Gidi Littwin, co-inventor of Apple’s FaceID and Vision Pro technology, has launched Hemispheric, a startup using AI to decode brain activity for cognitive disorder diagnosis. The company raised $52 million in funding after gathering data from 100,000 people to train deep learning models that analyze brain function without invasive procedures. Littwin left Apple in 2020 and joined his cofounder Hagai Lalazar, who had been developing AI to study the brain non-surgically. Littwin said, 'There were massive data collection operations behind these projects and we knew we had to build something very similar at Hemispheric, and we have.'
Hemispheric’s AI model uses electrical activity data from 250,000 hours of brain activity collected from volunteers in Asia, Tel Aviv, and Boston. The data was gathered through activities resembling games that activated different brain regions. The model infers brain function from electrical signals, similar to how large language models analyze text. The team tested the model on individuals with PTSD, schizophrenia, and depression, reporting accurate deductions about brain health. They are now working on a clinical study to assess the model’s ability to diagnose and predict Alzheimer’s. The first product, aimed at studying PTSD, will be submitted to the FDA for approval early next year, with a public rollout planned for 2027.
Hemispheric’s AI-assisted diagnostic tool will help clinicians interpret brain signals from a lightweight EEG headset worn for 15 minutes while using a tablet app. The model will assist in making diagnoses, predicting treatment effectiveness, and monitoring progress. Lalazar described the future as 'a blood test' for mental health, with the device being affordable and widely distributed in clinics and hospitals. The company also plans to develop its own brain scanners to gather more detailed data than traditional EEGs, as Littwin noted, 'These devices were never built for machine learning and definitely not deep learning.'
Source: wired
