Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Systems and Biomedical Engineering

In this thesis, we presented and evaluated two new nonlinear analysis techniques to enhance the early detection and diagnosis of heart diseases especially the ones related to the heart valves by introducing new nonlinear feature extraction methods. The first method is based on the phase space analysis that we reconstructed from the heart sound signal. The density matrix of this phase space portrait is generated and further the nonlinear features are extracted. In the second method, we extracted the nonlinear features based on the texture analysis of the two dimensional phase space plot using the co-occurrence matrix. The results of the proposed nonlinear analysis techniques showed that they are promising candidates to be used in computer aided diagnosis (CAD) or monitoring tools for heart diseases. Where, the accuracy of the phase space density matrix method and the phase space co-occurrence matrix methods are 93.96% and 91.63 respectively

Issued also as CD