Fog Computing Based on Hybrid MLP-LSTM for Hypertension Attack Detection

A prevalent condition that poses significant threats to the general health of people, hypertension requires early and precise diagnosis to prevent complications. Standard methods of hypertension monitoring that have been put in practice before face difficulties and inefficiencies in delivering signals on time. To mitigate the deficiencies, the study proposes a fog computing-based model for real-time identification of hypertension attacks utilizing an integrated MLP-LSTM model. Compared to cloud-based systems, fog computing lowers latency and enhances dependability since computations are performed at the network's periphery. Due to the need to enhance the detection accuracy, the identified hybrid MLP-LSTM model has been made to identify the nonlinear as well as the linear structure of the physiological data. Data obtained from wearable devices and used in the study are collected and cleaned to ensure that they are fit for use. Regarding data, training data is used, and a data validation is also used. Concerning models, various metrics are employed to evaluate these models. The outcomes prove that detection efficiency and response time of the proposed hybrid MLP-LSTM model implemented in the fog computing environment are tremendously high compared with the conventional algorithms in terms of 98% accuracy. Python is implemented in the proposed approach. Besides enhancing continuous tracking and early hypertension attack detection, this method offers a rapid and feasible way for healthcare IoT applications.