AI News AI资讯 2d ago Updated 19h ago 更新于 19小时前 80

China has approved the world’s first invasive brain-computer chip—here’s what’s next 中国批准了世界上第一个侵入式脑机接口芯片——接下来会发生什么

Forget the sci-fi trailers of thought-controlled cars and telepathic internet. The future of brain-computer interfaces just arrived not with a bang, but with a pen stroke in a Henan courtyard. Dong Hui, paralyzed for six years, wrote his name. That mundane act is the earth-shattering news. His device, NEO, developed by Shanghai-based Neuracle Technology, has become the first invasive BCI in the world to gain commercial approval, moving beyond clinical trials to treat real patients. This isn't a 忘掉那些关于意念控车或心灵感应网络的科幻预告吧。脑机接口的未来并非以轰鸣登场,而是在河南某处庭院里随着一笔一划悄然降临。瘫痪六年的董辉写下了自己的名字。这看似平凡的举动,却是震动世界的突破。他所使用的NEO设备由上海博睿康科技研发,成为全球首个获得商业化批准的侵入式脑机接口,正式从临床试验阶段迈向真实患者治疗。这绝非实验室里的猎奇品——它已是市场中的成熟产品,标志着神经科学领域的一次结构性变革。

85
Hot 热度
80
Quality 质量
85
Impact 影响力

Analysis 深度分析

Forget the hype from Elon Musk’s Neuralink. The first invasive brain-computer interface to get out of the lab and into the market isn’t a sleek, American invention promising to merge us with AI. It’s a coin-sized implant from a Shanghai startup, approved by Chinese regulators, already letting a paralyzed man write his name. This isn’t just a medical milestone; it’s a seismic shift in where and how frontier neurotechnology gets built and deployed.

The story of Dong Hui is powerful in its simplicity. A car accident, six years of paralysis, and a brain implant that, after months of training, allowed him to grip a ball and, later, a pen. It’s a modern miracle of neural engineering. But the real story isn’t the individual triumph, however moving. It’s the regulatory and commercial leapfrog. While Neuralink has dominated headlines with grandiose visions and a PR machine fueled by celebrity, Neuracle Technology has been quietly executing. They’ve achieved what Musk’s company has not: actual, approved clinical product for humans.

This is a brutal lesson in the reality of medtech versus the fantasy of Silicon Valley disruption. Building a brain-computer interface is one thing. Getting a government to stamp “approved for patient use” is an entirely different, grinding battle of safety data, manufacturing consistency, and regulatory navigation. China’s National Medical Products Administration has now drawn first blood. They’ve set the benchmark for the world, signaling that BCIs are transitioning from academic curiosities to regulated medical devices. The precedent is huge.

Critically, the NEO device isn’t trying to be a universal mind-machine conduit. It’s targeted. It’s for limb paralysis due to spinal injuries, using electrodes placed on the dura mater—not penetrating the brain tissue itself, a less invasive approach than some competitors. This is smart pragmatism. It prioritizes a defined, severe medical need and a plausible safety profile over sci-fi promises. The 2.5-hour daily training regimen with a robotic glove underscores a truth often glossed over: this is rehabilitation, not magic. The brain must learn, and the system must be calibrated, through repetitive, laborious practice. Dong’s excitement over a single successful grab speaks to the immense, granular challenge of this field.

And yet, the implications stretch far beyond China’s borders. We’re witnessing the potential bifurcation of a critical technology’s future. Will the path to BCI integration be defined by a Chinese model of state-backed, rapidly deployed medical devices for defined conditions? Or by a Western model of private, venture-fueled companies chasing transformative—and riskier—applications? Neuracle’s approval suggests the former is currently faster to market. This raises urgent questions. Will there be divergent standards for safety and data privacy? Could a “BCI-as-a-service” model emerge, tied to specific healthcare systems?

The race is no longer just technical; it’s geopolitical. For years, the narrative around cutting-edge AI and hardware has been one of US-China rivalry. Now, China is demonstrating it can also lead in the profoundly personal realm of human-machine integration. This isn’t about a single startup’s success. It’s about a ecosystem that can take a complex idea from a university lab (Tsinghua, in this case), shepherd it through rigorous trials, and get it to paying patients. That pipeline is formidable.

We should be cautiously optimistic. For people like Dong, this technology is a window back to autonomy. But we must resist the overblown rhetoric that often accompanies neurotech. This isn’t the dawn of cyborgs next year. It’s the painstaking beginning of a new chapter in restorative medicine, where progress will be measured in regained hand movements, not uploaded consciousness. The real revolution isn’t in the implant itself, but in the fact that it’s the first of its kind, anywhere, to be deemed safe and effective enough to leave the research setting. That’s how real progress happens—not with a tweet, but with a regulatory approval. The world’s neurotech race just got a new leader, and it wasn’t the one most people were watching.

