BCI-Sonics Secures Hundreds of Millions in Angel Round Funding to Accelerate Development of Next-Generation AI Ultrasound Brain-Computer Interface Platform | 36Kr Exclusive 「华超神控」获亿元天使系列融资，加速打造新一代AI超声脑机接口平台 | 36氪首发

## Summary
A 100-million-yuan angel round sounds like plating the delicate instruments of a lab with a layer of gold. When the capital winter sets in, and hard tech becomes one of the few sectors capable of telling a sexy story, the brain-computer interface (BCI) field—which already sounds sufficiently "sci-fi"—naturally becomes a hot favorite. BCI-Sonics and its backers are clearly betting on the "ultrasound" path, perceived as more elegant and human-centered: a way to communicate with deep brain regions without opening the skull.

## Deep Analysis
A 100-million-yuan angel round sounds like plating the delicate instruments of a lab with a layer of gold. When the capital winter sets in, and hard tech becomes one of the few sectors capable of telling a sexy story, the brain-computer interface (BCI) field—which already sounds sufficiently "sci-fi"—naturally becomes a hot favorite. BCI-Sonics and its backers are clearly betting on the "ultrasound" path, perceived as more elegant and human-centered: a way to communicate with deep brain regions without opening the skull.

The core of this story is indeed captivating. Imagine a Parkinson's patient no longer needing to endure the fear and risks of invasive electrode implantation. Instead, they could lie on a clinic bed, receive a painless, invisible ultrasound "massage," stabilize trembling hands, and regain dignity in life. This is precisely the ideal picture of "precision neuromodulation" that BCI-Sonics paints. They have chosen low-intensity focused ultrasound (tFUS/LIFU), recognizing its potential for being non-invasive and deep-reaching, attempting to find a sweet spot between safety and efficacy—a point that both traditional invasive BCIs and less effective non-invasive technologies have failed to reach.

The technological ambition primarily focuses on two "self-developed barriers." The first is a transcranial phase correction algorithm, which claims to achieve a precision of 1.5 millimeters. This sounds impressive, but the devil is in the details. The human skull is like an "acoustic labyrinth" of varying shapes and thicknesses, causing ultrasound distortion that differs from person to person. How many samples has this algorithm been validated on? In real, noisy clinical environments, how long and how stably can this millimeter-level precision be maintained? This is the most brutal test when transitioning from the "potential" in papers to the "performance" in products. Any showboating in a lab environment must face a big question mark when confronting the human brain—the most complex black-box system.

The other selling point is an "AI closed-loop neuromodulation platform," encompassing "reading," "modulation," and "feedback." The concept is closed-loop, very AI-driven, and aligns with the current investment narrative where everything can be looped. However, decoding neural signals (the "reading") is itself an unconquered fortress. How much AI can truly help here—is it an enabler or just a gimmick? More critically, dynamic modulation at the millisecond scale requires extraordinary precision in sensing, computing, and execution. Each additional link in this closed-loop chain exponentially increases potential errors and uncertainties. For now, this seems more like a blueprint for a future technology architecture rather than an immediately deployable clinical solution.

Of course, we cannot entirely dismiss this due to the gap between technological ideals and reality. BCI-Sonics has a star-studded team: PhDs from the Chinese Academy of Sciences and Fraunhofer, a former China R&D head of GE Healthcare, and multidisciplinary backgrounds from Tsinghua, Shanghai Jiao Tong, and Imperial College. The value of these "veterans" lies not only in their technical understanding but also in their grasp of the entire chain from principle to registration in medical devices, knowing how to walk the tightrope between regulation, engineering, and clinical practice. This is far more reliable than a group of scientists with only impressive papers starting a business. The "consensus first, then entrepreneurship" model valued by investors also suggests that the team is not a hastily assembled outfit chasing trends.

However, the path to success is fraught with specters. First is the regulatory maze. What are the safety and efficacy review standards for non-invasive neuromodulation devices? How do we define its "therapeutic effect"? Is it a replacement for DBS, or a complementary or early intervention approach? This path lacks mature approval precedents globally. BCI-Sonics and its domestic peers may well become co-explorers of the rules, which is even more time-consuming and labor-intensive than developing the technology itself.

Second is the real clinical demand and willingness to pay. The article uses Parkinson's as an example, but DBS therapy is relatively mature with established insurance payment pathways. How can a completely new, non-invasive technology with yet-to-be-validated efficacy convince doctors, patients, and the healthcare system to pay for it? Initially, breakthroughs may only be found in areas where existing treatments are ineffective, such as treatment-resistant depression, intractable pain, and substance addiction. Market education costs will be extremely high.

Finally, and most fundamentally: the foundational understanding of brain science remains a "desert." Our understanding of how the brain works is still like seeing only a part of the picture. Does precisely targeting ultrasound at a certain nucleus guarantee the expected, stable, and positive effects? What are the long-term effects and potential unknown impacts? Until a paradigm shift occurs in neuroscience, all modulation methods are like cautiously probing from outside a "black box."

