Our lab's pioneering work investigates the intricate connection between the eye and the brain, particularly in the context of neuropsychiatric disorders like schizophrenia and bipolar disorder. Through advanced retinal imaging techniques, we explore how subtle changes in the eye's microvasculature and structure can serve as powerful biomarkers for understanding, classifying, and even predicting brain health.

Our key findings demonstrate:

• The Eye as a Diagnostic Window: Retinal vascular abnormalities, including changes in caliber, tortuosity, and fractal dimension, are consistently associated with schizophrenia and bipolar disorder.

• Insights into Brain Structure & Function: We've established links between retinal measures and brain white matter lesions, as well as cognitive impairments like working memory deficits.

• Predictive Potential: Our deep learning models show promise in classifying psychiatric disorders based on retinal images, paving the way for non-invasive diagnostic tools.

• Broader Applications: Our systematic reviews and meta-analyses extend these observations to other conditions like major depressive disorder, and even explore macular perfusion in diabetic retinopathy, showcasing the eye's broad utility in systemic health.

Ultimately, our research underscores the profound utility of the eye as an accessible "window" for gaining crucial insights into complex brain conditions, aiming for earlier detection, better understanding, and improved patient outcomes. 

Currently, we are working on First Degree relatives of Psychiatric Disorders, Retinal Vasculature extraction, and Deep Learning Algorithms for classfication.

in Collaboration with Ivy Tso Lab, OSUMC, USA

This part of the research focuses on developing AI-driven diagnostic models with explainable AI (XAI) to enhance transparency, accuracy, and speed in medical imaging and disease classification. The research work spans mental health, oncology, neuroscience, and agricultural disease detection, addressing challenges in early detection, interpretability, and computational efficiency.