Interpretation of Heart Sound Signal through Automated Artifact-Free Segmentation

Purpose: Digital recording of heart sounds commonly known as Phonocardiogram (PCG)signal, is a convenient primary diagnostic tool for analyzing condition of heart. Phonocardiogramaids physicians to visualize the acoustic energies that results from mechanicalaspect of cardiac activity. PCG signal cycle segmentation is an essential processing step towardsheart sound signal analysis. Sound artifacts due to inappropriate placement of stethoscope,body movement, cough etc. makes segmentation difficult. Artifact-free segmented heartsound cycles are convenient for physicians to interpret and it is also useful for computerizedautomated classification of abnormality.Methods: We have developed a framework which selects good quality heart sound subsequenceswhich are artifact-free and reused the features involved in this processing in segmentation.In this work, we have used information contained in frequency subbands by decomposingthe signal using Discrete Wavelet Packet Transform (DWPT). The algorithm identifies theparts of the signal where artifacts are prominent and it also detects major events in heart soundcycles.Results: The algorithm shows good results when tested on normal and five commonly occurringpathological heart sound signals. An average accuracy of 93.71% is registered for artifactfreesubsequence selection process. The cycle segmentation algorithm gives an accuracy of98.36%, 98.18% and 93.97% respectively for three databases used in the experiment.Conclusions: The work provides a solution for artifact-free segmentation of heart sound cyclesto assist interpretation of heart sound by physicians in objective analysis through recordingin a computer. It is also useful for development of an automated decision support system onheart sound abnormality. MAIN KEY FINDINGS• Artifact free subsequence detection is preferable over attempt to reduce the effect of artifactsdue to overlap of information content.• DWPT is useful for detection of artifact contaminated subsequences due to its ability to providefor more detailed information of higher frequency components.• DWPT features of subsequence detection can be reused for automatic segmentation of heartsound.• The artifact-free segmented heart sound cycle detection system can work in real-time.