BCI-Sonics raises $28M Pre-A for non-invasive ultrasound brain-computer interface with AI closed loop

机器之心 · wechat · 2026-09-22

BCI-Sonics, a Shanghai startup founded in 2025 by Li Xin, has raised a 200M yuan ($28M) Pre-A round co-led by Sequoia China and Yunqi Capital, bringing total funding to nearly 300M yuan.

Approach: non-invasive + AI

Write: low-intensity transcranial focused ultrasound (TUS) reaches deep brain regions through the skull; reported focusing precision has improved from 1.5mm to under 1mm, with phase compensation in 5 seconds
Read: multimodal sensing combining EEG and functional ultrasound (fUS)
Decode: AI handles navigation, acoustic field computation, phase correction, brain-state recognition and parameter optimization, aiming for a closed "read–understand–modulate–feedback–re-optimize" loop

Progress

Two generations of FUS-BCI products completed with safety reports; a closed-loop system enters clinical trials in Q1 2027
Launched an AI platform for ultrasound neuromodulation parameter interaction plus research tools; 4-5 devices already in research labs
Commercial path: research tools first, clinical work on pain, Parkinson's, Alzheimer's, addiction and sleep, and eventually consumer wearables for sleep/mood/focus
Long-term NeuroAI vision: building neural data foundations and exploring what neuroscience can teach next-generation AI

Original post →

BCI-Sonics raises $28M Pre-A for non-invasive ultrasound brain-computer interface with AI closed loop

More from Venture