Classification of cardiotocography signals using machine learning

Abstract

. Cardiotocography has been used to record and monitor fetal heartbeat and uterine contractions, both antepartum and intrapartum for several decades now, albeit not without considerable controversy. The International Federation of Obstetrics and Gynecology (FIGO) guidelines were the first set of universally accepted classification guidelines for CTG signals. During labor, changes in the CTG are useful as indicators of fetal conditions. This paper aims to utilize CTG signal parameters to classify fetuses into three fetal states: normal, suspect and pathological and into 10 morphological patterns.