Preterm Birth Detection

This project investigates non-invasive preterm birth risk assessment using electrohysterography (EHG), which records uterine electrical activity from the maternal abdomen. The work is being carried out through a multidisciplinary collaboration involving the Department of Obstetrics & Gynaecology and the Department of Community Medicine of the Faculty of Medicine, together with the Department of Sociology at the University of Peradeniya. The research combines signal processing and machine learning to identify EHG patterns associated with preterm delivery. Initial studies use established public datasets to develop and rigorously evaluate prediction methods, while the next phase will establish a prospective Sri Lankan EHG dataset for local validation and model development. A dedicated higher-density EHG acquisition device has also been developed to support prospective data collection, richer spatial analysis, and future clinically relevant preterm birth risk assessment.