Prediction of fetal heart disease detection using support vector machine classifier and comparing with decision tree classifier.

The primary goal of this study is to predict fetal heart disease detection using Magnetic Resonance Imaging (MRI) scan images and comparing with decision tree (DT) classifiers to improve accuracy, specificity, and disease type. Materials and Methods: In this research 20 Magnetic Resonance Imaging (MRI) scan images in a view of samples of SVM (N=10) and DT (N=10). The sample size is calculated for each group with 80% of G power, 95% confidence interval, and 0.05 error rate (Alpha). Results: The proposed method SVM classifier achieves a high accuracy of 93%, specificity of 89% and DT classifier 82% accuracy, 79% specificity. The significance rate of accuracy is (p=0.001) and specificity is (p=0.018). Conclusion: The Novel Support Vector Machine classifiers have achieved significantly better accuracy, specificity when compared with the decision tree (DT) classifier to predict fetal heart disease. [ABSTRACT FROM AUTHOR]