Mining fetal magnetocardiogram data for high-risk fetuses

Abstract

The fetal magnetocardiogram (fMCG) contains a wealth of information regarding the health of a fetus. The purpose of this study is to classify fMCG data into the following two groups: high-risk and normal. In this presentation the authors first describe how the feature vector containing both time and frequency domain attributes is built from the time-series fMCG data. Second, the classification process using support vector machine (SVM) tools to identify the high-risk fetuses is described. Experimental results from 272 data sets taken from 118 fetuses demonstrate the SVM classifier's ability to distinguish between the high-risk and normal fetuses. Artificial neural networks and decision trees are used to validate the SVM results and receiver operating characteristic curve analysis and blind tests are employed to show the strength of the model. The model currently attains a sensitivity of 0.67 and a specificity of 0.65. While this study remains a work in progress, the authors are refining the process to improve the aforementioned results.