声波在穿透异质颅骨时会发生折射与散射。我们的混合仿真引擎可计算并补偿相位差。
Refraction & Defocusing (散射)
Precise Focusing in 1.5mm
基于自研 AI 模型（AI），针对每位患者的颅骨结构自动计算，生成个性化治疗参数。
解决个体差异导致的疗效不一致
通过多层环阵与相控偏转技术，实现从深部丘脑到浅表皮层的全网络覆盖调控。
*Supported by recent studies (Mesik et al., 2024)
传统的"电"与"磁"神经调控技术面临着“深度浅”与“精度低”的物理瓶颈。
tFUS (经颅聚焦超声) 是目前唯一能同时实现深部脑区覆盖与毫米级精准靶向的无创技术。
经颅磁刺激
局限性：
空间分辨率较低（厘米级），难以在不刺激浅表皮层的情况下精准触达深部核团。
经颅电刺激
局限性：
电流在大脑中广泛扩散，弥散性强，缺乏聚焦性，难以针对特定神经环路进行干预。
时间干涉电刺激
局限性：
物理聚焦原理限制了空间分辨率（难达亚毫米级）。调控策略个性化更改难度高，难以灵活适配个体差异。
低强度聚焦超声
全能冠军：
利用声波物理聚焦。兼具深部直达、<1.5mm 精度与AI 灵活性。真正实现“指哪打哪”。
BCI-Sonics 坚定选择 LIFU (低强度聚焦超声) 路线。与产生高温破坏组织的 HIFU 不同，我们利用超声波的纯粹机械效应来调控神经活动。
颅骨异质性是巨大挑战。我们的系统可在 <10秒内完成超声经颅骨三维声场重建与相位补偿， 将靶点定位精度控制在 1.5mm以内。
全球首个实现 深部核团（如丘脑）与 浅表皮层 协同刺激的系统，临床获益更显著。
拒绝千篇一律。BCI-Sonics 利用全栈自研的 AI 引擎，
将传统的“经验医学”转化为基于数据的“精准计算”，确保每一次治疗都精准、有效。
传统声场仿真耗时极长。利用自研的 AI 模型与物理信息神经网络，我们可在 <10秒 内完成完整的三维声场重建与颅骨声衰减校正。
无创神经调控的最大挑战是“疗效不一致”。我们的 AI 引擎在海量参数空间中，为每一位患者寻找最优解。
AI 不仅调整功率，更在声强、占空比、脉冲时序、波束成形等亿级参数组合中进行全局寻优。
根据患者颅骨透声率和靶点深度，自动微调入射角度，确保靶点处的有效刺激剂量一致。
显著提高治疗响应率，解决行业长期存在的“同样参数，不同效果”的难题。
实时捕捉脑电信号（如 Alpha/Theta 节律）。AI 根据反馈判断神经兴奋度，自动调整刺激时序 (Phase-Locked)。
实时监听微泡空化信号。一旦检测到惯性空化，系统毫秒级降低声强，确保全程处于安全的稳态空化区间。
汇聚神经科学、声学物理与人工智能的顶尖智慧。
核心团队拥有从原理样机、临床注册到商业化落地的完整医疗器械全生命周期经验。
Scientific Visionaries
Engineering Architects
Global Strategy
精选代表性成果展示
提出轻量化模型，在普通设备上实现医疗影像的即时分割与重建，为术前快速规划奠定基础。
在数据受限情况下实现高精度的器官/脑区分割算法，核心技术已迁移至脑区靶点识别中。
解决了复杂脑血管网络的自动识别难题，为避开血管的安全路径规划提供算法支撑。
攻克了复杂声学环境下的快速仿真难题，验证了电磁兼容性方案。
深厚的科研积淀
累计发表 50+ 篇同行评议论文，总影响因子 > 150
期待您的垂询，携手推进脑科学与智能医疗的未来。
办公地址 (Office Location)
上海巿长宁区金钟路968号
天会商务广场6号楼605室 (邮编: 200335)
合作与咨询 (General Inquiry)
info@bcisonics.com
The ultimate solution for "individual variability" in non-invasive neuromodulation.
Integrating tFUS Technology with
AI to automatically generate precise treatment
protocols for every patient.
