients with Stargardt disease or retinitis pigmentosa at Sydney Eye Hospital, Australia, with estimated primary completion in March 2029. SE013, SE023
CE042 Earlier PRIMA feasibility trials included NCT03333954 (Europe, original Pixium/Stanford) and NCT03392324 (US), with results published in Ophthalmology (2020) and Nature Communications (2022). SE001, SE016
CE043 The PRIMAvera NEJM paper (DOI 10.1056/NEJMoa2501396) was authored by Frank G. Holz (University of Bonn, lead), Yannick Le Mer, Mahiul M.K. Muqit, and 16 additional co-authors including Daniel Palanker (Stanford) and José-Alain Sahel (University of Pittsburgh) as senior authors. SE009, SE017, SE008
CU001 Science Corporation has no revenue-generating commercial customers as of June 2026; every current relationship is pre-commercial. SU001, SU012
CU002 The PRIMAvera trial (NCT04676854) enrolled 38 patients aged 60 or older across 17 European clinical sites in five countries. SU019, SU011
CU003 84.4% of PRIMAvera trial participants regained the ability to read letters, numbers, and words using the PRIMA system. SU020, SU022
CU004 32 of 38 enrolled patients (84.2%) completed the 12-month primary follow-up in the PRIMAvera trial. SU020, SU019
CU005 The average vision gain in PRIMAvera was approximately 5 lines on the ETDRS letter chart; 26 of 32 evaluable patients (81%) achieved the primary endpoint. SU020, SU022
CU006 Moorfields Eye Hospital in London was the UK's sole PRIMAvera trial site, enrolling five patients under Dr. Mahi Muqit. SU003, SU007, SU024
CU007 Dr. Frank Holz of the University Hospital Bonn was the lead investigator and NEJM lead author for the PRIMAvera trial. SU003, SU017
CU008 UPMC Vision Institute in Pittsburgh was the first US center to implant the PRIMA device in a human patient in a clinical feasibility study. SU011, SU016
CU009 Science Corporation launched a patient registry covering individuals with AMD, retinitis pigmentosa, and Stargardt disease to connect patients to PRIMA and future BCI treatments. SU001, SU002
CU010 Approximately 5 million people worldwide are affected by end-stage atrophic AMD for which no approved vision-restoring therapy currently exists. SU001, SU010
CU011 The Science patient registry tracks functional outcomes including reading ability, face recognition, and mobility, in addition to standard clinical metrics. SU001, SU010
CU012 Registry participants receive priority access to future PRIMA clinical trials and emerging device generations. SU001, SU010
CU013 Science Corp has established 18 clinical partnerships as of its patient registry announcement, spanning European and US sites. SU001, SU002
CU014 PRIMA patient eligibility requires age ≥60, confirmed advanced dry AMD with geographic atrophy, BCVA ≤20/400 in the study eye, and retained useful vision in the fellow eye. SU010, SU018, SU023
CU015 The PRIMA chip contains 378 photovoltaic pixels in a 2×2 mm array, providing limited spatial resolution sufficient for letter and word recognition. SU004, SU003
CU016 Patients require months of intensive visual rehabilitation training to learn to interpret PRIMA's infrared-projected signals before achieving functional vision. SU004, SU007
CU017 The PRIMA implant surgical procedure can be performed by any trained vitreoretinal surgeon in under two hours using standard vitreoretinal toolkit with subretinal delivery. SU003, SU004
CU018 Dr. Mahi Muqit at Moorfields stated the PRIMA chip operation 'can safely be performed by any trained vitreoretinal surgeon in under two hours' as key to broad clinical access. SU003, SU004
CU019 Science Corporation submitted CE mark and FDA applications for PRIMA; the company's patient registry announcement confirmed submission in June 2025. SU001, SU021
CU020 Germany is anticipated as PRIMA's first commercial market after CE mark approval, based on its established pathways for new medical technologies. SU012, SU021
CU021 CE mark approval for PRIMA was widely anticipated by mid-2026 by industry analysts but had not been publicly confirmed as of the reporting date. SU012, SU021
CU022 UK commercial access for PRIMA requires sequential approvals from MHRA, NICE health technology assessment, and NHS England commissioning — a process likely to take one to two years post-CE mark. SU004, SU007
CU023 No US Medicare reimbursement code, national coverage determination, or local coverage policy exists for PRIMA as of mid-2026. SU013, SU009
CU024 New implantable devices in the US typically require FDA clearance, CPT/HCPCS code assignment, and CMS national coverage determination before covered clinical use begins. SU013, SU009
CU025 Science Corporation and Neurosoft Bioelectronics announced a multi-year BCI Ecosystem partnership on February 20, 2026. SU006, SU014
CU026 Neurosoft Bioelectronics is Science Corp's first publicly named foundry ecosystem customer. SU005, SU006
CU027 Building a full-stack clinical-grade BCI platform independently costs $75 million to $100 million; the Science BCI Ecosystem enables first-in-human trials for under $5 million. SU006, SU015
CU028 Neurosoft's ultra-soft neural probes had been used in a dozen human patients for tinnitus and epilepsy prior to the Science partnership. SU006, SU014
CU029 Max Hodak described the Neurosoft deal as 'the first of many' expected Science BCI Ecosystem partnerships, signaling a platform expansion strategy. SU006, SU005
CU030 26 serious adverse events occurred across 19 of 38 PRIMAvera participants; 95% of these resolved within two months of surgery. SU008, SU009
CU031 Adverse events in the PRIMAvera trial included retinal tears, increased intraocular pressure, retinal detachment, and proliferative vitreoretinopathy. SU008, SU009
CU032 Treated eyes in the PRIMAvera trial showed a larger increase in geographic atrophy area compared to untreated control eyes, raising unresolved questions about disease acceleration. SU008
CU033 PRIMA implantation is currently restricted to elite academic vitreoretinal centers with specialized surgical training and post-operative rehabilitation infrastructure. SU004, SU003, SU019
CU034 Patients need months of intensive rehabilitation to learn to use the PRIMA AR glasses and achieve functional reading capability. SU004, SU007
CU035 The PRIMA system addresses only central vision loss in patients with geographic atrophy; this narrow eligibility limits the commercial-ready patient pool substantially below headline AMD prevalence figures. SU010, SU023
CU036 Moorfields trial patient Sheila Irvine reported regaining the ability to read crosswords, prescription labels, and small print on tins, calling the implant 'a big difference.' SU007, SU025
CU037 Dr. Peter Bloomfield of the Macular Society described PRIMA results as 'encouraging' and stated the Society is watching closely for UK approval and potential NHS availability. SU007, SU025
CU038 Science Corp patient registry participation is available at no cost, with participants retaining full data control and subject to enterprise-grade privacy and security measures. SU001, SU010
CU039 Science Corp launched a new clinical trial (NCT07266584, 2026) to expand PRIMA to inherited retinal degenerations including retinitis pigmentosa and Stargardt disease. SU019, SU001
CU040 Science Corp's commercial distribution channel does not exist; no GPO contracts, hospital purchasing agreements, or commercial pricing has been published. SU001, SU012
CU041 No dedicated PRIMA surgical training session was confirmed in the public ASRS 2026 annual meeting schedule as of mid-2026. SU013
CU042 PRIMA candidates must retain useful vision in the fellow (non-implanted) eye, further narrowing eligibility among bilateral AMD patients. SU018, SU023
CU043 The AAO has recognized PRIMA as a photovoltaic implant capable of restoring functional vision in patients with advanced AMD, in an editors' choice review. SU009
CU044 UPMC was described as the first US center to implant the PRIMA device in a clinical trial; Dr. Sahel co-led the NEJM data analysis as a North American principal investigator. SU016, SU011
