Estimation of QT Interval Duration Using a Modernized Wavelet Transform based on Gauss-Hermite Functions

The relevance of developing new algorithms for diagnosing cardiovascular diseases is due to the fact that these diseases are the leading cause of death worldwide. One form of potential disease is an increase in the QT interval, which can be a harbinger of cardiac arrest or sudden death. Therefore, it is important not only to diagnose the presence of an elongated QT interval, but to timely detect a trend to this, thereby making a forecast about a person's condition in the future. The article describes the basic principles of constructing an algorithm for tracking the trend to an increase in the QT interval. The algorithm is based on the wavelet transform, where the Gauss- Hermite functions act as a basis. The wavelet transform is interpreted as correlation processing, and due to the ability to change the scale of the basis functions, it is possible to track the dynamics of changes in the duration of the QT interval. The result of such processing is a diagnostic function that is similar to a rhythmogram, where the ECG signal time is plotted along the abscissa axis, and the duration of the QT interval is plotted along the ordinate.