Automatic Segmentation of Twin Regions in Mo-Di Placentae Based on Geometric Analysis of Spatiotemporal BOLD MRI Signals
Tal Shnitzer1, S. Mazdak Abulnaga1, Carolina Bibbo2, P. Ellen Grant3, Polina Golland1, Justin Solomon1, and Esra Abaci Turk3
1Computer Science and Artificial Intelligence Laboratory, Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, 2Maternal and Fetal Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, United States, 3Fetal-Neonatal Neuroiaging and Developmental Science Center, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States
We propose an automatic segmentation method for delineating functional regions of the placenta responsible for each twin in Mo-Di placentae. The study of differences in MRI biomarkers between identical twins promises to elucidate placental function and fetal development. We combine temporal information from BOLD MRI time series and spatial information from the umbilical cord insertion in a flattened placenta representation. We demonstrate alignment of the automatic segmentation results with expert manual delineations and subsequent agreement of dynamic MRI signals in the identified regions with those derived from expert segmentations. Our method enables automatic localized analysis of the placenta.
This abstract and the presentation materials are available to members only;
a login is required.