CU045 The Macular Society reported that a 2022 Moorfields patient was among the first UK recipients of the earlier Pixium PRIMA device, establishing Moorfields' long-term engagement with the platform. SU025, SU007
CR001 Science Corporation holds FDA Breakthrough Device designation for PRIMA (secured 2023) and submitted a CE mark application to BSI in June 2025 with approximately 60GB of technical documentation. SR018, SR024
CR002 FDA PMA review for Class III implantable devices typically takes 180 days from substantive review initiation after filing, with a historical ultimate approval rate near 79%. SR004, SR014
CR003 The RAPID Coverage Pathway (announced April 23, 2026 by FDA and CMS) allows Breakthrough Device-designated products to receive Medicare coverage simultaneously with FDA approval, eliminating the typical 6-18 month post-approval reimbursement gap. SR024, SR025
CR004 Breakthrough Device designation accelerates FDA review scheduling and provides more frequent Agency feedback, but does not lower the safety-and-effectiveness evidentiary threshold required for PMA approval. SR016, SR024
CR005 The PRIMAvera GA acceleration finding — 8.5 mm² vs 2.5 mm² increase in implanted vs fellow eyes over 24 months — may be the primary clinical deficiency that FDA reviews most scrutinously in a PRIMA PMA submission. SR003, SR011
CR006 Science Corp's CE mark application submitted June 2025 is under review by BSI notified body; no public confirmation of a decision or timeline was found as of June 2026. SR018, SR004
CR007 EU MDR Regulation 2017/745 imposes post-market clinical follow-up (PMCF) requirements on Class IIb implantable devices; PRIMA's commercial surveillance program must integrate PMCF into the submission before CE mark approval. SR016, SR018
CR008 PRIMAvera reported 26 serious adverse events in 19 of 38 patients — a 50% incidence rate — with four classified as severe: macular hole, retinal detachment, proliferative vitreoretinopathy, and ocular hypertension. SR003, SR011
CR009 Geographic atrophy area increased 8.5 mm² in PRIMA-implanted eyes versus 2.5 mm² in control fellow eyes over 24 months — a statistically significant difference attributed by Science to surgical trauma from subretinal space entry. SR011, SR008
CR010 Twenty-one of 26 PRIMAvera SAEs occurred within the first two months post-implant, consistent with a surgical-phase event profile rather than a chronic device degradation mode. SR003, SR011
CR011 Subretinal PRIMA implantation requires vitrectomy and entry into the subretinal space within the foveal avascular zone; each procedural step introduces distinct failure modes including choroidal bleed, retinal perforation, and chip displacement. SR003, SR008
CR012 No long-term implant durability data beyond 36 months exists for any commercial subretinal photovoltaic device; PRIMA's second-generation chip design change may reset the durability evidence clock relative to earlier cohort data. SR003, SR014
CR013 Visual acuity and letter-reading gains in PRIMAvera persisted despite accelerated geographic atrophy spread; the GAprogressdissociation between anatomical and functional outcomes supports a functional benefit argument for the PMA. SR011, SR005
CR014 Neuralink's first human trial (May 2024) recorded wire retraction from brain tissue in its initial implant, illustrating that biotic-abiotic interface stability at long implant durations is an unresolved challenge across implantable neural devices. SR027
CR015 The PRIMAvera Data Safety Monitoring Board reviewed the 24-month data and recommended trial continuation without protocol modification, signaling acceptable benefit-risk balance in the DSMB's assessment. SR029, SR005
CR016 CMS has no CPT procedure code, Local Coverage Determination, or National Coverage Determination for implantable subretinal photovoltaic prostheses as of June 2026. SR019, SR025
CR017 Anthem Blue Cross's DME.00052 policy classifies BCI rehabilitation devices — including FDA-reviewed products — as investigational and not medically necessary; private payers typically follow CMS coverage precedent. SR025
CR018 The RAPID Coverage Pathway announced April 23, 2026 allows Breakthrough Device-designated products to receive Medicare coverage simultaneously with FDA approval, directly addressing the historical 6-18 month post-approval reimbursement gap. SR024, SR019
CR019 GAO-25-106952 (2025) found that BCI developers had historically found it difficult to engage with CMS about their devices, and that some clinical trial participants had BCIs physically removed after trials ended because no funding existed to maintain them. SR019, SR025
CR020 No public commercial pricing for PRIMA has been disclosed in any jurisdiction; cochlear implant device cost analogy ($30K-$100K range) and the absence of any US CPT code or payer rate make pricing risk assessment speculative. SR014, SR025
CR021 The RAPID pathway is new (announced April 2026) and has not been tested for any implantable visual prosthetic; the practical timeline for Science Corp to enroll and receive concurrent coverage is unknown. SR024, SR025
CR022 The typical gap between FDA approval of a novel Class III device and CMS National Coverage Determination has historically been 6-18 months for devices without pre-submission coverage engagement; the RAPID pathway explicitly targets elimination of this gap. SR024, SR019
CR023 Science Corporation is investing $65 million in a manufacturing facility in Durham, NC, targeting 150+ new life-sciences jobs; the facility was announced in 2024 and is intended as the primary PRIMA production site. SR028, SR004
CR024 PRIMA chips are photovoltaic CMOS/MEMS devices; Science Corp operates an in-house foundry business (hardware foundry) that manufactures specialty chips for external BCI clients and also fabricates PRIMA arrays. SR028, SR002
CR025 21 CFR Part 820 Quality System Regulation requires implantable device manufacturers to maintain documented design controls, process validation records, corrective-action systems, and production and process controls throughout device manufacture. SR016, SR017
CR026 No public confirmation of ISO 13485:2016 certification for Science Corp's Durham, NC manufacturing facility was found in research conducted through June 2026, despite the CE mark submission referencing extensive quality documentation. SR017, SR018
CR027 A manufacturing non-conformance finding during a pre-approval FDA or BSI inspection could trigger a production hold, require a PMA supplement, or delay CE mark decision by 12-24 months, compressing the entire commercial launch timeline. SR016, SR004
CR028 Science Corp's hardware foundry business serves external BCI clients with custom chip manufacturing; during PRIMA commercial ramp this external business may compete for specialized fabrication capacity and engineering talent. SR028, SR007
CR029 Science Corporation employs approximately 150 people as of March 2026 across four programs: PRIMA, biohybrid neuroscience, hardware foundry, and Vessel wearable BCI hub. SR002, SR028
CR030 No public succession plan, key-person insurance documentation, or CEO contingency arrangement for Max Hodak has been disclosed by Science Corporation or its investors as of June 2026. SR022, SR028
CR031 PRIMA is Science Corporation's sole program with Phase III clinical data and a near-term regulatory path to commercialization; all other programs (biohybrid, foundry, Vessel) are pre-clinical, pre-revenue, or early-stage. SR028, SR002
CR032 Science Corp's biohybrid brain sensor program is targeting first-in-human implantation before 2027 per comments from Yale collaborator Murat Günel; this program requires dedicated clinical trial infrastructure competing with PRIMA for engineering and management attention. SR007, SR009
CR033 Science Corporation has raised $490 million in total funding through the $230M Series C closed in March 2026, at a $1.5 billion post-money valuation with approximately 150 employees. SR028, SR002
