EPFL LogSync chip tracks brain network biomarkers

Researchers at EPFL have developed the LogSync chip, a device that records brain activity and extracts biomarkers describing the strength of brain waves and synchrony between signals from different brain regions. The system includes a machine-learning algorithm that analyzes these network-level biomarkers to distinguish cognitive and emotional states and generate stimulation triggers. Mahsa Shoaran, head of the Integrated Neurotechnologies Lab in EPFL's School of Engineering and Neuro-X Institute, led the project. The development was supported by a CHF 6.6 million grant from the US National Institute for Mental Health.

The team evaluated LogSync using human brain recordings from previous psychiatric experiments involving cognitive engagement and emotional arousal. In separate experiments with rats at the University of Minnesota, the implant tracked ongoing oscillations and generated stimulation triggers at predefined phases of brain activity. Shoaran noted that a major challenge was achieving functionality within the limited power budget of an implantable system. The LogSync system keeps power consumption under 200 microwatts by co-designing algorithms and hardware with implant constraints in mind. It uses logarithmic computations to simplify complex operations needed for brain-network analysis and real-time biomarker extraction.

The work was published in the IEEE Journal of Solid-State Circuits in September 2026. The laboratory is now developing new machine-learning methods that learn directly from neural recordings rather than predefined biomarkers. Recent work presented at ICML 2026 uses self-supervised learning to capture neural activity across regions, channels, and time. Infera Neuro, an EPFL spin-off co-founded by Shoaran, is involved in bringing intelligent neural interfaces closer to clinical applications. The LogSync chip's capability to extract brain-network biomarkers is broadly applicable to adaptive deep brain stimulation for movement disorders and closed-loop non-invasive stimulation. The project builds on the lab's previous work on brain-to-text decoding and closed-loop neuromodulation.