A comprehensive review on efficient artificial intelligence models for classification of abnormal cardiac rhythms...

Area of Science:

• Biomedical Engineering
• Artificial Intelligence
• Cardiology

Background:

• Deep learning has advanced electrocardiogram (ECG) waveform analysis for cardiac rhythm classification.
• Current AI models for ECG analysis are often computationally intensive, hindering real-time application and increasing costs.

Purpose of the Study:

• To review state-of-the-art AI models for ECG-based cardiac rhythm classification.
• To explore emerging AI methodologies for efficient, real-time heart rhythm diagnosis.
• To assess strategies for reducing ECG lead requirements without sacrificing accuracy.

Main Methods:

• Review of current AI models for ECG analysis.
• Discussion of novel, lightweight AI methodologies for real-time diagnosis.
• Summary of studies on reduced ECG lead configurations for classification.

Main Results:

• Existing AI models for ECG analysis face challenges in real-time implementation due to computational demands.
• Emerging AI models show potential for lightweight, efficient, and accurate cardiac rhythm classification.
• Reducing the number of ECG leads may be feasible without compromising diagnostic accuracy.

Conclusions:

• Lightweight and computationally efficient AI models are crucial for real-time ECG-based cardiac diagnosis.
• Future AI applications in precision medicine can enhance cardiovascular status prediction and diagnostics.
• Optimizing AI models and data acquisition (e.g., fewer ECG leads) can improve accessibility and cost-effectiveness.