EPFL builds implantable chip LogSync that decodes cross-brain-region signals
What happened
An October 24 report says an EPFL team developed LogSync, a low-power implantable neural chip that records brain activity and extracts biomarkers such as brainwave intensity and signal synchrony across brain regions. Machine learning algorithms then classify cognitive and emotional states and trigger stimulation. The chip measures about 4×2 mm, draws under 200 microwatts and uses logarithmic computation to simplify brain network analysis. The work was validated on human brain recordings and rat experiments, and the paper appeared in IEEE Journal of Solid-State Circuits.
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageNeural chip decodes interactions between brain regions
EPFL 团队开发出低功耗可植入神经芯片 LogSync，可记录脑活动并提取脑波强度与跨脑区信号同步性等生物标志物，再由机器学习算法区分认知与情绪状态并触发刺激。芯片尺寸约 4x2 mm，功耗低于 200 微瓦，采用对数计算简化脑网络分析。研究基于人类脑记录与大鼠实验验证，论文发表于 IEEE Journal of Solid-State Circuits。
Heat over time
Not enough continuous observations to draw a trend yet.