Wireless Sensor Network Based Home Elderly Care Medical Detection System using Deep Learning Model

Preventing falls is more important for senior health because these incidents can lead to lasting injuries, especially impact their well-being. The effective solution is elderly care for fall detection and prevention based on the Wireless Sensor Network (WSN). However, most existing fall detection methods detect and notify after an elderly fall occurs. To solve this problem, a Deep Learning (DL) model Long ShortTerm Memory, and BackPropagation (LSTM-BP) are proposed to detect and prevent elderly falls by using wireless sensors. Initially, the collected data are fed as input to the preprocessing, and features are then extracted by the butterfly loop, Further the features are passed to feature selection method and finally with the useful features, the proposed LSTM combined with the BackPropagation network minimizes the error and enhances the detection process effectively. The experimental analysis indicates promising results for fall detection with 98.6% of accuracy, 98.4% of precision, 98.1% of Specificity, and 98.3% of Sensitivity which is greater than the existing methods like LSTM, Precondition and Limit Threshold- Sequential Probability Ratio Test (PLT-SPRT), Convolutional Neural Network (CNN), Support Vector Machine (SVM), and CNN-LSTM.