AI POWERED SMART INTELLIGENCE-BASED POST COVID HEART DISEASE PREDICTION USING DEEP FEATURE HYPER CAPSULE NETWORK

The COVID-19 pandemic has significantly impacted global health, with its acute manifestations thoroughly studied; however, the long-term consequences particularly those related to cardiovascular health remain a pressing concern. Recent studies highlight a substantial increase in heart-related complications among post-COVID-19 patients, particularly in middle-aged populations. This emergent trend calls for intelligent, high-accuracy predictive systems capable of identifying at-risk individuals. In response, this paper introduces a novel AI-powered smart intelligence-based post-COVID heart disease prediction system, which integrates advanced machine learning and deep learning methodologies address limitations of traditional diagnostic approaches. At the core of the proposed system is a hybrid architecture combining a Support Vector Decision Scalar Algorithm (SVDCA) and a Deep Hypernet Capsule Convolution Neural Network (DHC-CNN). The system begins with Recursive Feature Z-score Normalization (RFZ-SN) to cleanse and standardize raw medical data, ensuring feature consistency and robustness. SVDCA then analyzes the Cardia Post Quadratic Impact Rate, extracting subtle yet critical acetate cardiac features and identifying deviations from ideal physiological norms. These extracted features undergo further refinement via Fuzzy Intensive Ant Colony Optimization, an intelligent feature selection algorithm that synergizes fuzzy logic with swarm intelligence to isolate the most relevant predictive variables. Subsequently, the refined feature set is input into the DHC-CNN, a sophisticated neural architecture that captures spatial hierarchies and dynamic relationships between features. The capsule-based structure preserves critical spatial dependencies in cardiac data, while the hypernet component allows the model to adjust internal parameters in response to input variations, enabling precise risks tratification for heart disease in post-COVID-19 patients. The system's performance is rigorously evaluated using metrics including precision, recal l, prediction accuracy, and false positive rate, and is benchmarked against traditional prediction models. Results demonstrate the superiority of the proposed system in accurately identifying at-risk individuals while minimizing diagnostic errors.