IoT Based System for Heart Monitoring and Arrhythmia Detection Using Machine Learning.

Area of Science:

• Biomedical Engineering
• Computer Science
• Health Informatics

Background:

• Internet of Things (IoT) enables remote data access and analysis for applications like remote patient monitoring.
• Electrocardiogram (ECG) monitoring is crucial for diagnosing heart conditions, but remote monitoring presents challenges.

Purpose of the Study:

• To propose and evaluate an IoT system for real-time ECG monitoring and arrhythmia detection.
• To assess the performance of different machine learning algorithms for classifying heart events.

Main Methods:

• Developed an IoT system integrating a Polar H10 heart sensor, machine learning models, and communication technology.
• Trained and evaluated Random Forest, Convolutional Neural Network (CNN), and K-Nearest Neighbors (KNN) algorithms for heart event classification.
• Collected and analyzed ECG data to identify various types of arrhythmias and normal/unclassifiable beats.

Main Results:

• The K-Nearest Neighbors (KNN) algorithm achieved the highest accuracy in classifying arrhythmias: premature ventricular contraction (94%), fusion of ventricular beat (81%), and supraventricular premature beat (82%).
• KNN also demonstrated high accuracy in distinguishing normal beats (93%) and unclassifiable beats (97%).
• The proposed IoT architecture facilitates remote data sharing and storage for patient information.

Conclusions:

• The developed IoT system, particularly with the KNN classifier, offers a viable solution for accurate remote ECG monitoring and arrhythmia detection.
• This technology can enhance patient care by enabling timely alerts for cardiac events, even from remote locations.
• Further research can explore integration with clinical decision support systems for enhanced diagnostic capabilities.