董慧坐在河南老家的院子里,重新握起笔写下自己名字的那一刻,他大脑皮层上那枚硬币大小的装置,已经静默地运行了十一个月。这不是科幻电影,而是2024年秋天一个真实的、略带笨拙却震撼人心的瞬间——一个瘫痪六年的人,因为一次开颅手术植入的设备,获得了重新与世界物理互动的可能。他太激动了,以至于签名时都漏了一笔。这个细节比任何完美的技术演示都更有力量,它赤裸地展示了科技在触达个体时,那种原始的、带着颤抖的喜悦。

董慧使用的设备叫NEO,由上海的脑虎科技和清华大学联合研发。它并非侵入式地插入大脑组织,而是被小心地放置在保护大脑的硬脑膜上,通过另一部分贴在颅骨上的设备传输信号。这套系统的精妙之处在于它的“非终极侵入”:既绕过了直接刺入脑组织的高风险与伦理雷区,又获得了比头皮电极清晰得多的脑电信号。这就像在窃听一场大脑内部的秘密会议时,选择了站在会议室门口的玻璃窗外,而不是直接钻到桌子底下。效果够用,且风险可控。从临床操作到日常训练,这套方案显示出一种罕见的工程学上的成熟与务实。

当这枚“中国造”的芯片在2025年3月拿到全球首张侵入式脑机接口的医疗注册证时,它实际上完成了一次对硅谷叙事的小型颠覆。埃隆·马斯克的Neuralink仍在为它的第一例人体试验和FDA的重重审批摇旗呐喊,而中国的团队已经让一个产品悄然走出了实验室,开始服务于真实的患者。这并非是技术代差的胜利,而是两种路径哲学的分野。一边是追求最前沿的极限,试图一步到位实现“全脑接口”或“心灵感应”;另一边则像董慧的训练一样,每天2.5小时,一遍遍地练习抓握水球,目标是让患者最终能自己穿上衣服、端起饭碗。后者不性感,不具有话题爆炸性,但它直接指向了病床上最迫切、最卑微的需求。中国的科研生态正在证明,解决一个具体问题的能力,有时比描绘一个宏伟蓝图更值得尊敬。

获批适应症明确指向了因脊髓损伤导致四肢瘫痪、但上肢仍有部分功能的人群。这是一个精妙的切入点。它避开了最复杂、最不可预测的神经退行性疾病,选择了一个相对“清晰”的战场——大脑发出的指令通常是完好的,只是传输通道被物理性地阻断了。NEO做的,就是搭建一个绕过断点的“外挂神经通道”。这种临床路径的务实,正是其能率先获批的关键。它不试图解释意识,不尝试解读情感,它只是忠实地将“我想握紧”的念头,转化为一只机械手套的夹紧动作。在伦理的钢丝上,它选择了最平稳的那段路先走。

当然,欢呼之余必须保持清醒。从“能抓球”到“能独立生活”,之间横亘着无数工程学与临床训练的巨壑。那每天2.5小时的康复训练,是患者用巨大的意志力在弥补系统尚不完美的现实。信号的稳定性、设备的长期生物相容性、对非目标神经信号的抗干扰能力,这些都是未来必须硬碰硬解决的硬骨头。而更深层的问题已经隐隐浮现:当我们的大脑可以直接与外部设备对话,思想的边界在哪里?谁有权访问这最私密的“内部频道”?如果设备可以读取运动意图,那么读取更复杂的心智活动,在技术上只是时间问题。今天的康复工具,明天会否成为无法预料的监控或影响工具?

董慧的故事有一个温暖的注脚:他装上设备后最迫切的愿望,是不再给年迈的父母添麻烦。技术最大的慈悲,或许正在于此——它最终指向的不是人类能力的无限扩张,而是将一部分人从绝对的无助中打捞出来,归还他们最基本的尊严与自主性。NEO的里程碑意义,不在于它有多么颠覆,而在于它如此扎实地落地。它像一颗精确制导的子弹,射向了“帮助具体的人”这个靶心。当科技圈还在为通往通用人工智能(AGI)的路径争执不休时,中国的一个团队和一位患者,用一笔一划的重新书写,提供了另一种同样重要的、关于“技术应该向何处去”的答案。这答案不在云端,而在院子里,在那张被泪水和墨水晕开的纸上。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

医疗AI 医疗AI 芯片 芯片 政策 政策
Share: 分享到: