IoT-Enabled Multifunctional Biomaterials for Continuous Health Sensing and Intelligent Digital Healthcare Service Loops

Abstract

The continuous monitoring of health using IoT-based multifunctional biomaterials for personalized healthcare is possible, but the existing systems are limited by noise, signal variations, multimodal complexity and unreliable health-risk prediction. To address the limitation, this research proposes an algorithm called Vital Intelligence Temporal Adaptation Learning Integrated Sensing Network (VITALIS-Net), which combines three key approaches, Adaptive Kalman Filter (AKF), Multimodal Temporal Transformer (MTT) and Deep Evidential Learning (DEL). Continuous signals generated by multifunctional wearable sensors based on biomaterial are collected using IoT network. In this context, AKF does adaptive denoising and measurement corrections, while MTT used in extracting temporal dependencies and fusion of multiple modalities from the sensors. Finally, DEL facilitates uncertainty-based health-risk predictions. The developed framework is a digital healthcare service loop of continuous sensing, health risk detection, intelligent feedback and re-monitoring. This research evaluates the performance of the framework with respect to predictive reliability against drifting sensors, missed measurements and changes in physiological states. Health-risk classification accuracy obtained using the proposed VITALIS-Net is 98.2%. Overall the VITALIS-Net can support and efficiently personalized healthcare with dependable adaptive signal denoising, multimodal temporal analysis, health risk predictions with uncertainty and intelligent monitoring.