M. Pfeifer, G. Lenis, and O. Dössel. A general approach for dynamic modeling of physiological time series. In Biomedizinische Technik. Biomedical Engineering, vol. 58(s1) , 2013
Dynamic modeling of physiological time series represents an auspicious approach in the arena of biomedical signal processing. This study illustrates a new methodology for identifying dynamic models that is based on stationary stochastic processes. The method is applied to time series extracted from the ECG. Simulations of the gained models yield physiologically plausible results.
Student Theses (1)
M. Pfeifer. Investigation of the heart rates influence to the QT-RR dynamics and the morphology of the T-wave in healthy subjects. Institute of Biomedical Engineering, Karlsruhe Institute of Technology (KIT). Bachelorarbeit. 2012
Introduction: Ventricular repolarization (VR) analysis plays a crucial role in the in- vestigation of cardiac diseases and drug safety studies. Most of the methods have been primarily directed towards the duration of the ventricles de- and repolarization: the QT interval. In newer times, morphological repolarization properties, represented by the T wave, are focused more frequently. Within numerous factors affecting the VR, the heart rates (HRs) influence remains vague in nowadays literature. Hence, in this study the HRs influence to the VR is investigated.Methods: The study included data recorded by a surface electrocardiogram (ECG) from 10 male subjects during a cardiac stress test. Based on the QT-RR hysteresis loops, expo- nential and elliptical descriptors were extracted. In order to describe the QT-RR relation dynamically, a first order system identification was applied. Moreover, morphological T wave descriptors representing width, symmetrical properties and slimness were calculated.Results: The exponential and elliptical descriptors showed a qualitative dependence, which proved difficult to quantify. By low model residuals, the first order system approx- imation was an appropriate approach with increasing goodness for less dynamics. The T wave descriptors revealed a significant HR correlation in the upper HR range, whereas no deterministic influence was observed for HRs below 100bpm.Conclusions: Dynamic first order systems represent a more appropriate approach in or- der to describe the QT-RR relation compared to static methods. Since higher HRs are attained quite seldom during Holter ECGs, the morphology of the T wave can be assumed independently from the HR in the clinical context.