LIFE‐CARE: IoT–Cloud‐Enabled Smart Heart Disease Prediction System for Smart Healthcare Environment Using Deep Learning

Heart disease prediction is a challenging task which assists medical authorities to make timely and precise decisions about patient health. Predictive analysis in smart healthcare can enable the reactive healthcare to be proactive with assistance of machine learning (ML) and artificial intelligence (AI) technologies. Existing works often lack complexity and heterogeneity as they trained similar inputs for their models and unaware of varied input data. This research proposed a fully automated heart disease prediction framework from the real‐time data named LIFE‐CARE that enables precise and timely predictive analysis. The propounded research utilized medical IoT sensors and electronic health records (EHRs) for acquiring the patient heart status and forwarded them to the multicloud server via gateway and distributed edge server (DES) for heart disease prediction. In between that, we preprocess the raw medical data in terms of data filtering and normalization at the DES. For heart disease prediction, we utilize collaborative transfer learning–enabled separable transformer (CTL‐ST) which processes the varied input by adopting domain adaptation and knowledge distillation techniques thereby achieving better prediction results. At that time of implementation, the proposed LIFE‐CARE model achieves performance results of accuracy of 98.72%, precision of 98.25%, recall of 98.79%, specificity of 98.55%, and F ‐score of 98.78% which is better than the state‐of‐the‐art heart disease prediction models.