We built OcuMind around one question: why do patients with MS, TBI, or Parkinson's describe exhaustion and brain fog long before any test can confirm it? We thought the answer might be in how their eyes move.
Using just a regular webcam, OcuMind tracks 468 facial landmarks in real time and measures four key eye movement metrics tied to neurological conditions like Parkinson's, MS, and TBI. The hardest part was dealing with noise. Raw webcam data is messy, and blinks would throw everything off, so we built logic that freezes tracking during blinks and calibrates to each user's specific setup. That got us to 95% tracking stability on standard hardware.
The biggest surprise was realizing that once you treat gaze as a muscle that can be conditioned, and not just a symptom to be measured, the whole thing opens up. We started seeing the potential for everyday eye training. By gamifying our smooth pursuit and fixation exercises, users could actively combat severe digital eye strain and screen fatigue. We even started prototyping targeted routines for people with astigmatism, using micro-movement training to relieve the ocular muscle tension that makes focusing so difficult. That shift from a strictly clinical monitoring tool to an active, daily regimen for eyesight improvement and strain relief was something we didn’t plan for, but it ended up being one of the most exciting parts of the whole project.
Log in or sign up for Devpost to join the conversation.