A Wearable Electrocardiogram Telemonitoring System for Atrial Fibrillation Detection

Area of Science:

• Biomedical Engineering
• Cardiology
• Digital Health

Background:

• Atrial fibrillation (AF) detection requires continuous monitoring.
• Existing methods may lack long-term monitoring capabilities.
• Wearable technology offers a solution for remote patient monitoring.

Purpose of the Study:

• To develop and evaluate a wearable ECG telemonitoring system for AF detection.
• To integrate smartphone and cloud computing for real-time AF analysis.
• To assess the performance of a machine learning classifier for AF identification.

Main Methods:

• A wearable ECG patch transmitted data via Bluetooth to an Android smartphone.
• An Android application displayed ECG waveforms and sent data to a cloud server.
• A CatBoost machine learning model was employed for AF classification on the cloud server.
• Physician feedback was integrated into the system via a web browser interface.

Main Results:

• The CatBoost classifier achieved an F1 score of 0.92 on a test set of 7,270 samples.
• The system demonstrated real-time ECG waveform display and data transmission.
• AF detection results and doctor's diagnoses were successfully relayed back to the user.

Conclusions:

• The proposed wearable ECG system effectively detects atrial fibrillation using smartphone and cloud-based machine learning.
• The system facilitates remote, long-term ECG telemonitoring for AF.
• The technology holds promise for improving AF management and patient care.