mmWave-Enabled Human Activity Monitoring for Telemedicine-Based Remote Patient Monitoring

Abstract

Telemedicine technologies have been advanced due to the growing need of constant and non-contact monitoring of patients. The given paper introduces a Human Activity Monitoring (HAM) system that is built on the millimeter-wave (mmWave) radar and telemedicine-based Remote Patient Monitoring (RPM) platforms. The system is capable of detecting and classifying the key activities of the patient with high spatial resolution and motion-sensing capabilities of mmWave radar, such as fall detection, posture recognition, gait analysis, and respiration, without wearable devices or cameras. This guarantees reliability and privacy of the patient. The proposed framework is based on the integration of edge computing to real-time signal processing, deep learning-based activity recognition (CNN-LSTM), and the HL7 FHIR-compliant telemedicine interfaces. Efforts to experimentally test a 60 96.3, fall detection sensitivity 94.7, respiration rate error error 3breaths/min experimental evaluation with a 60 96.3, fall detection sensitivity 94.7, respiration rate error and end-to-end latency for clinician notifications over a 5G/mmWave link was performed in a simulated home baseline and a clinical baseline environment. The wearable-based systems were found to be more private, have fewer false alarms (approximately 12 percent), and work well in the dark and hidden environments in comparison to the camera-based systems. Such findings substantiate the fact that mmWave radar is a feasible, privacy-preserving sensing modality of telemedicine-driven healthcare, which will underpin future AI-enabled, 6G-integrated, and contactless remote patient monitoring protocols.