Investors often talk about "defining the next decade" and "benefiting patients and the public." The slogans are always appealing. But history tells us that the gap from a lab concept to a reliable medical product that can change patients' lives is not just a few years, but a path paved with countless failures, compromises, and restarts—a path of blood and tears. BCI-Sonics holds a good hand of cards, but the game has just begun. The opponents are the most complex human organ, as well as the even more complex realities of human trials, regulatory approvals, and commercial promotion. It might become the spotlight that illuminates the deep brain, or it could just be another brilliant yet fleeting tech firework at a capital feast.

一亿天使轮，听起来像是给实验室里的精密仪器镶了层金边。当资本寒冬里，硬科技成为少数几个能讲出性感故事的赛道时，脑机接口这个听上去就足够“科幻”的领域，自然成了香饽饽。华超神控（BCI-Sonics）和它背后的资方，显然押注的是“超声”这条被认为更优雅、更人文的路径——不开颅，就能与深脑对话。

这故事内核确实迷人。想象一下，帕金森患者不再需要经历开颅植入电极的恐惧与风险，而是躺在门诊的床上，接受一阵无形无痛的超声波“按摩”，就能稳住颤抖的手，重获生活的尊严。华超神控描绘的，正是这样一幅“精准神经调控”的理想图景。他们选择低强度聚焦超声（tFUS/LIFU），看中的正是其“无创、深部可达”的潜力，试图在安全性与有效性之间，找到一个传统侵入式BCI和效果有限的非侵入式技术都未能企及的甜蜜点。

技术上的野心，主要集中在两个“自研壁垒”上。一是经颅相位校正算法，号称能把精度收敛到1.5毫米。这听起来相当厉害，但魔鬼在细节里。人的颅骨如同形状各异、厚度不均的“声学迷宫”，超声波穿过时产生的畸变千人千面。这个算法在多少样本上验证过？在真实的、嘈杂的临床环境中，这个毫米级的精度能保持多久、多稳定？这是从论文里的“potential”（潜力）变为产品里的“performance”（性能）时，最残酷的试金石。任何在实验室环境下的炫技，面对人脑这个最复杂的黑箱系统，都得打上一个大大的问号。

另一个卖点是“AI闭环神经调控平台”，从“读”到“调”再“反馈”。概念很闭环，很AI，很符合当下万物皆可闭环的投资叙事。但神经信号的解码（“读”）本身就是尚未攻克的堡垒，AI在此究竟能发挥多大作用，是助力还是噱头？而更关键的是，在毫秒级尺度上动态调控，这需要何等精度的传感、计算与执行能力？这闭环的链条每多一环，潜在的误差和不确定性也呈指数级增加。目前看来，这更像是一个面向未来的技术架构蓝图，而非一个立即能投入临床的解决方案。

当然，不能因为技术理想与现实的差距就全盘否定。华超神控的团队配置堪称豪华：中科院与弗劳恩霍夫的博士，GE医疗的前中国区研发负责人，清华、交大、帝国理工的多学科背景。这群“老炮儿”的价值，不仅在于技术理解，更在于他们清楚医疗器械从原理到注册转化的完整链条，知道如何在法规、工程、临床之间走钢丝。这比一堆只有漂亮论文的科学家创业，靠谱得多。资方看重的“先有共识，再创业”的模式，也至少说明团队不是为了追风口临时拼凑的草台班子。

然而，通往成功的路上，幽灵重重。首先是监管的迷宫。无创神经调控设备，其安全性和有效性的审评标准是什么？如何定义其“疗效”？是替代DBS，还是作为一种补充或早期干预手段？这条路径在全球都缺乏成熟的审批先例，华超神控和国内同行们，很可能要成为规则的共同探索者，这比研发技术本身更耗时耗力。

其次是临床的真实需求与支付意愿。文中用帕金森举例，但DBS疗法已相对成熟，且有明确的保险支付路径。一个全新的、无创的、但疗效可能需要验证的技术，如何说服医生、患者和医保体系为其买单？初期很可能只能在难治性抑郁症、顽固性疼痛、药物成瘾等现有手段效果不佳的领域寻找突破口，市场教育成本极高。

最后，也是最根本的：脑科学的底层认知仍是“荒漠”。我们对大脑工作机制的理解还停留在管中窥豹的阶段。超声波精准打到某个核团，一定能产生预期中的、稳定积极的效果吗？它带来的长期效应和潜在的未知影响是什么？在神经科学取得范式突破之前，所有调控手段都更像是在“黑箱”外小心翼翼地试探。

投资人们都在说“定义下一个十年”、“造福患者及广大人群”，口号总是动听。但历史告诉我们，从实验室概念到一款能改变患者生活的可靠医疗产品，中间隔着的不是几年时间，而是一条由无数失败、妥协、重启组成的血泪之路。华超神控手握一副好牌，但牌局才刚刚开始，对手是人类最复杂的器官，以及更加复杂的人体试验、监管审批和商业推广的现实世界。它或许能成为那盏照亮深脑的探照灯，但也可能，只是资本盛宴中，又一个灿烂而短暂的科技烟花。