An Intelligent IoT and AI Framework for Real-Time Remote Patient Monitoring and Healthcare Management

Abstract

This study proposes an integrated Artificial Intelligence (AI)-driven Internet of Things (IoT) framework for real-time remote patient monitoring and intelligent healthcare management. The proposed framework brings together physiological data acquisition, preprocessing, intelligent analysis, and healthcare decision support within a unified monitoring architecture. To establish the data foundation for the proposed framework, the study analyzes 53 patient recordings from the BIDMC PPG and Respiration Dataset. The recordings contain complementary cardiovascular, respiratory, oxygenation, and demographic information, including heart rate, pulse rate, respiratory measurements, oxygen saturation, age, and gender. Descriptive analysis was conducted to examine the availability and characteristics of the physiological measurements and to identify differences in the completeness of individual variables. The analysis demonstrates that combining multiple physiological parameters can provide a broader basis for continuous remote monitoring than relying on a single measurement. Variations in data availability also emphasize the need for appropriate preprocessing and data-quality management before intelligent analysis is applied. Based on these findings, the proposed framework establishes a structured pathway from physiological sensing and data communication to preprocessing, AI-oriented analysis, and healthcare decision support. The framework is intended to support future development of automated monitoring and timely identification of potential physiological abnormalities. However, the study does not report empirical AI model training, clinical validation, or physical IoT deployment. Future research should therefore focus on developing and validating AI models, implementing real-time IoT connectivity, incorporating explainable AI and secure communication mechanisms, and evaluating the framework using larger and clinically diverse datasets.