Integration of neural network in illness of heart disease for healthcare monitoring system

Since Heart Disease (HD) is a major global cause of mortality, improving patient outcomes requires both early detection and ongoing monitoring. The opportunity to implement an Internet of Things (IoT) solution has increased with the popularity of smart wearable devices. Regrettably, those who have unexpected heart attacks have a poor chance of surviving. In this paper, a Hybrid Salp Swarm Intelligence-Multi-Featured Recurrent Neural Network (HSSI-MFRNN) is proposed to enhance the ability to diagnose and track cardiac disease with greater precision and efficiency. First, patient health records are verified and uploaded using Run-Length Encoding (RLE) to compress and store the data. The Rivest-Shamir-Adleman (RSA) method is used to secure the transmission of patient data, offering a practical approach to protect sensitive medical data. The gathered data contains 70,000 records, each representing a synthetic patient healthcare record. The preprocessing step includes min-max normalization to ensure data consistency and dependability. This system uses the HSSI to optimize feature selection, identifying the most critical variables from extensive patient data and feeding these selected features into a sophisticated MFRNN architecture, which captures temporal dependencies and patterns in the data. The findings of this research show that the suggested system is beneficial in numerous areas, including data compression (DC) and secure data transfer. The comparative analysis, which has 0.921341 an accuracy, 0.934189 a precision, a (0.90135) F1-score, and 0.959532 an Area under the Curve (AUC) metric, shows that the proposed approach outperforms other traditional methods in HD diagnosis. The suggested method involves monitoring, making it a valuable tool for improving patient outcomes and reducing the financial impact of heart disease on healthcare providers.