Researchers at the Ecole Polytechnique Fédérale de Lausanne have developed a low-power, implantable neural system-on-chip named LogSync that records brain activity and extracts network-level biomarkers to treat severe psychiatric illnesses. Most existing deep brain stimulation systems operate in an open-loop manner using predefined settings, while closed-loop systems typically rely on signals from a single brain region. The new chip processes cross-regional synchrony and brain wave oscillations to generate stimulation triggers under 200 microwatts of power.
LogSync uses logarithmic computations for real-time brain analysis
Mahsa Shoaran, head of the Integrated Neurotechnologies Lab in EPFL’s School of Engineering and Neuro-X Institute, states that psychiatric disorders stem more from communication disruptions between brain regions than from dysfunction in any isolated area. To address this, the research team co-designed the algorithms and hardware of LogSync to operate within tight implant constraints. The device utilizes logarithmic computations to simplify complex operations required for real-time brain-network analysis, biomarker extraction, and state classification on a single chip.
NIMH grant funds validation in humans and rats
The development was funded as part of a CHF 6.6 million grant from the US National Institute for Mental Health aimed at controlling oscillatory brain networks for mood and anxiety disorders. The work was published in the IEEE Journal of Solid-State Circuits. Researchers evaluated the system using human brain recordings from previous experiments that captured cognitive engagement during mental tasks and emotional arousal during video viewing. Additional experiments in rats at the University of Minnesota showed that the implant successfully tracked ongoing oscillations and generated stimulation triggers at predefined phases of brain activity.
Clinical Applications and Infera Neuro
While LogSync’s initial validation focused on psychiatric applications, its capability to extract network biomarkers and trigger precisely timed stimulation applies broadly to adaptive deep brain stimulation for movement disorders, non-invasive stimulation, and focused-ultrasound interfaces. In parallel, the laboratory is developing machine-learning methods using self-supervised learning to capture neural activity across regions, channels, and time. This work supports broader efforts to bring intelligent neural interfaces toward clinical use, including through Infera Neuro, an EPFL spin-off co-founded by Shoaran.
Frequently Asked Questions About LogSync
How much power does the LogSync implant consume?
The system consumes under 200 microwatts of power, which is a fraction of the energy typically required by many implantable neural-interface systems.
What specific funding body supported this research?
The US National Institute for Mental Health provided a CHF 6.6 million grant to support the development of implantable systems capable of measuring and controlling oscillatory brain networks.
Where was the animal testing for the implant conducted?
Further experiments to track ongoing oscillations and generate stimulation triggers were conducted in rats at the University of Minnesota.
Keep reading