Detection and categorization of severe cardiac disorders based solely on heart period measurements

Area of Science:

• Cardiology
• Biomedical Engineering
• Data Science

Background:

• Cardiac disorders pose significant mortality risks.
• Current diagnostic methods like electrocardiograms (ECGs) are infrequent and require expert interpretation.
• Accessible, continuous cardiac monitoring is needed.

Purpose of the Study:

• To develop an automated system for early cardiac disorder detection using heart period measurements.
• To evaluate novel heart rate variability (HRV) metrics for improved diagnostic accuracy.
• To assess the system's performance using short-duration recordings.

Main Methods:

• Analysis of 1-day ECG recordings from over 1,000 subjects.
• Examination of conventional HRV metrics and two newly proposed metrics: local variation and local-global variation.
• Development of an automated detection system based on these metrics.

Main Results:

• A local variation metric showed higher efficiency in alerting to cardiac disorders compared to conventional HRV metrics.
• The novel local-global variation metric demonstrated superior discrimination between premature contractions and atrial fibrillation.
• The new system achieved better diagnostic performance with just a 1-minute recording than conventional methods using 1-day recordings.

Conclusions:

• The developed automated system offers a promising approach for accessible cardiac disorder detection.
• Novel HRV metrics, particularly local-global variation, enhance diagnostic capabilities for specific arrhythmias.
• Short-duration heart period recordings can be sufficient for effective cardiac monitoring using advanced analytical methods.