QT interval variability

Under normal resting conditions, beat-to-beat changes in QT interval are very small with a standard deviation typically below 5 ms. Digital high resolution ECG sampled at 300 Hz or higher and dedicated computer algorithms are required for QTV assessment.[2] Template-based algorithms that use parts of, or the entire ECG waveform usually deliver good results;[3] template stretching or warping techniques[4] perform comparably well in the presence of noise.

A number of metrics have been proposed for QTV quantification. The QT variability index (QTVi) has been most frequently reported in the literature: ${\displaystyle QTVi=\log {\frac {SDQT^{2}/(QT_{mean})^{2}}{SDHR^{2}/(HR_{mean})^{2}}}}$, where ${\displaystyle SDQT}$, ${\displaystyle SDHR}$, ${\displaystyle QT_{mean}}$, and ${\displaystyle HR_{mean}}$ denote standard deviation and mean of QT interval and heart rate time series, respectively.[5]

More advanced approaches that take into account the relationship between QTV and heart rate variability include vector autoregressive process models[6] and information domain approaches.[7]

Example traces of heart rate and QT interval variability in a normal heart and after myocardial infarction.