EPFL Implantable Neural Chip Decodes Brain Region Interactions

Researchers at EPFL have developed LogSync, a low-power implantable neural chip that records signals from multiple brain regions, extracts network biomarkers, and uses machine learning to identify cognitive and emotional states in real time. The device, created by Mahsa Shoaran’s team at EPFL’s Geneva campus, aims to improve adaptive deep brain stimulation for psychiatric disorders by detecting patterns distributed across neural networks rather than relying on a single brain region. Published in the IEEE Journal of Solid-State Circuits in September 2026 and supported by a CHF 6.6 million grant from the US National Institute of Mental Health, LogSync measures brain-wave strength and cross-regional synchrony, then triggers stimulation when target states are detected. In tests using human psychiatric recordings and rat experiments at the University of Minnesota, the chip classified task-related cognitive engagement and emotional arousal and delivered phase-specific stimulation. The team co-designed the hardware and algorithms and used logarithmic computations to keep power consumption below 200 microwatts, far lower than many implantable neural-interface systems. Although initially validated for severe mood and anxiety disorders, the approach could also support adaptive stimulation for movement disorders, non-invasive stimulation and ultrasound-based neuromodulation, bringing intelligent neural interfaces closer to clinical use.