Performance Analysis of Classification of Cardiotocograms Using Support Vector Machine based Classifier

Abstract

Fetal heart rate (FHR) and uterine contractions (UC) are simultaneously recorded by Cardiotocography (CTG). The CTG, which is one of the most common diagnostic techniques used to evaluate maternal and etal well-being during pregnancy and before delivery. By observing the Cardiotocography trace patterns doctors can understand the state of the fetus. There are several signal processing and computer programming based techniques for interpreting a typical Cardiotocography data. Even few decades after the introduction of cardiotocography into clinical practice, the predictive capacity of these methods remains controversial and still inaccurate. In this research work I propose an integrated methodology for CTG analysis and classification. A novel set of features, derived from the time and frequency domains, is used to feed the new powerful tool for pattern classification, named Support Vector Machines (SVMs). We used Accuracy, Specificity, NPV, Precision, Recall and ROC as the metric to evaluate the performance. The arrived results prove that, even though the traditional clustering methods can identify the Normal CTG patterns, they were incapable of Suspicious and Pathologic patterns. It was found that, the Support Vector Machines based classifier was capable of identifying Normal, Suspicious and Pathologic condition, from the nature of CTG data with very good accuracy..