Predictive Healthcare Monitoring Using Iot-Integrated AI Systems with Sensor Data

Abstract

Predictive healthcare monitoring systems have attracted much interest due to the introduction of artificial intelligence and the Internet of Things (IoT). This research paper presents a highly effective predictive healthcare monitoring system by processing multivariate physiological signals using the WESAD (Wearable Stress and Affect Detection) dataset. The proposed method involves the processing of physiological signals using sliding window segmentations and important preprocessing steps like z-score normalization, linear interpolation, and low-pass filtering. Statistical features are generated and used by a Temporal Fusion Transformer (TFT), which is capable of handling temporal correlations along with the importance of the features through its attention mechanism. The use of a classifier that uses the sigmoid function helps in identifying whether the data set represents normal or abnormal cases. According to the experimental results, the proposed model outperforms other conventional Machine Learning and Deep Learning models with accuracy, precision, recall, and F1-score values greater than 98%. The results highlight the effectiveness of the proposed framework for accurate and efficient predictive healthcare monitoring.