Shanghai-based BCI-Sonics completes RMB 100 million angel series financing for non-invasive brain-computer interface.

ME AI News: In January this year, Merge Labs, the brain-computer interface company co-founded by OpenAI CEO Sam Altman, emerged from stealth with a $252 million seed round, valuing the company at $850 million. The firm avoids electrodes and craniotomy, aiming to read and modulate the brain through the skull using ultrasound and molecular techniques. Less than six months later, the same vision has taken root in China: Shanghai-based BCI-Sonics has just completed a RMB 100 million angel series round led by Matrix Partners China, with the angel+ round co-led by Delian Capital and Daoyuan Capital, and Xunguang Capital serving as exclusive financial advisor. This funding will support technical validation and industrialization. BCI-Sonics is pursuing the same direction as Merge Labs: using transcranial focused ultrasound (tFUS-LIFU) to non-invasively target deep brain regions and applying AI to decode signals and create closed-loop feedback—in other words, reaching deep into the brain with sound waves without opening the skull. Why has ultrasound suddenly gained momentum? A look at the two dominant approaches in today’s brain-computer interface landscape makes it clear. Neuralink, co-founded by Elon Musk and long established as the most well-known name in BCI, follows an invasive craniotomy route: a surgical robot removes a small portion of the skull and implants thousands of flexible electrode threads into the subject’s motor cortex. In June last year, Neuralink closed a $650 million Series E round at a $9 billion valuation and announced plans to begin mass production by 2026. Another company, Synchron, delivers stent-based electrodes through blood vessels into the brain—bypassing craniotomy but still remaining invasive. While its signal quality is superior to non-invasive methods, it carries surgical risks and faces scalability challenges. Ultrasound offers a different path: working through an intact skull, it trades surgical precision for non-invasiveness. In Altman’s view, brain-computer interfaces should do only one thing: “read.” When you think something, ChatGPT should be able to catch it instantly… It sounds like science fiction, but the underlying science isn’t new. Low-intensity focused ultrasound (LIFU) has long been used clinically to non-invasively reach deep brain structures that TMS cannot access and that previously required invasive DBS surgery. In an early trial targeting addiction, researchers stimulated the nucleus accumbens with LIFU; participants reported reduced cravings after a single session, with effects lasting 90 days. Clinical trials for depression and anxiety are also underway. BCI-Sonics is tackling the most challenging engineering hurdle on this ultrasound path. Injecting ultrasound into the brain isn’t difficult—the real challenge is ensuring the sound remains precisely focused on the intended target after passing through the skull. Every individual’s skull varies in thickness and curvature; ultrasound waves undergo phase distortion and focal drift during transmission, and energy attenuates. If not properly corrected, the theoretical target point diverges from the actual focus point, compromising efficacy and reproducibility. This step is known in the industry as transcranial phase correction—and it’s the critical barrier separating academic research from deployable devices. The company claims its proprietary phase correction algorithm can achieve individualized calibration in minutes, enabling transcranial focusing precision as fine as 1.5 mm—and asserts it is China’s first ultrasound-based BCI system with this algorithm ready for quasi-commercial deployment. Beyond modulation, BCI-Sonics aims to build a closed-loop system: using passive cavitation detection (PCD) to monitor real-time brain responses, combined with AI to recognize brain states and auto-generate stimulation protocols—creating a full cycle from reading to feedback to optimization. Rather than building just a stimulation device, its ambition is broader: a whole-brain, self-evolving “neuromodulation operating system.” Within China’s broader context, this funding round is no accident. 2025 is widely regarded as China’s Year of Brain-Computer Interfaces: the National Medical Insurance Bureau has separately funded BCI initiatives; seven ministries including the Ministry of Industry and Information Technology issued an industrial implementation plan; in January this year, China’s first BCI medical device standard came into effect; BCI was included in China’s “15th Five-Year Plan” as a future industry—and for the first time mentioned in the government work report; in March, the world’s first invasive BCI product received market approval in China and obtained a medical insurance code. Massive capital is pouring into this field. According to statistics, domestic BCI funding reached RMB 3.8 billion in Q1 2026—exceeding last year’s total—and surpassed RMB 4.4 billion by May; annual funding may break RMB 10 billion. Non-invasive unicorn StrongBrain raised approximately RMB 2 billion at the start of this year—one of the largest single investments globally outside Neuralink; semi-invasive company BrainCon is preparing for a STAR Market listing. Today, China’s BCI industry is maturing rapidly—with over 200 companies already active. The sector has moved beyond conceptual hype and is now focused on practical implementation and industrialization. The era when simply sketching a vision could secure funding has largely passed; investors are now concentrating on companies capable of delivering at least clinical-grade results. Leading BCI-Sonics is serial entrepreneur Li Xin, who holds a Ph.D. in biomedical engineering jointly awarded by the Chinese Academy of Sciences and Germany’s Fraunhofer IGD Institute. He previously served as head of GE Healthcare’s China R&D center and has founded multiple medical and tech startups. The core team comes from Tsinghua University, Imperial College London, Shanghai Jiao Tong University, Zhejiang University, and others—with expertise spanning neuroscience, acoustics, AI, and medical device registration. Li Xin’s bet on this venture is clear: The next decade of neuroscience will belong to platforms that can truly close the loop between “reading the brain” and “precise intervention.” (Source: Ifnar)