T-wave morphology parameters based on principal component analysis reproducibility and dependence on T-offset...

Area of Science:

• Cardiovascular Physiology
• Electrocardiography
• Computational Biology

Background:

• T-wave morphology analysis using principal component analysis (PCA) aids in understanding QT prolongation and torsades de pointes.
• Assessing the repeatability and T-end position influence on PCA parameters is vital for their clinical application.

Purpose of the Study:

• To evaluate the repeatability of PCA-derived T-wave morphology parameters over various time scales.
• To determine the impact of T-end cursor positioning on PCA parameter accuracy.

Main Methods:

• Digital ECGs from 30 subjects were analyzed for short-term, circadian, and long-term repeatability of PCA parameters.
• The T-end cursor position was systematically varied (+/- 8 ms) to assess its influence.
• Key parameters calculated included QRS-T angle, PCA ratio, and T-wave residuum (TWR).

Main Results:

• Long-term repeatability showed coefficients of variation of 11.3% (QRS-T angle), 11.7% (PCA ratio), and 23.0% (TWR).
• Variations in T-end cursor position resulted in minimal impact: 0.42% (QRS-T angle), 1.00% (PCA ratio), and 4.0% (TWR).

Conclusions:

• PCA-based T-wave morphology parameters are generally reproducible, with TWR and QRS-T angle being exceptions.
• PCA demonstrates robustness and minimal sensitivity to T-end cursor position.
• Findings are significant for safety pharmacology trials involving ECG analysis.