Prediction of ECG-Biomarkers for Fetal Arrhythmia Using Non-invasive Fetal ECG

Abstract

A noninvasive Fetal Electrocardiogram (ECG) is supposed to be a potential prognostic tool in fetal arrhythmia identification and post-treatment. Fetal ECG data generated cannot be real-time accumulated, processed, and used for enterprise-level healthcare and wellness services with the existing fetal heart monitoring system used in hospitals. This study aims to quantify the ECG biomarkers and predict Fetal arrhythmia using Non-invasive fetal ECG data. We investigated the recordings of a total of 24 pregnant women using the Non-Invasive Fetal ECG Arrhythmia Database (NIFEA DB) (February 19, 2019) from physionet.org. We extracted ECG Fiducial Features and performed various statistical analyses on them to quantify ECG biomarkers. After performing statistical analysis we can conclude that Fetal arrhythmia ECG changes are associated with the ECG fiducial features. Machine-learning algorithms were investigated to predict fetal arrhythmia through Noninvasive Fetal ECG signals. The Overall accuracy of various Machine Learning Models is as followed: C5.0 is 95%, KNN is 94 %, CHAID is 90%, Neural Network is 81 %, and CART is 79 %. A noninvasive Fetal ECG-based Fetal Arrhythmia prediction approach is expected to utilize in a wearable fetus monitoring system.