A Narrowband IoT Personal Sensor for Long-Term Heart Rate Monitoring and Atrial Fibrillation Detection

Area of Science:

• Biomedical Engineering
• Cardiology
• Internet of Things (IoT)

Background:

• Atrial fibrillation (AF) is a widespread cardiac condition requiring long-term patient monitoring for effective detection.
• Traditional Holter electrocardiographic (ECG) monitoring has limitations due to short recording durations.
• Current smartwatch photoplethysmography offers extended monitoring but provides only intermittent measurements.

Purpose of the Study:

• To develop and evaluate an Internet of Things (IoT) sensor system for continuous, long-term monitoring and real-time detection of atrial fibrillation (AF) episodes.
• To assess the performance of the proposed IoT sensor in accurately identifying AF arrhythmias.

Main Methods:

• An IoT sensor integrating an ECG Analog Front End (MAX30003), microcontroller (STM32F401RE), and narrowband IoT module (STEVAL-STMODLTE) was developed.
• Real-time heart rate extraction and transmission of inter-beat intervals to an IoT cloud platform (ThingSpeak) every two minutes.
• Atrial fibrillation detection software, implementing the Lorentz algorithm, was deployed on the cloud platform for analysis using the MIT-BIH Atrial Fibrillation Database.

Main Results:

• The developed IoT system demonstrated continuous recording and efficient transmission of heart rhythms.
• The system achieved high performance in recognizing atrial fibrillation episodes, with an overall accuracy of 0.88, sensitivity of 0.71, and specificity of 0.99.
• The proposed method effectively identifies AF episodes in real time.

Conclusions:

• The developed IoT sensor system offers an effective and efficient solution for continuous, long-term patient monitoring.
• The system shows considerable performance in real-time recognition of atrial fibrillation episodes.
• This technology has the potential to improve the detection and management of atrial fibrillation.