Improved Classification of Phonocardiography Signal Using Optimised Feature Selection

Noise Reduction of Simplified Phonocardiograph Through Flexural Tube Adjustment

“…Phonocardiography is a graphic display that shows the sound of the heart, where the sound of the heart is monitored through the surface of the human skin and a tool for obtaining Phonocardiography data is called the Phonocardiograph (PCG) [1], [2]. Phonocardiography can be an invasive and noninvasive sensor where non-invasive sensing shows a graph of heart rate and vibration of the skin surface, which is the effect of changes in pressure from the blood vessel system [2].…”

“…A normal cardiac cycle contains two major audible sounds: the first heart sound (S1) or systole and the second heart sound (S2) or diastole named as Fundamental Heart Sounds (FHS), where the remaining two sounds S3 and S4 are generally not audible [1], [13]. The example of determination of S1 and S2 and charateristics of heart beat sound for normal patient and patient with murmur depicted in figure 1 below The heart sounds obtained from PCG signals of (a) normal patient; and (b) patient with murmur [4], [14].…”

Multi-Modal Feature Integration in Machine Learning Predictions for Cardiovascular Diseases

“…Results reveal that the suggested selection strategy boosts classifier efficiency by boosting accuracy, precision and sensitivity. 4 In a research study released in 2020, Devjyoti Chakraborty et al offer a novel method for collecting crucial information in PCG and then categorizing it into normal and abnormal groups utilizing deep learning techniques. After converting the signals to spectrograms, they used deep convolutional networks to extract information from the spectrogram.…”