A real-time cardiac arrhythmia classification system with wearable sensor networks.

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Continuous electrocardiogram (ECG) monitoring in free-living environments is crucial for preventing heart attacks and other cardiovascular diseases.
• Existing systems often face challenges with noise and unobtrusive data acquisition.
• Real-time analysis of cardiac arrhythmias requires efficient signal processing and classification algorithms.

Purpose of the Study:

• To design and develop a real-time wearable ECG monitoring system.
• To implement advanced algorithms for unobtrusive ECG and motion signal acquisition.
• To achieve real-time classification of multiple cardiac arrhythmias using a smartphone-integrated system.

Main Methods:

• Development of an ECG analog front-end and on-node digital processing to minimize noise and bias.
• Design of a wearable sensor node for unobtrusive monitoring of ECG and motion signals.
• Real-time ECG digitalization, Bluetooth transmission to a smartphone, and classification using a layered hidden Markov model.

Main Results:

• The wearable system successfully captured clean and reliable ECG waveforms even under stressed conditions.
• The integrated layered hidden Markov model demonstrated competent real-time classification of cardiac arrhythmias.
• The system enables unobtrusive, continuous monitoring and immediate analysis of cardiac health.

Conclusions:

• The developed real-time wearable ECG monitoring system offers a promising solution for continuous cardiac health assessment.
• The system's ability to acquire clean data and perform real-time arrhythmia classification enhances diagnostic capabilities.
• This technology has the potential to significantly improve the prevention and management of heart conditions.