CARDIOSIGHT: A MULTIMODAL FRAMEWORK FOR INTEGRATED CARDIOVASCULAR RISK ASSESSMENT

Abstract

Cardiovascular diseases remain a significant global health challenge, particularly in areas with limited access to early diagnosis and specialized care. Conventional diagnostic approaches often rely on invasive procedures, fragmented clinical tests, and expert judgment, which can delay timely detection. Retinal imaging has recently emerged as a powerful noninvasive tool, as retinal blood vessel patterns reflect underlying cardiovascular conditions. When combined with structured clinical data such as blood pressure, cholesterol levels, and glucose measurements, prediction accuracy can be greatly enhanced. However, most existing systems only provide binary results, failing to assess disease severity