CR034 Science Corp's burn rate is estimated at $20-30 million annually, implying approximately 7-8 years of runway at the current consumption rate — sufficient for a 3-year regulatory path but potentially compressed by commercialization capital requirements. SR002, SR021
CR035 Jacques-Alain Sahel serves as Science Corp's Chief Scientific Advisor and is the world's leading retinal prosthetics researcher; his departure would impair the credibility of PRIMA's regulatory submissions and clinical network access. SR003, SR022
CR036 Science Corp's Vessel platform (wearable BCI integration hub) was announced in 2024 and represents a fourth distinct regulatory and commercial pathway requiring its own development, regulatory clearance, and market adoption program. SR028, SR007
CR037 Second Sight Medical Products abandoned its Argus II epiretinal bionic eye product line in 2019, leaving more than 350 patients with unsupported implants; Barbara Campbell's Argus II failed in a New York City subway station, illustrating the patient-safety consequences of device abandonment. SR001, SR010, SR030
CR038 Pixium Vision entered Paris Commercial Court liquidation proceedings in 2024 and was acquired by Science Corporation for approximately €4 million; Retina Implant AG (Germany) discontinued business after 16 years citing an innovation-hostile regulatory and reimbursement climate. SR006, SR012, SR013, SR020, SR026
CR039 Every developer of a subretinal or epiretinal retinal prosthetic that reached clinical proof-of-concept before Science Corporation ultimately exited the commercial market; the common failure modes were inadequate payer reimbursement, insufficient addressable patient volume, and unsustainable capital intensity. SR001, SR006, SR010, SR020
CR040 Science Corp's PRIMA uses subretinal photovoltaic stimulation targeting partially sighted patients (advanced AMD with residual peripheral vision), which differentiates it technically from Second Sight's epiretinal Argus II and Pixium's early Alpha designs. SR003, SR028
CR041 GAO-25-106952 identified data security, cognitive liberty, and neurosurveillance concerns as among the 12 most pressing policy challenges for BCI technology; these ethical/privacy risks apply to all implantable BCI programs including PRIMA and Science's biohybrid sensor. SR019, SR025
CR042 An institutional payer catch-22 operates in the retinal prosthetics sector: limited commercial revenue suppresses the evidence base needed for positive coverage determinations, while coverage gaps prevent the commercialization needed to generate that evidence. SR014, SR019, SR025
CV001 Science Corporation closed a $230 million Series C financing round in March 2026 at a post-money valuation of approximately $1.5 billion. SV012, SV013, SV001
CV002 Science Corporation has raised approximately $490 million in total capital since its 2021 founding, as of the March 2026 Series C close. SV012, SV014
CV003 Neuralink raised $656.5 million in a Series E round in May 2025 at a post-money valuation of $9.6 billion, making it the highest-valued private BCI company by a wide margin. SV003, SV006, SV011
CV004 Synchron raised $200 million in a Series D round in November 2025, bringing total funding to approximately $345 million and implying a post-money valuation of approximately $1 billion per Forbes reporting. SV006, SV019, SV020
CV005 Science Corporation submitted a CE Mark application for PRIMA to BSI in June 2025 with approximately 60 gigabytes of technical documentation; CE Mark approval is expected in mid-2026 but has not been confirmed as of the research date. SV016, SV013, SV029
CV006 Medtech hardware companies in 2026 trade at EV/Revenue multiples of 3.5–5.5x and EV/EBITDA multiples of 11–14x for MDR-ready regulatory-stage assets. SV002, SV004, SV008
CV007 The March 2026 Series C was described as oversubscribed by company communications and third-party reporting, indicating demand exceeding the $230 million target. SV012, SV013
CV008 All five Series C investors — Lightspeed Venture Partners, Khosla Ventures, Y Combinator, IQT, and Quiet Capital — were repeat investors from earlier Science Corp rounds, providing no independent new-investor price discovery. SV012, SV013
CV009 Science Corporation raised a $104 million convertible note in April 2025 led by Khosla Ventures; the note's conversion cap, discount rate, and mechanics were not publicly disclosed. SV025, SV013
CV010 Science Corporation's cap table structure, including individual investor ownership percentages, preference stack depth, participation rights, and liquidation waterfall, has not been publicly disclosed as of June 2026. SV021, SV022
CV011 Science Corporation had approximately 150 employees as of the March 2026 Series C close, according to TechCrunch reporting. SV013, SV022
CV012 Science Corporation is pre-revenue globally as of mid-2026; no revenue, ARR, unit economics, or financial guidance has been publicly disclosed. SV015, SV021, SV022
CV013 Precision Neuroscience raised $102 million in a Series C in December 2024, bringing total funding to approximately $155 million; multiple analyst estimates imply a post-money valuation of approximately $500 million. SV030, SV024
CV014 Blackrock Neurotech received a $200 million majority investment from Tether in April 2024, funding a pivot from academic research tools to clinical products and scale manufacturing. SV024
CV015 The six major invasive BCI companies have collectively raised over $3 billion in cumulative capital by mid-2026, with Neuralink accounting for approximately $1.3 billion of that total. SV024, SV007
CV016 An oversubscribed round closed at $1.5 billion in a quarter when overall medtech VC volume cooled from its 2025 peak is a positive demand signal, but the absence of new-investor pricing discipline limits the informational value of the valuation mark. SV007, SV013
CV017 ABCMoney analysis estimates that Science Corp must reach $500 million or more in annual revenue within 7–10 years to justify the $1.5 billion entry valuation at typical medtech exit multiples. SV021
CV018 Investors Lightspeed and Khosla Ventures need Science Corp to exit above approximately $15 billion for strong fund-level 10x return performance on a $1.5 billion entry. SV021
CV019 Second Sight Medical Products invested over $300 million in the Argus II retinal prosthesis; revenue peaked below $5 million per year and the company abandoned the technology in 2020 with 350+ patients still implanted. SV026, SV028
CV020 Pixium Vision — the predecessor developer of the PRIMA technology now held by Science Corp — was placed into liquidation proceedings in 2024 and acquired by Science Corp for approximately €4 million after spending over €50 million in development capital. SV027
CV021 Retina Implant AG discontinued operations after approximately 16 years of development despite achieving CE Mark for the Alpha IMS device, citing an innovation-hostile regulatory and reimbursement environment in Europe as a primary factor. SV027
CV022 All three CE-marked implantable retinal prosthetics companies that preceded Science Corp — Second Sight (Argus II), Pixium Vision (Prima Alpha and PRIMA), and Retina Implant AG (Alpha IMS) — failed commercially without achieving sustainable revenue. SV026, SV027, SV028
CV023 Cencora acquired Retina Consultants of America for $4.6 billion in 2024 at an implied EBITDA multiple of 15x–23x, representing the highest-value retina-sector acquisition in recent history. SV009, SV010
CV024 Carl Zeiss acquired D.O.R.C., a Dutch retinal surgical device company, for $1.09 billion in 2024, establishing a significant M&A comparable in the retinal device sector. SV010
CV025 The global BCI market was valued at approximately $3.2 billion in 2025 and is projected to reach $12.9 billion by 2034, growing at approximately 17% CAGR. SV003
