Enhancing ophthalmic examinations with real-time eye position monitoring using SVM

This work proposes OptoCheck, a modular and compact solution to monitor eye state and patient cooperation during ophthalmic exams. Our solution integrates an optical architecture, including a CMOS board-level camera and a dichroic mirror, and employs a Support Vector Machine (SVM)-based classification algorithm for real-time image analysis. The system is engineered to be compatible with current ophthalmic instruments, such as pupillometers and Optical Coherence Tomography (OCT), with minimal modifications. We tested the system on eight healthy subjects to evaluate its ability to detect eye movements, blinks, and small misalignments in real time. This approach improves the reliability of ophthalmic tests, shortens examination times, and supports more accurate and timely diagnoses, particularly in settings involving non-cooperative patients, such as pediatric or geriatric cases.