Performance Evaluation of QT-RR Adaptation Time Lag Estimation in Exercise Stress Testing.

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Slower QT interval adaptation to heart rate changes is a risk marker for ventricular arrhythmias.
• Exercise stress testing allows for estimating the QT-RR adaptation time lag.

Purpose of the Study:

• To develop and evaluate methods for estimating the QT-RR adaptation time lag.
• To assess the performance of different noise models (Gaussian, Laplacian) for time lag estimation.
• To investigate the impact of T-wave delineation accuracy and learning window positioning on estimation performance.

Main Methods:

• Developed maximum likelihood time lag estimators based on memoryless transformations of QT and RR intervals.
• Utilized an ECG simulator to model various physiological parameters and assess estimation accuracy.
• Investigated the influence of learning window placement on nonstationary data during exercise.

Main Results:

• The proposed estimation method accurately reflects the true time lag in simulated ECGs.
• Performance is reliable for heart rate trends below a specific frequency.
• Optimal learning window positioning reduces mean estimation error across various time lags.
• The Laplacian-based estimator demonstrated superior discrimination in a clinical dataset for coronary artery disease risk.

Conclusions:

• The developed method provides accurate estimation of the QT-RR adaptation time lag.
• The approach is robust for various heart rate trends and noise conditions.
• This technique can aid in identifying patients at risk for ventricular arrhythmias and coronary artery disease.