IoT-Based Heart Disease Monitoring Using Neural Networks for Detecting Arrhythmia and Assessing Hypertrophic Cardiomyopathy(HCM) Risk

Advancements in deep learning and IoT technology have the potential to revolutionize healthcare, as evidenced by recent developments in remote health monitoring systems. Even if they are cutting-edge, current systems frequently fall short regarding cost-effectiveness, patient-centered treatment, and real-time accuracy. Conventional approaches that rely on hospital-centric models are less accessible and resource-intensive, especially in rural areas. Furthermore, a lot of IoT-based solutions have trouble with early intervention and accurate data, both of which are necessary for managing chronic disorders like heart disease. This research offers a novel IoT system for monitoring remote cardiac illnesses, such as arrhythmia and hypertrophic cardiomyopathy (HCM). The MLX90614 non-contact body temperature sensor, the AD8232 ECG sensor, and the heart rate sensor are integrated into the system along with NodeMCU for data collecting and storage. A Deep Neural Network (DNN) model for heartbeat classification is trained using the information gleaned from these sensors, which is then saved in Excel sheets for examination. Our methodology, which shows an impressive 98% accuracy rate, emphasizes using a simplified DNN model trained on Excel data to achieve accurate heartbeat predictions, in contrast to traditional CNN-based approaches. Additionally, a web application has been provided to improve accessibility and facilitate early detection of potential health issues, optimizing patient care. With continued study and development, the suggested approach promises to improve accuracy and efficiency even more, marking a substantial improvement in home healthcare technology.