M2AI-CVD: Multi-modal AI approach cardiovascular risk prediction system using fundus images

Area of Science:

• Cardiology
• Artificial Intelligence
• Medical Imaging

Background:

• Cardiovascular diseases (CVD) are a leading global health concern, often diagnosed late.
• Non-invasive CVD detection methods, like retinal imaging, are under-researched in regions such as Qatar.
• Early detection is crucial for timely intervention and improved patient outcomes.

Purpose of the Study:

• To develop and validate a novel Multi-Modal Artificial Intelligence for Cardiovascular Disease (M2AI-CVD) system.
• To enhance the accuracy and efficiency of non-invasive CVD identification.
• To address the limited research on advanced CVD detection techniques in specific geographical areas.

Main Methods:

• Image quality assessment and processing for lower-quality inputs.
• Entropy-based Fuzzy C Means (EnFCM) for precise image segmentation.
• Multi-Modal Boltzmann Machine (MMBM) for feature extraction and Genetic Algorithm (GA) for feature selection.
• ZFNet Convolutional Neural Network (ZFNetCNN) for classifying CVD and Non-CVD cases.

Main Results:

• The M2AI-CVD system demonstrated high performance across five datasets.
• Achieved an accuracy of 95.89%, sensitivity of 96.89%, and specificity of 98.7%.
• The multi-modal AI approach proved effective in distinguishing between CVD and Non-CVD cases.

Conclusions:

• The M2AI-CVD system offers a promising, accurate solution for early cardiovascular disease detection.
• This multi-modal AI approach can significantly improve patient outcomes through timely interventions.
• Highlights the potential of AI in advancing non-invasive diagnostic tools for cardiovascular health.