Automated Heart Disease Diagnosis and Monitoring System Using Similarity-Navigated Graph Neural Networks with Leopard Seal Optimization in IoT Healthcare Applications

One of the leading causes of death worldwide continues to be heart disease; therefore, early detection and precise diagnosis are essential to better patient outcomes. The conventional methods of diagnosing chemotherapy-induced nail changes involve a great deal of laboratory testing, which is more than exhaustive. In order to tackle these obstacles, this research introduces a novel methodology for the automated diagnosis and continuous tracking of heart diseases in the IoT -based health care system, integrated with deep learning algorithms. The proposed system includes several state-of-the art techniques to improve diagnostic reliability and speed. First of all, min-max normalization is used as a pre-processing technique that aims at pre-scaling the data of patients to a consistent level and, as a result, enhancing the performance of the model. The Mud Ring Algorithm is then used for feature selection, whereby the best features are chosen together with the reduction in dimensionality to decrease computational intensity, thus improving diagnostic accuracy. The core diagnostic engine is based on Similarity Navigated Graph Neural Networks (SNGNN) and has been enhanced with Leopard Seal Optimization (LSO), a new meta-method motivated by the hunting patterns of leopard seals. This combination enables the system to relate various features and give the correct prognosis for heart diseases. These techniques, applied within the IoT framework, enable real-time monitoring and diagnosis and therefore have the benefit of being a scalable solution for modern health care systems. The results show that the proposed method for providing diagnosis of heart disease is accurate, timely, and reliable and thus has the potential to revolutionize patient management in healthcare based on the IoT.