Phonocardiography nonlinear multiple measurements in discovering the abnormalities in the bioprosthetic heart valves
Hussam Mustafa, Harold H Szu, N. Kyriakopoulos, Mohammed Ameen · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2001
This paper describes the benefit for potentially rectifying the multiple nonlinear Phonocardiography measurements so that we can apply the linear independent component analysis (ICA) to separate blindly the sources of heart valve murmuring in order to find a noninvasive way to discover abnormalities in the Bioprosthetic heart valves. At the beginning the paper will discuss a new design of measurement of the PCG signal based on the (ICA). The second part of the paper will show a comparison between the classification using the classical Fourier transform as a source of features and using the ICA sources of the same features. The last part of this paper will discuss various ways to overcome the nonlinear square detect law, and the lack of multiple independent readings, which is essential in the implementation of the ICA algorithm.