CV026 Total medtech M&A deal value reached $97.6 billion in 2025, the highest in over a decade, while Q1 2026 PE deal count declined 41% year-on-year as capital became more selective. SV004, SV007
CV027 Probability-weighted analysis assigning 30% to the bull case, 50% to the base case, and 20% to the bear case yields an expected enterprise value of approximately $1.0B–$1.25B, approximately 17–33% below the $1.5B March 2026 entry valuation. SV021, SV005
CV028 The bull scenario for Science Corp assumes CE Mark approval Q3 2026, German launch by Q4 2026, FDA RAPID pathway approval by 2027–2028, and implied enterprise value of $2.5B–$3.5B at 7–10x forward 2028 revenue of $120M–$200M. SV022, SV023
CV029 The bear scenario for Science Corp assumes CE Mark delay past 2026, EU reimbursement failure within 3 years, and an implied value of $150M–$400M consistent with distressed exit or down-round precedent in the retinal prosthetics sector. SV021, SV015
CV030 The investment recommendation for Science Corp at the $1.5B March 2026 entry is TRACK with a research-more posture; the expected-value gap is negative at current scenario probabilities but improvements in CE Mark timing and EU pricing data could shift the call to BUY. SV021, SV005
CV031 The FDA and CMS RAPID Coverage Pathway, announced April 2026, enables simultaneous FDA approval and Medicare coverage determination for Breakthrough Device-designated products, materially reducing the historical 12–18 month post-approval reimbursement gap. SV017
CV032 Anthem's 2026 policy explicitly classifies all artificial retinal devices as investigational and not medically necessary for all indications, representing the current default major US private-payer stance prior to any Science Corp commercial presence. SV017
CV033 Germany provides an established AMNOG early-access pathway for CE-marked medical devices that can bridge reimbursement in the first 12 months post-CE Mark approval, before formal HTA reimbursement review is completed. SV023
CV034 The most strategically aligned potential acquirers for Science Corp include Medtronic (neurology and neuromodulation platform), Boston Scientific (post-Penumbra neuro/vascular scale), Johnson & Johnson MedTech (retina and surgical robotics), and Abbott (neurostimulation and diagnostics). SV004, SV005
CV035 EdgeMedical completed a $2.2 billion IPO in January 2026 for its surgical robotics platform, representing the first major medtech public listing of the year and a potential reopening of the medtech IPO window for category leaders. SV007
CV036 A realistic IPO timeline for Science Corp requires FDA approval, at least 12–18 months of commercial revenue, and a 2028–2031 window based on current regulatory and commercial milestones. SV005, SV021
CV037 At the $1.5B March 2026 entry valuation, the probability-weighted expected return is marginally negative; a $1.0–$1.2B entry valuation would generate a positive expected return under the same scenario distribution. SV021, SV005
CV038 Science Corporation's estimated annual cash burn of $20–$30 million, applied to the $230 million Series C net proceeds, implies approximately 7–11 years of runway not accounting for the $65 million foundry expansion capex or commercial launch costs. SV021, SV013
CV039 Private-market VC conditions in the 2025 cycle, which saw medtech VC funding reach a high of $16.1 billion, contributed to elevated valuation marks for pre-revenue neurotech companies; 2026 represents a moderation phase with capital selectivity increasing. SV007
CV040 Science Corp's earliest plausible first commercial revenue event — initial German implant procedures under AMNOG early-access — is unlikely before Q4 2026 contingent on CE Mark approval timing and hospital onboarding. SV023
CV041 A strategic acquisition of Science Corp by a major medtech acquirer at FDA approval could be valued at $3–$8 billion based on platform optionality, IP portfolio, and manufacturing infrastructure, but this range requires commercial traction and is highly speculative at current stage. SV005
CV042 The undisclosed $104 million convertible note's conversion mechanics may create meaningful preference overhang that reduces the effective economic entry point for Series C investors below the nominal $1.5B post-money valuation. SV010, SV021
来源
编号出版方标题引文
SO001 Science Corporation Science Closes $230 Million Series C to Fund Commercialization of PRIMA BCI Retinal Implant and Expansion of Clinical Trials The deal was oversubscribed with demand in excess of our capital needs, and brings the total capital invested into the company to approximately $490 million since our founding in 2021.
SO002 MassDevice Science Corp. closes $230M Series C to support BCI tech
SO003 Business Wire Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision to Patients with Geographic Atrophy Caused By Age-Related Macular Degeneration, a Leading Cause of Blindness, The New England Journal of Medicine Reports 80% of patients demonstrated meaningful improvement of visual acuity and were able to read letters, numbers, and words.
SO004 Science Corporation The New England Journal of Medicine Reports Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision to Patients with Geographic Atrophy
SO005 STAT News Science Corp. raises $230 million to bring retinal implant to Americans
SO006 Business Wire Science Corporation Acquires Pixium's PRIMA Retinal Implant, Extending and Diversifying Its Commitment to Vision Restoration Through Brain-Computer Interface Technology The PRIMA retinal implant, developed by Pixium and based on research done at Stanford University, shows great promise.
SO007 Crunchbase News Science Corp., Another Braintech Startup Founded By Neuralink Alums, Raises $230M Series C
SO008 MassDevice BCI maker Science acquires retinal implant from Pixium Vision
SO009 MassDevice Science Corp. reports positive results for vision-restoring BCI
SO010 MassDevice Science submits BCI for vision restoration for CE mark
SO011 MedTech Dive Science Corporation raises $230M for Prima vision restoration device
SO012 Fierce Biotech BCI retinal implant medtech Science closes major $230M series C round
SO013 Ophthalmology Times PRIMA retinal implant restores vision in patients with advanced GA
SO014 Longevity Technology Science Corp lands $230m to commercialize vision restoration chip
SO015 Inside BCI Science Corp Raises $230 Million to Commercialise PRIMA Retinal Implant
SO016 Clinical Trials Arena BCI implant's trial success prefigures new era in eye care
SO017 Science Corporation Alan Mardinly, PhD | Science Corporation
SO018 Science Daily Tiny AI-powered eye implant helps the blind see again
SO019 Tech Funding News Can a chip in the eye restore sight? Former Neuralink founder's Science zips $230M
SO020 University of Pittsburgh School of Medicine Retinal Implant Restores Central Vision in Patients with Age-Related Macular Degeneration
SO021 Medical Device Network Science Corporation acquires retinal implant from Pixium Vision
SO022 Neurofounders Science's $230M Series C Tests PRIMA Commercial Scale The financing values the company at roughly $1.5 billion and comes as the team grows to around 150 employees.
SO023 Eyewire News NEJM Publishes Study Demonstrating Vision Restoration with Brain-Computer Interface Retinal Implant
SO024 Yahoo Finance Science Corporation Closes $230 Million Series C to Accelerate Commercialization of Its PRIMA BCI Retinal Implant
SO025 ClinicalTrials.gov Restoration of Central Vision With the PRIMA System in Patients With Atrophic Age-Related Macular Degeneration (PRIMAvera) — NCT04676854
SO026 Macular Society Q&A - your questions answered on the PRIMA System clinical trial for dry AMD
SO027 Science Corporation Science Announces Positive Preliminary Results For Vision Restoration In Pivotal Clinical Trial
SO028 Science Corporation Science | Science Corporation
SO029 Fierce Biotech Science Corp.'s retinal implant restores some vision in patients with macular degeneration The device provides only black-and-white vision, and, for now, the resolution is relatively low.
