Vital sign analysis and correlation of physiological parameters using edge AI and IoT technology

Individuals encounter discomfort in their lifestyle with vital signs and problems that lie in their ability to identify and respond to life-saving situations with advanced health care technologies. Vital signs are essential parameters that provide an early warning of a diagnosis before further deterioration in health. Progress in health care technologies, such as automated health monitoring systems and telehealth systems, which can detect, collect real-time data, and diagnose remotely. The present article aims to measure vital signs such as heart rate (HR), temperature (T), oxygen saturation (SpO 2 ), and blood pressure (BP) using a designed medical device by implementing a quantized Gradient Boosting (GB) model deployed on the ARM Cortex-A72 processor for automated clinical risk stratification. The proposed system was evaluated on two benchmark datasets: The Medical Information Mart for Intensive Care for Emergency Department (MIMIC-IV-ED v2.2) dataset and a real-world clinical dataset. The accuracy of measured parameters was evaluated using parametric statistical analysis of baseline metrics and Gaussian distribution fitting of the recorded data from MIMIC-IV-ED. The (GB) algorithm achieved an accuracy of 99.84%, precision of 99.40%, and attained a lower false positive rate (FPR) of 0.06%. The quantized edge AI (INT8) model, when deployed on the processor, attained a latency of 0.596 ms. Hence, the designed system provides an efficient, low-latency solution for real-time vital sign monitoring and clinical risk-based classification.