Israeli startup looks to harness AI to decode the ‘language’ of the brain

Israeli startup Hemispheric, co-founded by Gidi Littwin and Hagai Lalazar, has developed an AI model called Descartes to decode the brain's electrical activity for diagnosing neurological and psychological conditions like depression and PTSD. The model was trained on over 250,000 hours of non-invasive EEG brain data from 100,000 volunteers, creating the world's largest dataset of its kind. Patients wear a dry EEG headset for 15 minutes while interacting with a gamified app, allowing the AI to interpret brain function.

"The problem that we are coming to solve is really paramount to humanity, which is that our brain is the most complicated organ in our body, the largest disease burden, but we have no quantitative, non-invasive measurement of its function, of its health, and as a neuroscientist — that really bothered me," said Lalazar.

Hemispheric recently raised $52 million to pursue regulatory approvals in the US, Europe, and Israel, and plans to submit its first application for PTSD to the FDA early next year, aiming for a product rollout in late 2027. While promising, experts like Dr. Nirit Lev from Clalit-Meir Medical Center note the need for extensive clinical validation across diverse demographics and acknowledge the inherent limitations of EEG data.