SO030 Neurofounders PRIMA Retinal Implant Restores Functional Vision in Macular Degeneration
SM001 National Eye Institute (NEI) Eye Health Data and Statistics The NEI tracks eye disease burden and prevalence including AMD and geographic atrophy across the US population.
SM002 Centers for Disease Control and Prevention (CDC) VEHSS Modeled Estimates: Age-Related Macular Degeneration (AMD) Among Medicare beneficiaries aged 65 and older, GA prevalence rate is approximately 0.73%, consistent with the widely used 1 million total US estimate.
SM003 IEEE Spectrum Their Bionic Eyes Are Now Obsolete and Unsupported Ross Doerr and Barbara Campbell were both delighted with the retinal implants that gave them artificial vision. Then the company behind the implants, Second Sight, stopped making them — leaving more than 350 users with devices that became effectively obsolete.
SM004 American Academy of Ophthalmology (AAO) Age-Related Macular Degeneration PPP 2024 — Updated 2026 The AAO Preferred Practice Pattern provides evidence-based guidance for retina specialists on AMD management, including the role of advanced surgical interventions and specialist referral criteria.
SM005 Institute for Clinical and Economic Review (ICER) Report at a Glance: Retinitis Pigmentosa — May 2025 RP affects about one in 4,000 individuals worldwide with an estimated 80,000–110,000 people affected in the United States. About 12% develop advanced RP with severe vision loss such that they can only count fingers or detect hand motion.
SM006 Mordor Intelligence Retinal Implants Market Size, Share, Analysis | Industry Report 2031 The retinal implants market size is expected to grow from USD 56.53 million in 2025 to USD 61.28 million in 2026 and is forecast to reach USD 91.72 million by 2031 at 8.4% CAGR over 2026-2031.
SM007 Grand View Research Brain Computer Interface Market Size | Industry Report, 2033 Considering invasive BCIs, the total addressable market was estimated at USD 168.27 billion in 2025; considering non-invasive BCIs, the serviceable obtainable market was estimated at USD 397.59 million in 2025 at a CAGR of 8.73% from 2026 to 2033.
SM008 European Commission — Health Technology Assessment Implementation of the Regulation on Health Technology Assessment The EU HTA Regulation (EU 2021/2282) entered phased implementation in January 2025; joint clinical assessments for high-risk medical devices (Class IIb and III) are scheduled to begin from 2026 under the phased rollout.
SM009 Retina Today The Burden of Geographic Atrophy Geographic atrophy currently affects approximately 5 million people around the world, including about 1 million people in the United States. A majority of those diagnosed with GA will lose substantial vision within 2 to 3 years.
SM010 American Academy of Ophthalmology (AAO) Medicare Reimbursement 2026: the Good* and the Problematic The CMS 2026 Physician Fee Schedule updated conversion factors with the non-APM participant rate at $33.40, with modest increases for office-based care but significant payment cuts for facility procedures in ambulatory surgical centers.
SM011 Polaris Market Research Age-Related Macular Degeneration (AMD) Market Value 2026–2034 The global AMD market was valued at USD 12.64 billion in 2025 and is expected to grow at a CAGR of 8.0% during the forecast period; 2026 market size projected at USD 13.58 billion.
SM012 Healio Epidemiology and Risk Factors | Geographic Atrophy | Clinical Guidance Geographic atrophy is accountable for approximately 5 million to 8 million cases of AMD worldwide. In the United States, about 1 million people are affected by GA, with 160,000 new cases occurring each year. AMD is predicted to affect 288 million people globally by 2050.
SM013 The Ophthalmologist Betting on Bionic Implants Second Sight had to discontinue both their Argus I and Argus II models in 2019 due to financial difficulties, leaving more than 350 blind users with implanted technology that was essentially obsolete. The challenge isn't in the technology; the challenge will be in the ability to efficiently manufacture the implant.
SM014 The Insight Partners Retinal Prosthesis Market Growth, Share & Trends by 2034 The global Retinal Prosthesis Market size is projected to reach US$922.5 million by 2034 from US$392.53 million in 2025, registering a CAGR of 11.27% during the forecast period.
SM015 ClinicalTrials.gov (US National Library of Medicine) PRIMAvera: Efficacy and Safety Trial of the PRIMA Photovoltaic Subretinal Implant System — NCT04676854 NCT04676854 (PRIMAvera) is Science Corporation's pivotal Phase 3 trial of the PRIMA photovoltaic subretinal implant in geographic atrophy; primary endpoint data published in NEJM in October 2025.
SM016 Macular Society Q&A: Your Questions Answered — Revolutionary Bionic Eye Chip
SM017 BusinessWire Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision to Patients with Geographic Atrophy — The New England Journal of Medicine Reports
SM018 Science Corporation Science Corporation Closes $230 Million Series C
SM019 STAT News Science Corp Raises $230 Million as It Awaits FDA Decisions on Retinal Implant
SM020 FierceBiotech Science Corp.'s Retinal Implant Restores Some Vision in Patients With Macular Degeneration
SM021 Ophthalmology Times PRIMA Retinal Implant Restores Vision in Patients With Advanced GA
SM022 Science Corporation PRIMAvera Trial Preliminary Results
SM023 MedTech Dive Science Corporation Raises $230M for PRIMA Vision Restoration Device
SM024 EyeWire News NEJM Publishes Study Demonstrating Vision Restoration With Brain-Computer Interface Retinal Implant
SM025 MassDevice Science Submits BCI Vision Restoration CE Mark
SM026 ScienceDaily Retinal Implant Restores Central Vision in Patients With Age-Related Macular Degeneration
SP001 Centre for Eye Research Australia (CERA) Bionic eye trial reveals substantial vision improvements 97% of electrodes still functioning up to 2.7 years after implantation
SP002 MedTech Dive Vivani to spin off neurostimulation business
SP003 Vivani Medical (investor relations) Vivani Subsidiary Cortigent Presents Promising 6-Year Early Feasibility Study Results for the Orion System at NANS 2026 All six study participants showed functional gains in visual assessments
SP004 MassDevice Synchron raises $200M Series D to support BCI commercialization, next-gen tech
SP005 MedTech Dive Synchron raises $200M to prepare for brain computer interface launch
SP006 MedTech Dive Precision Neuroscience receives FDA clearance for brain implant
SP007 NeuroTech.com Precision Neuroscience Closes $120M Series D to Accelerate Commercialization
SP008 Neuralink Neuralink Updates
SP009 ClinicalTrials.gov Study NCT06429735 — Precise Robotically IMplanted Brain-Computer Interface (PRIME)
SP010 NeurFounders How Capital Is Consolidating Around the Big 6 Invasive BCI Companies
SP011 Inside BCI BCI Sector Closes Record First Quarter With Over $960 Million Raised
SP012 TechCrunch Science Corp. raises $230M as it races to bring its brain implant to market
SP013 Implantable Device Journal Science acquires assets and clinical studies from Pixium Vision SA
SP014 Nurse.org Blind Patients Gain Vision Thanks to Bionic Eye Breakthrough in Clinical Trials
SP015 Tesorb The 2026 Global Brain-Computer Interface Race
SP016 Neurotech Reports Nader Pouratian, Orion, Cortigent, Second Sight, Argus retinal implant, ReVision, Neuralink
