Remote Patient Activity Monitoring System by Integrating IoT Sensors and Artificial Intelligence Techniques

Area of Science:

• Biomedical Engineering
• Internet of Things (IoT)
• Artificial Intelligence (AI)

Background:

• Patient monitoring, especially for vulnerable groups and COVID-19 cases, presents significant challenges.
• Existing tools struggle with real-time tracking of vital signs and patient location.

Purpose of the Study:

• To develop and evaluate a wearable IoT device for continuous monitoring of COVID-19 vital indicators.
• To implement real-time GPS tracking for quarantine compliance and automatic alerts to medical authorities.
• To introduce a novel deep neural network for human activity recognition within the monitoring system.

Main Methods:

• A three-tier architecture comprising wearable IoT sensors, cloud data processing, and an Android mobile application.
• Integration of a novel deep neural network, CNN-UUGRU, combining convolutional and updated gated recurrent units for activity recognition.
• Real-time data analysis for vital sign assessment and GPS tracking for location monitoring.

Main Results:

• The CNN-UUGRU model achieved high accuracy (97.7%), precision (96.8%), and F-measure (97.75%) on diverse datasets.
• The system successfully monitored vital indicators and tracked patient GPS data.
• The system demonstrated superior performance compared to existing deep neural models and products.

Conclusions:

• The proposed wearable IoT system offers an effective solution for remote patient monitoring and quarantine management.
• The integrated CNN-UUGRU model significantly enhances human activity recognition capabilities.
• This technology holds promise for improving healthcare delivery and public health surveillance, particularly during pandemics.