ECG-based heartbeat classification for arrhythmia detection: A survey

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Electrocardiogram (ECG) is a non-invasive tool for assessing heart electrical activity.
• Analyzing ECG waveforms aids in detecting cardiac abnormalities.
• Automated heartbeat classification methods have advanced significantly.

Purpose of the Study:

• To provide a comprehensive survey of state-of-the-art automated ECG-based heartbeat classification methods.
• To detail preprocessing, segmentation, feature extraction, and learning algorithms.
• To discuss evaluation databases and propose a future research workflow.

Main Methods:

• Literature review of automated ECG-based heartbeat classification techniques.
• Analysis of ECG signal preprocessing and heartbeat segmentation.
• Examination of feature description methods and machine learning algorithms.
• Discussion of AAMI-standardized databases (ANSI/AAMI EC57:1998/(R)2008).

Main Results:

• Identification of key components in automated ECG analysis: preprocessing, segmentation, feature extraction, and classification.
• Overview of common algorithms and evaluation methodologies.
• Highlighting of limitations and challenges in current literature.

Conclusions:

• Automated ECG analysis is crucial for cardiac diagnostics.
• Standardized evaluation using databases like AAMI is essential.
• Future work should address current limitations and refine evaluation processes.