Machine learning algorithms in the healthcare domain can improve healthcare and clinical practice ethically and reliably. The fetal heart rate (FHR) and the uterine contraction (UC) activity are recorded by using the technique called Cardiotocography (CTG).It provides support for the obstetricians to obtain complete physiological information about new-borns.In this paper, machine learning classification algorithms such as Artificial Neural Network (ANN), Naive Bayes,Decision Tree, Random Forest, Support Vector Machine (SVM), and Adaptive Boosting (Adaboost) are applied to predict fetal status as normal, suspicious or pathologic. The performance of the algorithms has been evaluated based on training and testing Accuracy, Precision, Recall, Specificity, ROC (Receiver Operation Characteristics), and Kappa Statistics. The obtained result shows that the majority of classification algorithms perform better. It was found that Random Forest has provided the highest accuracy of 99% in training and 93% in testing.
Volume 12 | 08-Special Issue
Pages: 637-643
DOI: 10.5373/JARDCS/V12SP8/20202565