Advanced Health Care Monitoring in IoMT Systems through Integrated CNN and LSTM

Introduction: The developments in Internet of Medical Things (IoMT), communications technologies, and wearable sensors have made it possible for humans to live wiser lives, all thanks to pervasive computing. This has led to better healthcare services. The IoMT has the ability to completely transform the healthcare system. Caretakers, medical professionals, patients, and wearable sensors linked to software and ICT make up IoMT. Notable among the many growing industries with enormous demand is healthcare. Objectives: This study uses the Chaotic Satin Bowerbird Optimization Algorithm (CSBOA) for feature selection in conjunction with Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to create an optimal heart disease diagnosis system. Improving classification performance while lowering computational complexity is the aim. Methods: The incorporation of cognitive behaviour into IoT technology has also attracted the attention of several researchers. Combining CNN and LSTM networks with a Chaotic Satin Bowerbird Optimisation Algorithm (CSBOA) for feature selection is the proposed hybrid classical for heart disease detection in this study. To improve computing performance and decrease dimensionality, the CSBOA isolates the most important attributes. Results: Convolutional neural network–long short-term memory (CNN–LSTM) construction enables strong classification by capturing data patterns in both space besides time. The suggested model outperformed previous approaches using the UCI heart disease dataset, achieving superior presentation indicators such as high recall, accuracy, precision, besides F1-score. Conclusions: By using CSBOA, feature selection was further optimised, achieving a happy medium between precision and computational burden. Offering a dependable and efficient tool for heart disease detection, this hybrid technique shows great promise for real-world healthcare applications.