Hybrid Adam Coati Optimization with DLSTM for Biomedical Parameter Prediction Using Sensor

A real-time patient monitoring system procesfses a wide range of biomedical parameters and can store and safely capture biomedical parameters. It lowers the diagnosis time of the several diseases that are prevalent in healthcare services. Accurate prediction of biomedical parameters can significantly reduce healthcare costs and assist in effective decision-making. This research proposes a deep learning approach with hybrid optimization for biomedical parameter prediction by using sensors. At first, a healthcare centre with many patients having various disorders is considered. Several biomedical sensors are placed on the patients and these sensors include a respiratory Rate (RR) sensor, Heart rate (HR) sensor, Pulse Rate (PR) sensor, Ambulatory Blood Pressure (ABP)-sensor, and Saturation of Peripheral Oxygen (SpO2). Based on these sensed parameters, biomedical parameter prediction by using sensor based Deep Long Short-Term Memory (DLSTM)is executed. The DLSTM is trained using the proposed Hybrid Adam Coati Optimization(HAdCO). Here, the HAdCO is developed by incorporating of Coati Optimization Algorithm (COA) with Adam Optimizer. Furthermore, the HAdCO-DLSTM is evaluated by utilising metrics, like Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Squared Error (MSE). The HAdCO-DLSTM had a minimal value of MAE at 0.0063, MSE at 0.0062, and RMSE at 0.0078.