Velocity tracking of cardiac vector loops to identify signs of stress-induced ischaemia

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Coronary artery disease (CAD) is a major global cause of mortality.
• Current diagnostic methods, including electrocardiogram (ECG) stress tests and cardiac imaging, have limitations in diagnostic performance.
• There is a need for improved, non-invasive tools to assess CAD risk.

Purpose of the Study:

• To propose and evaluate novel quaternion methods for analyzing ECG signals during stress tests.
• To assess alterations in ventricular depolarization and repolarization patterns indicative of ischaemia.
• To develop an automated algorithm for calculating cardiac vector velocities during exercise and recovery.

Main Methods:

• Utilized the Bruce protocol treadmill stress test for data acquisition.
• Developed an algorithm to automatically detect exercise start and end points.
• Computed angular and linear velocities of the cardiac vector during depolarization and repolarization loops.
• Employed statistical analysis for feature selection and classification of ischaemic versus non-ischaemic patients.

Main Results:

• Maximum linear velocity during depolarization (p < 5E-9) and maximum angular velocity during repolarization (p < 5E-16) were significant markers for ischaemia.
• The angular velocity marker achieved a sensitivity/specificity of 78%/92% (AUC 0.89).
• A linear classifier identified a reduction in cardiac vector velocity post-exercise in at-risk patients, with 86% sensitivity and 100% specificity.

Conclusions:

• Quaternion analysis of cardiac vector dynamics during stress tests offers a sensitive and specific method for detecting CAD.
• Observed reductions in dynamic changes post-exercise may be linked to ischaemia-induced trajectory deviations in depolarization/repolarization loops.
• These findings suggest potential for non-invasive, complementary tools to improve CAD risk assessment.