Every year, approximately 15 million fetuses are born preterm (when a baby is born before 37 weeks). For instance, approximately 10% of fetuses were born preterm in USA in 2019 (CDC.GOV). Preterm birth is considered a leading factor of death among children under 5 years of age. Predicting if a fetus may be born earlier than he/she should, may allow clinicians to provide special care to such pregnancies and may hopefully reduce preterm birth. In this project, we study how to predict preterm birth from multiple fetal ultrasound acquisitions using machine learning.