New Neural Implant Adapts to Real-Time Brain Network States

EPFL researchers have created an implantable neural system that monitors real-time communication between brain regions to identify cognitive and emotional states. Called LogSync, the device uses machine-learning algorithms to extract biomarkers describing brain wave patterns and synchronization across networks, then triggers adaptive stimulation when specific states are detected. This network-level approach addresses a key limitation of existing systems, which either rely on fixed stimulation settings or monitor only single brain regions.

The innovation achieves all this while consuming under 200 microwatts of power by co-designing algorithms and hardware together and employing logarithmic computations that simplify the complex math needed for network analysis. Validation on human brain recordings showed the system could distinguish cognitive engagement and emotional arousal from neural signals alone, while animal experiments confirmed it could track ongoing brain activity and generate precise stimulation triggers. The approach has immediate applications for psychiatric disorders like depression and anxiety, where disrupted communication between regions matters more than single-region dysfunction.

Beyond psychiatry, LogSync’s ability to extract brain-network biomarkers and time stimulation precisely opens possibilities for adaptive treatment of movement disorders and emerging neurotechnologies. The work, published in the IEEE Journal of Solid-State Circuits and funded by the US National Institute of Mental Health, builds on the lab’s earlier research in brain-to-text decoding and closed-loop neuromodulation, extending those concepts to network-level analysis for future clinical applications.