Sound waves refract and scatter when penetrating heterogeneous skulls. Our hybrid simulation engine calculates and compensates for phase aberrations.
Refraction & Defocusing
Precise Focusing in 1.5mm
Utilizing our proprietary AI model, we automatically calculate and generate personalized treatment parameters based on each patient's skull structure.
Solving inconsistency caused by individual differences
Multi-layer ring arrays and phased deflection technology enable full network coverage modulation from the deep thalamus to the superficial cortex.
*Supported by recent studies (Mesik et al., 2024)
Traditional "Electric" and "Magnetic" technologies face physical bottlenecks of "Shallow Depth" and
"Low Precision".
tFUS (Transcranial Focused Ultrasound) is the only non-invasive technology capable
of achieving both deep brain coverage and millimeter-level precision.
Transcranial Magnetic Stimulation
Limitation: Low spatial resolution (cm level), difficult to reach deep nuclei accurately.
Transcranial Direct Current
Limitation: Current diffuses widely, lacking focus.
Temporal Interference
Limitation: Limited spatial resolution, hard to customize.
Low Intensity Focused Ultrasound
All-Around Champion: Deep Reach, <1.5mm Precision, AI Flexibility.
BCI-Sonics firmly chooses the LIFU (Low-Intensity Focused Ultrasound) route. Unlike HIFU which destroys tissue with high heat, we utilize the pure mechanical effects of ultrasound to modulate neural activity.
Skull heterogeneity is a huge challenge. Our system can complete transcranial 3D acoustic field reconstruction and phase compensation in <10 seconds, controlling target positioning accuracy within 1.5mm.
The world's first system to achieve synergistic stimulation of deep nuclei (such as the thalamus) and the superficial cortex, resulting in significantly better clinical outcomes.
Rejecting the "one-size-fits-all" approach. BCI-Sonics uses a full-stack self-developed AI engine
to transform traditional "empirical medicine" into data-based "precision computation", ensuring every treatment is precise and effective.
Traditional acoustic field simulation takes extremely long. Utilizing our self-developed AI model and physics-informed neural networks, we can complete full 3D acoustic field reconstruction and skull attenuation correction in <10 seconds.
The biggest challenge in non-invasive neuromodulation is "inconsistent efficacy". Our AI engine searches for the optimal solution for each patient in a massive parameter space.
AI adjusts not only power but also globally optimizes across billions of parameter combinations like intensity, duty cycle, pulse timing, and beamforming.
Automatically fine-tunes the incident angle based on patient skull acoustic transmittance and target depth to ensure consistent effective stimulation dose at the target.
Significantly improves treatment response rates, solving the long-standing industry problem of "same parameters, different results".
Real-time capture of EEG signals (e.g., Alpha/Theta rhythms). AI judges neural excitability based on feedback and automatically adjusts stimulation timing (Phase-Locked).
Real-time monitoring of microbubble cavitation signals. Once inertial cavitation is detected, the system lowers intensity in milliseconds, ensuring the entire process remains in the safe stable cavitation range.
Bringing together top wisdom from neuroscience, acoustic physics, and artificial intelligence.
The core team possesses complete medical device lifecycle experience from
prototype, clinical registration to commercialization.
Scientific Visionaries
Engineering Architects
Global Strategy
Selected Representative Achievements
Proposed lightweight models enabling instant segmentation and reconstruction of medical images on standard devices, laying the foundation for rapid preoperative planning.
Achieved high-precision organ/brain segmentation algorithms with limited data; core technology has been transferred to brain target identification.
Solved the challenge of automatic identification of complex cerebrovascular networks, providing algorithmic support for safe path planning to avoid blood vessels.
Overcame challenges in rapid simulation under complex acoustic environments and validated electromagnetic compatibility solutions.
Deep Research Heritage
Published 50+ peer-reviewed papers, Total Impact Factor > 150
We look forward to your inquiry and working together to advance the future of brain science and intelligent medicine.
Office Location
Room 605, Building 6, Skybridge HQ
No. 968 Jinzhong Road, Changning District
Shanghai, China (Zip: 200335)
General Inquiry
info@bcisonics.comWe are always looking for top talents in acoustic physics, neural engineering, and AI algorithms. Join us in defining the next generation of medical technology.