SP017 From the Interface Science Corp and the quest for a Vision BCI
SP018 MD+DI Online Science Corporation to Acquire Pixium's Retinal Implants
SP019 European Society of Cataract and Refractive Surgeons (ESCRS) Learning from retinal implants Despite CE mark approval in Europe, outcomes varied greatly among patients and reimbursement proved difficult to sustain
SP020 BioRegio Stern Retina Implant AG discontinues business activities
SP021 MobiHealthNews Brain-computer interface company Synchron raises $200M
SP022 Digital Watch Observatory Neuralink's first human trial reveals critical issue with brain implant wire retractions approximately 85% of the electrodes had retracted from his brain tissue
SP023 PubMed Central / NIH Brain-computer interface commercialization Second Sight generated approximately $32M in revenue from roughly 350 patients globally against $300M in investments over two decades
SP024 Inside BCI Medicare Has No Plan for Brain-Computer Interfaces
SP025 U.S. Government Accountability Office (GAO) Brain-Computer Interfaces: Applications, Challenges, and Policy Options
SP026 MassDevice Cortigent reports positive 6-year findings for brain implant for providing artificial vision
SP027 Knowing Neurons The Aftermath of Abandoned Neurotech: Ethical and Regulatory Challenges in the Dawn of Brain-Computer Interfaces
SP028 NeurFounders Synchron Raises $200 Million Series D to Advance Brain-Computer Interface Technology
SP029 NeurFounders Science Submits CE Mark Application for PRIMA Retinal Implant
SP030 Science Corporation PRIMA Clinical Trial Data Safety Monitoring Board Votes to Recommend PRIMA Device for Submission for European Approval
SI001 Science Corporation / Business Wire Science Corporation Closes $230 Million Series C to Accelerate Commercialization of Its PRIMA BCI Retinal Implant This financing brings the company's total capital raised to approximately $490 million since its founding in 2021.
SI002 NC Biotech Center Science Corp. set to expand RTP facility
SI003 CBS17 (Nexstar Media) Science Corporation to invest $65 million in Durham County, create jobs in life sciences manufacturing
SI004 U.S. Food and Drug Administration CMS and FDA Announce RAPID Coverage Pathway to Accelerate Patient Access to Life-Changing Medical Devices The RAPID pathway is designed to accelerate Medicare coverage for FDA Breakthrough Device-designated medical technologies.
SI005 Wealth-X Neuralink Rival Science Corp. Raises $104 Million Led By Khosla
SI006 DeepNewz Science Corp., Founded by Max Hodak, Raises $110 Million for Retina and Brain Implant Development Led by Khosla Ventures
SI007 ABCMoney Series C Funding Round Closes $230M for Neuralink Rival Medical device startups burn $20–30 million annually through FDA approval processes. Science raised enough runway for 7–8 years at that burn rate.
SI008 Southpoint Access Science Corporation Plans $65M Expansion in Durham
SI009 Anthem Blue Cross Blue Shield Artificial Retinal Devices — Medical Policy SURG.00113 The use of artificial retinal devices is considered investigational and not medically necessary for all indications.
SI010 MarketScreener Science Corp acquired Intellectual property and related assets for the Prima retinal implant from Pixium Vision SA
SI011 Review of Ophthalmology Coding and Reimbursement: 2026 Update
SI012 NS Medical Devices Science acquires PRIMA retinal implant technology from Pixium Vision
SI013 Science Corporation Durham County Commissioners Unanimously Approve Incentives for Science Foundry Expansion Our long-term goal is to become the go-to manufacturing partner for the next generation of medical and brain-computer interface devices.
SI014 MedTech Impact Partners Breakthrough Device Reimbursement in 2026: Acceleration or Restriction? The Policy Paradox Deepens.
SI015 Fierce Biotech 'Fantastic technology and a lousy company': Second Sight's 'bionic eyes' have gone obsolete while still implanted, report finds Second Sight invested over $300 million in its Argus II retinal prosthetic but generated less than $32 million in revenue.
SI016 Retinal Physician CMS Final 2026 OPPS Rule Brings Modest Increases for Retina Surgery
SI017 Silicon Valley Invest Club Science — Company Profile and Investment Analysis
SI018 Yahoo Finance Science Corporation Closes $230 Million Series C
SI019 Longevity Technology Science Corp lands $230M to commercialize vision restoration chip
SI020 MedTech Dive Science Corporation raises $230M for Prima vision restoration device
SI021 MassDevice Science closes $230M Series C BCI round
SI022 STAT News Science Corp raises $230 million for retinal implant as it pushes for FDA decision
SI023 InsideBCI BCI Sector Closes Record First Quarter with Over $960 Million Raised
SI024 InsideBCI Medicare's Missing Coverage Pathway for Brain-Computer Interfaces
SI025 Science Corporation Science Closes $230 Million Series C to Fund Commercialization of PRIMA BCI Retinal Implant
SI026 ClinicalTrials.gov / NIH PRIMAvera — PRIMA System Study in Geographic Atrophy (NCT04676854)
SE001 Science Corporation PRIMA Visual Prosthesis | Science Corporation PRIMA has 378 light-powered pixels. The implant has a honeycomb pattern of independently controlled pixels that convert infrared light, sent by a pair of glasses, into signals that activate neurons in the retina.
SE002 Science Corporation Biohybrid | Science Corporation Channel count is a measure of the number of neurons or brain regions a brain-computer interface is able to stimulate and/or record at the same time... a million embedded neurons could form a billion synapses with the brain.
SE003 Science Corporation Science Chip Families | Science Corporation NYX: Low-noise and low-power mixed-signal programmable ASIC for neural recording. LUX: Small form-factor, low power microLED drivers designed for optogenetic stimulation.
SE004 Science Corporation Science Foundry | Science Corporation
SE005 Science Corporation Announcing the Science Eye | Science Corporation The Science Eye is a visual prosthesis targeted at retinitis pigmentosa (RP) and dry age-related macular degeneration... by placing an extremely high resolution (~8x denser than an iPhone 13 Super Retina display!) microLED film directly over the retina that moves with the eye, we can maintain a constant pixel-to-cell mapping.
SE006 Science Corporation Prioritizing the Science Eye and PRIMA | Science Corporation It's clear now that the PRIMA implant is by far the biggest advancement in this field in decades, and also has a much faster path to market... We're going to defer proceeding with dosing a first-in-human study for the Science Eye gene therapy until after PRIMA is on the market somewhere in the world.
SE007 Science Corporation Science Corporation — GitHub Organization synapse-python: Python client library and command line utility for interacting with Synapse devices. Updated Jun 26, 2026. 18 repositories in Science Corporation GitHub organization.
SE008 Stanford Medicine Eye prosthesis is the first to restore sight lost to macular degeneration He is also engineering chips that will offer higher resolution vision. Currently, the pixels are 100 microns wide, with 378 pixels on each chip. The new version, already tested in rats, may have pixels as small as 20 microns wide, with 10,000 pixels on each chip. A chip with 20-micron pixels could give a patient 20/80 vision.
