Resource optimization in smart electronic health systems using IoT for heart disease prediction via feedforward neural networks

Smart healthcare is the result of recent developments in cloud computing, artificial intelligence (AI), and the Internet of Things (IoT). These technologies have replaced the traditional healthcare system. Medical services can be enhanced by merging important technologies like artificial intelligence and the Internet of Things. The healthcare industry has several potential as a result of the convergence of artificial intelligence and the Internet of Things. From this angle, a convergence-based model for diagnosing heart disease based on IoT and artificial intelligence is presented in the current research study for a smart healthcare system. This paper’s primary objective is to use Internet of Things and artificial intelligence convergence techniques to create a disease detection model for cardiac disorders. The phases in the model that is being described involve gathering data, pre-processing, classifying, and configuring parameters. Wearables and other IoT devices make it easy to collect data, and AI tools analyze that data to diagnose diseases. For illness diagnosis, the suggested approach makes use of the cascaded short-term memory model (CuSO-MuLSTM), which is based on the Cuckoo search optimization algorithm. CuSO is used to modify the MuLSTM model’s “weight” and “bias” parameters in order to better classify medical data. Furthermore, outliers have been eliminated from this study project using the isolation forest technique (iForest). The diagnostic outcomes of the MuLSTM model are significantly improved by the application of CuSO. Using medical data, the CuSO-MuLSTM model’s performance was verified. The CuSO-MuLSTM model that was presented during the trial was able to diagnose diabetes and heart disease with maximum accuracy of 98.32% and 98.19%, respectively. As a result, the suggested CuSO-MuLSTM model can be used to smart healthcare systems as an appropriate illness diagnosis tool.