Extraction and analysis of fetal heart signals with abnormalities an Eigen-analysis based approach

The scope of this paper is to extract an accurate fetal heart signal and to further improve fetal heart signal abnormality detection capability by analyzing the extracted signal waveform. In order to obtain a more informative fetal heart signal, unwanted noisy signals were mitigated by using an Eigen filter based subspace separation technique, cascaded with an adaptive Wiener filter. In addition to fetal heart signal extraction this research was focused on the use of subspace tool for possible fetal heart signal abnormality detection. Therefore, to detect abnormalities in a fetal heart signal, an Eigenvector based subspace matching system was used. To evaluate the performance of abnormality detection capability in the proposed system, a Mitral Stenosis affected fetal heart signal was assumed. Thereby it was shown that subspace matching and analysis can become a vital tool in the future for abnormality detection in fetal heart signals. The proposed methodology though introduced to Mitral Stenosis has the capability to be extended to other abnormalities. The proposed method can be used as a guideline to warrant further investigation through other means, if a certain parameter exceeds certain boundary levels.