Sheba Medical Center, in collaboration with Israeli startup Revealense, has unveiled a new artificial intelligence-powered diagnostic platform designed to transform the way autism and ADHD are assessed. The system uses a standard webcam to analyze subtle behavioral cues through video, enabling fast, accurate and clinically validated diagnoses from home.
Dr. Omer Bar-Yosef, director of Sheba’s Child Development Center, said demand for neurodevelopmental assessments has soared over the past two decades, while the number of qualified professionals has not kept pace.
“The gap between the number of children needing evaluations and the availability of specialists has grown dramatically,” he said. “Today, diagnostic processes are cumbersome, requiring multiple in-person sessions with neurologists, psychologists, and psychiatrists and families often face months-long waits.”
According to Bar-Yosef, roughly 3% of children in the U.S. are diagnosed with autism, while 10% to 15% are diagnosed with ADHD. Traditional assessments involve lengthy interviews, standardized tests and complex perceptual evaluations that can stretch over weeks or even months.
Revealense’s AI platform, developed over several years, integrates patented technologies to analyze hundreds of micro-behavioral markers, such as facial expressions, tone of voice, micro-movements, gaze patterns and reaction times.
Unlike existing diagnostic tools that focus on isolated signals — like voice, eye tracking, or facial cues — the system combines multiple streams of data into a unified neurological profile.
“The innovation lies in its simplicity,” said Revealense CEO Dov Donin. “All you need is a webcam. There are no sensors, no wires and no lab visits. Our algorithms cross-analyze countless data points, giving us a diagnostic accuracy rate exceeding 93%, which was nearly impossible until now.”
The system works by presenting patients with short video-based tasks designed to trigger emotional, cognitive and social responses. These range from identifying human and animated figures to interpreting emotional scenarios and solving visual puzzles.
Revealense CEO Dov DoninPhoto: Eyal TuagSubtle differences in focus, empathy and problem-solving strategies — often imperceptible to the human eye — are captured and analyzed by the platform.
For example, Donin explained, children on the autism spectrum often fixate on background objects rather than human faces, or focus more on the mouth than the eyes when observing people. Similarly, children with ADHD may struggle to track moving objects or maintain attention on the central figure in a scene.
The AI engine adapts assessments across cultural contexts and is designed to minimize biases by cross-referencing behavioral baselines against large, diverse datasets. It also supports repeated testing to improve reliability.
“Unlike traditional evaluations, where a child’s mood on a single day can skew results, our system allows for multiple assessments over time,” said Dr. Bar-Yosef.
The joint venture between the hospital’s ARC innovation center and Revealense aims to broaden the platform’s use beyond autism and ADHD to cover a wide range of neurological, cognitive, and developmental conditions. Revealense plans to release the system to the public in the first half of 2026, pending regulatory approvals.