Fetal Distress Classification

LSTM classifier to predict fetal distress based on fetal heart rate and uterine contraction time series data

Jan 2019 - March 2019

Fetal distress is a condition that occurs before and during childbirth when a fetus receives inadequate oxygen. We developed a system that classifies a fetus as “distressed” or “normal” using fetal heart rate and uterine contraction time series data obtained from cardiotocography. We used the CTU-CHB Intrapartum Cardiotocography Database which consists of 552 readings. We extracted features such as accelerations and decelerations and trained an LSTM classifier to predict fetal distress. We won first place in our category at the 2019 Smart India Hackathon.