Research Projects

Neurological disorders, particularly those affecting the face and eyes, are not widely understood by the general public, yet they can cause significant long-term distress for individuals. Currently, diagnosing these disorders is largely limited to specialists, such as neurologists, who detect symptoms based on visible signs. However, emerging technologies like computer vision offer the potential to enable computational devices to perform similar diagnostic tasks, making both diagnosis and rehabilitation more accessible and affordable for patients while also saving valuable time for healthcare professionals. This proposal focuses on two specific neurological disorders: Bell's Palsy and Blepharospasm. Both conditions involve abnormal blinking patterns. In Bell’s Palsy, patients struggle to blink the eye on the affected side, leading to asymmetric blinking between the two eyes. In Blepharospasm, patients experience involuntary, repetitive, and forcible eyelid contractions, resulting in prolonged, tight blinking. The primary scientific objectives of this proposal are twofold: (i) to model these distinct blinking patterns, and (ii) to develop a diagnostic and rehabilitation system that leverages these models for improved patient care. Achieving these goals requires the development of a robust blink detection system capable of identifying even subtle, tight blinks, and creating specialized detection models for both disorders. To address challenges such as limited labeled data, we will employ a combination of deep learning and traditional machine learning techniques. With the rising prevalence of neurological disorders over the past two decades, partly due to increasing stress levels, this research aims to improve the management and treatment of these conditions. Currently, there are very few blink-based diagnostic and rehabilitation systems, making this area a promising field for exploration and development.