AI-Enable Heart Sound Analysis: PASCAL Approach for Precision-Driven Cardiopulmonary Assessment

A large number of medical professionals depended on manual methods to assess the features of heart sounds. Because of the hands-on nature of this method, it was necessary for them to generate waveforms representing heart sounds and then carefully analyses the various components of these sounds. In contrast, the findings of our most recent research provide an improved methodology that automates the process of separating heart sound data and extracting the relevant factors from those signals. This update is very helpful for applications of machine learning in thoracentesis, and it aligns well with the developments of the most recent clinical imaging technologies. A novel “accordion” approach is presented in our methodology, which we have dubbed PASCAL. This method not only makes use of reference data that is determined by sudden shifts, but it also incorporates an improved classification model that has been customised for the highest possible level of precision. This is particularly clear when looking at the Dataset that We Have for Characterising Heart Rate. It is of the utmost importance to emphasise the fact that the performance as well as the false positive ( \(F_{p}\) ) rates of our classification phase were thoroughly evaluated. While this was going on, we looked at how the features of the ECG signal changed over time. The findings of our investigation are really positive. A fantastic F1-score of 94.38% and a respectable accuracy rate of 93.52% were reached by our suggested technique for identifying and classifying the capabilities of a healthy cardiopulmonary system. The primary sound ( \(S_{1}\) duration), the subsequent sound ( \(S_{2}\) duration), the overall cardiac cycle, the duration of ventricular systole, the time taken for ventricular contraction, and the ratio between the systolic and diastolic phases were some of the aspects of heart sounds that we examined in great detail. This ground-breaking method makes it possible to conduct an accurate study of the two most important peaks of heart sounds, which are \(S_{1}\) and \(S_{1}\) . It is important to note that these peaks may demonstrate significant changes from one individual sample to the next. This is mostly because of the different positions that the sphygmomanometer is placed.