SE009 Ophthalmology Times PRIMA System: A major step in the battle against GA The PRIMA implant is 2 x 2 mm and 30 μm thick... a thin crystalline silicon array comprising 378 photovoltaic pixels that are each 100 μm wide. The array is implanted subretinally within the atrophic lesion. A frame-mounted camera on the PRIMA glasses captures images and projects them, after processing, onto the implant with the use of near-infrared light (wavelength 880 nm).
SE010 IEEE Spectrum Science Corp's Biohybrid BCI Adds Neurons to the Brain a million embedded neurons could form a billion synapses with the brain which would provide unparalleled insights into the structure and communication patterns of our brains.
SE011 TechCrunch Max Hodak's Science Corp. is preparing to place its first sensor in a human brain Alan Mardinly, a co-founder and the company's chief science officer, has led development of Science's biohybrid sensor with a team of 30 researchers. The final device will be embedded with lab-grown neurons. Those neurons can be stimulated with pulses of light and are designed to naturally integrate with the neurons in a patient's brain.
SE012 News-Medical Wireless retinal implant restores sight in patients blinded by advanced macular degeneration
SE013 ClinicalTrials.gov PRIMA Products in Stargardt Disease and Inherited Retinal Degeneration (NCT07266584)
SE014 ScienceDaily Stanford's tiny eye chip helps the blind see again
SE015 University College London Pioneering eye device restores reading vision to blind eyes
SE016 ClinicalTrials.gov NCT04676854: PRIMAvera — Safety and Efficacy of the PRIMA Implant
SE017 Science Corporation via BusinessWire Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision — NEJM Reports Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision to Patients with Geographic Atrophy Caused By Age-Related Macular Degeneration.
SE018 Science Corporation PRIMAvera Clinical Trial Preliminary Results | Science Corporation
SE019 Ophthalmology Times PRIMA Retinal Implant Restores Vision in Patients with Advanced GA
SE020 MassDevice Science submits BCI vision-restoration CE Mark application
SE021 MassDevice Science posts positive results for vision-restoring BCI
SE022 ScienceDaily Science Corp.'s PRIMA BCI implant restores vision
SE023 MedTech Dive Science Corporation raises $230M for PRIMA vision restoration device
SE024 NCBiotech Science Corp. set to expand RTP facility
SE025 InsideBCI Science Corp Raises $230 Million Series C for PRIMA Retinal Implant
SU001 Science Corporation Connecting the Blind to Groundbreaking BCI Vision Restoration Technology Through Science Corporation's Patient Registry for Retinal Diseases We submitted our applications for PRIMA's commercial availability in the US and Europe in June and are working with regulators to ensure we get to market as soon as possible.
SU002 MassDevice Science Corp. opens up patient registry for sight-restoring BCI
SU003 NIHR Moorfields Biomedical Research Centre Pioneering PRIMA eye device restores reading vision to blind eyes The PRIMA chip operation can safely be performed by any trained vitreoretinal surgeon in under two hours - that is key for allowing all blind patients to have access to this new medical therapy.
SU004 HTN Health Tech News Interview: Mahi Muqit, consultant at Moorfields Eye Hospital and the Institute of Ophthalmology at UCL, on the European clinical trial of a new bionic eye implant In the UK, we'll be looking for approvals from the MHRA, NICE and NHS England in the future, so we can offer this device to patients with the condition. That process is ongoing, and is likely to take one or two years.
SU005 Science Corporation Science Corp. and Neurosoft Bioelectronics Announce BCI Ecosystem Partnership
SU006 Business Wire Science Corp. and Neurosoft Bioelectronics Announce Novel BCI Ecosystem Partnership It can cost upwards of $75–100 million to build a full-stack clinical-grade BCI platform. The Ecosystem partnership gives innovators access to a state-of-the-art BCI platform that enables first-in-human trials for under $5 million and in a fraction of the time.
SU007 Macular Society (UK) Revolutionary implant allows patients with dry AMD to read again It's made a big difference. Reading takes you into another world, I'm definitely more optimistic now.
SU008 Review of Optometry Retinal Implant Restores Vision in Geographic Atrophy Treated eyes showed a larger increase in geographic atrophy area compared to untreated control eyes, raising unresolved questions about device-related acceleration of the underlying disease.
SU009 American Academy of Ophthalmology A Photovoltaic Implant May Restore Functional Vision in Patients With Advanced AMD
SU010 Science Corporation Science for Patients
SU011 UPMC Physician Resources New Study Shows Central Vision Restoration
SU012 IndexBox PRIMA Vision Implant: $230M Funding & 2026 European Approval Target
SU013 American Society of Retina Specialists New Payment Proposals for the 2026 Medicare Program New implantable ophthalmic devices require separate applications and a national coverage determination; no specific code or coverage policy for novel retinal BCI devices has been assigned for 2026.
SU014 BioSpace Science Corp. and Neurosoft Bioelectronics Announce Novel BCI Ecosystem Partnership
SU015 Engineering.com Science and Neurosoft partner on BCI development
SU016 BioEngineer.org Retinal Implant Restores Central Vision in Advanced AMD Patients, Pitt-Led Study Finds
SU017 MedicalXpress Retinal implant restores central vision in patients with advanced AMD, study shows
SU018 University of Pittsburgh Department of Ophthalmology PRIMA Bionic Vision System Clinical Feasibility Research Study
SU019 ClinicalTrials.gov (US National Library of Medicine) Study Details NCT04676854: Restoration of Central Vision With the PRIMA System
SU020 Business Wire Science Corp.'s PRIMA BCI Implant Restores Functional Central Vision to Patients with Geographic Atrophy — The New England Journal of Medicine Reports
SU021 MassDevice Science submits CE Mark application for vision-restoring BCI implant
SU022 MassDevice Science Corp. reports positive results for vision-restoring BCI
SU023 Science Corporation PRIMA Visual Prosthesis
SU024 UCL (University College London) Pioneering eye device restores reading vision to blind eyes
SU025 Macular Society (UK) Q&A — your questions answered on the revolutionary bionic eye chip
SR001 IEEE Spectrum Bionic Eye Technology Has Gone Obsolete — While Still Implanted in People The company abandoned the technology with 350+ patients still implanted, leaving them with no support and potentially dangerous aging devices.
SR002 TechCrunch Science Corp closes $230M round as it pushes to get its brain implant to patients
SR003 Retinal Physician PRIMA Implant Shows Clinical Potential — PRIMAvera Safety and Outcomes Review Twenty-six serious adverse events occurred in 19 of 38 patients, with four classified as severe — including macular hole, retinal detachment, PVR, and ocular hypertension.
SR004 MedTech Dive Science Corporation raises $230M for PRIMA vision-restoration device
SR005 Clinical Trials Arena BCI implant trial success prefigures new era in eye care
SR006 Bionic Vision Pixium Vision — Company History and Liquidation
SR007 TechCrunch Max Hodak's Science Corp is preparing to place its first sensor in a human brain
SR008 EMPR AAO: PRIMA implant and glasses improve visual acuity in geographic atrophy
SR009 The Next Web Science Corp's biohybrid brain sensor gets first human implant nod
SR010 FierceBiotech Second Sight's bionic eyes have gone obsolete while still implanted in people, report finds Second Sight abandoned its bionic eye product and left more than 350 patients with no support.
