Enhancing Cardiac Disease Prediction Through Data Recovery and Deep Learning Analysis of Electronic Sensor Data

Remote health monitoring plays a pivotal role in tracking the health of patients outside traditional clinical settings. It facilitates early disease detection, preventive interventions, and cost-effective healthcare, relying on electronic sensors to collect essential data. The accuracy of medical data analysis is paramount for early disease identification, patient treatment, and optimizing social services, particularly as data utilization expands within the biomedical and healthcare sectors. However, the presence of incomplete or inconsistent data hampers the accuracy of analysis. This paper introduces a novel approach, employing Grey Wolf Optimization-based Convolutional Neural Networks (GW-CNN), to recover missing data and enhance cardiac disease identification. The proposed method combines data imputation techniques for identifying and predicting missing values in electronic sensor data, followed by feature extraction to capture relevant information. The CNN model leverages Grey Wolf Optimization to improve its predictive capabilities for cardiac disease. Comparative evaluation against existing models assesses the new model’s performance in terms of specificity, accuracy, precision, recall, and F1 score.