Leveraging ML for Fetal Acquisition and Analysis

Speaker: Ellen Grant, MD, Director, Fetal-Neonatal Neuroimaging and Developmental Science Center; Borjan Gagoski, PhD, Faculty MR Physician, Fetal-Neonatal Neuroimaging and Developmental Science Center; Junshen Xu, PhD, Student, at MIT


Fetal magnetic resonance imaging is challenging to perform as the fetus continually moves during image acquisition. As a result, the technologist must know how to "chase the fetus" to get images orthogonal to the fetal brain and repeat these acquisitions until images without motion are obtained. In this talk we will discuss how ML approaches are being used to accelerate and automate fetal imaging acquisition. Fetal imaging can be a challenge, but fetal motion also provides insight into the neurological and musculoskeletal development of the fetus. We will also describe how we have used ML approaches to track and quantify fetal motion and the potential neuroscientific and clinical applications.