SR011 Medscape Wireless Subretinal Implant Shown to Improve Central Vision (PRIMAvera NEJM Results) Geographic atrophy area increased 8.5 mm² in implanted eyes vs 2.5 mm² in fellow eyes over 24 months.
SR012 MD+DI Online A New Vision for Pixium's Retinal Implants
SR013 Medical Device Network Science Corporation Acquires Retinal Implant Technology from Pixium
SR014 From the Interface Science Corp Vision Interface — Strategic Analysis and Risk Assessment
SR015 Financial Content / Business Wire Science Corporation Closes $230 Million Series C to Accelerate Commercialization
SR016 US Government Publishing Office 21 CFR Part 820 — Quality System Regulation for Medical Devices (CFR Title 21, Vol. 8)
SR017 International Organization for Standardization (ISO) ISO 13485:2016 — Medical devices — Quality management systems
SR018 Neurofounders Science Submits CE Mark Application for PRIMA Retinal Implant Science submitted approximately 60GB of technical documentation to BSI as part of its CE mark application under EU MDR.
SR019 US Government Accountability Office GAO-25-106952: Brain-Computer Interfaces — Technology Assessment and Policy Options BCI developers found it difficult to engage with CMS about their devices; some trial participants had BCIs removed when studies ended due to lack of funding.
SR020 BioRegio STERN Management GmbH Retina Implant AG Discontinues Business Activities
SR021 InsideBCI Science Corp raises $230M Series C for PRIMA retinal implant
SR022 TIME Magazine Max Hodak — TIME100 Health 2026
SR023 Google Patents / USPTO US9592396B2 — Biocompatible bonding method and electronics package suitable for implantation
SR024 US Food and Drug Administration CMS and FDA Announce RAPID Coverage Pathway to Accelerate Patient Access to Medical Devices The RAPID Coverage Pathway allows Breakthrough Device-designated products to receive Medicare coverage simultaneously with FDA approval.
SR025 InsideBCI Medicare's Missing Coverage Pathway for Brain-Computer Interfaces Anthem Blue Cross's DME.00052 policy classifies BCI rehabilitation devices as investigational and not medically necessary, despite FDA review.
SR026 MarketScreener Science Corp Acquired PRIMA Retinal Implant from Pixium Vision — Paris Commercial Court
SR027 Dig.Watch / ITU Neuralink's First Human Trial Reveals Critical Issue With Brain Implant Wire Retractions
SR028 Science Corporation Science Corporation Closes $230 Million Series C Financing
SR029 Science Corporation PRIMAvera Data Safety Monitoring Board Review and Recommendation
SR030 Axios Bionic eye recipients left in the dark as technology becomes obsolete Barbara Campbell's Argus II implant failed suddenly in a New York subway station, leaving her temporarily unable to see in a dangerous environment.
SV001 Axios Science Corp. raises $230 million to reverse blindness Science Corp., a brain implant startup led by Neuralink co-founder Max Hodak, raised $230 million in Series C funding at a $1.5 billion valuation.
SV002 healthcare.digital HealthTech and MedTech M&A 2026 Valuation Multipliers MedTech Hardware (MDR-Ready): EV/Revenue 3.5x–5.5x; EV/EBITDA 11x–14x.
SV003 Acquinox Capital Neuralink: Investor Insights into the BCI Opportunity Neuralink raised $656.5 million in Series E funding in May 2025 at a post-money valuation of $9.6 billion, more than doubling its valuation in under two years.
SV004 Bain & Company M&A in Medtech: The Boom in Portfolio Reshaping Total M&A deal value for medtech reached $97.6 billion in 2025 — the highest in over a decade.
SV005 Phoenix Strategy Group MedTech Exit Strategies: M&A vs IPO
SV006 Forbes Full Circle: Synchron's $200 Million Signals Revolution By Evolution The round values Synchron at about a billion dollars.
SV007 Private Markets Insights Medtech Funding Cools From Its Peak — But the Brain-Computer Interface Race Heats Up VC funding cooled from 2025's $16.1B high, reflecting a more disciplined investor stance.
SV008 Equidam EBITDA Multiples by Industry in 2026
SV009 Scope Research Ophthalmology M&A & EBITDA Multiples — Cencora Acquires Retina Consultants of America for $4.6B Cencora's acquisition of Retina Consultants of America implies an EBITDA multiple of 15x–23x, extending above the top of the historical range for office-based physician specialist platforms.
SV010 Becker's ASC Review Ophthalmology dealmaking slows after PE boom: 7 notes
SV011 PremierAlts Neuralink Valuation: $9.7B (2026)
SV012 BusinessWire Science Corporation Closes $230 Million Series C to Accelerate Commercialization of Its PRIMA BCI Retinal Implant Science Corporation Closes $230 Million Series C to Accelerate Commercialization of Its PRIMA BCI Retinal Implant
SV013 TechCrunch Science Corp closes $230M round as it pushes to get its brain implant to patients The financing values the company at roughly $1.5 billion.
SV014 Crunchbase News Science Corp., Another Braintech Startup Founded By Neuralink Alums, Banks A Series C
SV015 STAT News Science Corp. raises $230 million to bring retinal implant to Americans Science Corp. raises $230 million as it awaits FDA decision — the company is pre-commercial and dependent on regulatory outcomes that remain uncertain.
SV016 MassDevice Science closes $230M Series C for BCI retinal implant
SV017 InsideBCI Science Corp Raises $230 Million Series C — PRIMA Retinal Implant
SV018 InsideBCI BCI Sector Closes Record First Quarter With Over $960 Million Raised
SV019 MassDevice Synchron raises $200M Series D to support BCI commercialization, next-gen tech
SV020 MedTech Dive Synchron raises $200M to prepare for brain computer interface launch
SV021 ABCMoney Series C Funding Round Closes $230M for Neuralink Rival At $1.5 billion post-money on $489 million raised, Lightspeed and Khosla need Science to exit above $15 billion for strong fund performance. The math only works if Science reaches $500 million-plus in revenue within 7-10 years.
SV022 Silicon Valley Investclub Science | Silicon Valley Investclub
SV023 Neurofounders Science's $230M Series C Tests PRIMA Commercial Scale
SV024 Neurofounders How Capital Is Consolidating Around the Big 6 Invasive BCI Companies
SV025 WealthX Neuralink Rival Science Corp Raises $104 Million Led by Khosla
SV026 IEEE Spectrum Bionic Eye Technology Has Gone Obsolete — While Still Implanted in People The company abandoned the technology with 350+ patients still implanted, leaving them with no support and potentially dangerous aging devices.
SV027 Bionic Vision Organization Pixium Vision — Bionic Vision Company Database
SV028 FierceBiotech 'Fantastic technology and a lousy company': Second Sight's bionic eyes have gone obsolete while still implanted 'Fantastic technology and a lousy company' — the device worked clinically but the company could not sustain commercial operations.
SV029 MedTech Dive Science Corporation raises $230M for Prima vision restoration device
SV030 MassDevice Precision Neuroscience raises $102M to support BCI
SV031 US Patent and Trademark Office / Google Patents US9592396B2 — Photovoltaic retinal prosthesis with subretinal photosensitive pixel array