Analysis of the QT-RR variability interactions using the NARMAX model

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• The QT and RR intervals are crucial indicators of cardiac electrical activity and autonomic nervous system function.
• Understanding the complex interactions between QT and RR intervals is vital for diagnosing various heart pathologies.
• Existing methods may lack the precision to differentiate subtle variations in these interval dynamics.

Purpose of the Study:

• To develop and validate a new hybrid model, the non-linear autoregressive moving average with exogenous input (NARMAX), for quantifying QT-RR interval interactions.
• To assess the model's ability to differentiate between normal, long QT, and short QT interval cardiac conditions.
• To introduce and analyze the QT variability index (QTVI) as a measure of QT-RR interval relationship variability.

Main Methods:

• Utilized a two-step NARMAX model for identifying RR and QT series, involving linear parametric (MA) and non-linear (NARX) identification.
• Computed power spectral density (PSD) using the monovariate MA model and analyzed QT-related RR series via the bivariate NARX model.
• Employed cross-spectral and coherence functions for result confirmation and compared findings with the Poincaré plot method.

Main Results:

• The NARMAX model identified distinct low frequency (LF) and high frequency (HF) components in all analyzed cardiac cases.
• LF predominated in normal and long QT interval cases, while HF was significantly larger in short QT interval cases.
• The NARMAX model demonstrated high precision (p < 0.001) in distinguishing between normal and pathological states, outperforming the Poincaré plot method in this regard.

Conclusions:

• The NARMAX model provides a robust and precise method for evaluating QT-RR interval interactions and cardiac autonomic function.
• Frequency band analysis (LF/HF) using NARMAX effectively differentiates cardiac conditions, offering diagnostic potential.
• The QTVI quantifies variability, showing distinct patterns in short and long QT interval cases, further aiding in cardiac assessment.