Development of a Criterion for Selecting the Level of Wavelet Decomposition for QRS Detection in Electrocardiogram Signals Using Energy and Entropy

Analysis of electrocardiographic (ECG) signals is one of the key areas of modern biomedical engineering, providing the possibility of early detection of cardiac disorders such as arrhythmias, ischemia, and other cardiovascular pathologies. The most informative segment of the ECG signal is the QRS complex, which reflects the electrical activity of the heart ventricles and serves as an important marker for medical diagnostics. This work proposes a approach to selecting the optimal wavelet decomposition level based on quantitative analysis of energy and entropy characteristics of the signal at each level. The aim of the study is to construct a formalized criterion of informativeness for decomposition levels and evaluate the effectiveness of different types of mother wavelets for the task of extracting QRS complexes from ECG signals. Within the study, signals from the open MIT-BIH Arrhythmia Database were analyzed, and values of energy, entropy, and their ratio were calculated for each level of discrete wavelet transform. The results indicate that levels with high energy contribution and low entropy best reflect the localization of QRS complexes. It was shown that the choice of mother wavelet type significantly affects the distribution of these characteristics. The Daubechies wavelet was found to be the most effective for automated QRS detection, particularly at levels d3–d5. The scientific novelty of the work lies in the integration of the energy-entropy approach into the automated process of decomposition level selection without expert involvement. The practical significance is in the potential implementation of this method in computer diagnostic systems to improve their accuracy, reliability, and adaptability to various types of biomedical signals.