Shape-aware Segmentation of the Placenta in BOLD Fetal MRI Time Series

S. Mazdak Abulnaga1,20000-0001-7902-3640, Neel Dey30000-0003-1427-6406, Sean I. Young2,3, Eileen Pan1, Katherine I. Hobgood3, Clinton J. Wang1, P. Ellen Grant40000-0003-1005-4013, Esra Abaci Turk4, Polina Golland10000-0003-2516-731X
1: CSAIL/EECS, Massachusetts Institute of Technology, Cambridge, MA, USA, 2: MGH/HST Martinos Center for Biomedical Imaging, Harvard Medical School, Boston, MA, USA, 3: CSAIL, Massachusetts Institute of Technology, Cambridge, MA, USA, 4: Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Harvard Medical School, Boston, MA, USA
Publication date: 2023/12/08
PDF · Code · arXiv


Blood oxygen level dependent (BOLD) MRI time series with maternal hyperoxia can assess placental oxygenation and function. Measuring precise BOLD changes in the placenta requires accurate temporal placental segmentation and is confounded by fetal and maternal motion, contractions, and hyperoxia-induced intensity changes. Current BOLD placenta segmentation methods warp a manually annotated subject-specific template to the entire time series. However, as the placenta is a thin, elongated, and highly non-rigid organ subject to large deformations and obfuscated edges, existing work cannot accurately segment the placental shape, especially near boundaries. In this work, we propose a machine learning segmentation framework for placental BOLD MRI and apply it to segmenting each volume in a time series. We use a placental-boundary weighted loss formulation and perform a comprehensive evaluation across several popular segmentation objectives. Our model is trained and tested on a cohort of (91) subjects containing healthy fetuses, fetuses with fetal growth restriction, and mothers with high BMI. Biomedically, our model performs reliably in segmenting volumes in both normoxic and hyperoxic points in the BOLD time series. We further find that boundary-weighting increases placental segmentation performance by 8.3% and 6.0% Dice coefficient for the cross-entropy and Signed Distance Transform objectives, respectively.


Placenta · Segmentation · BOLD MRI · Shape

Bibtex @article{melba:2023:017:abulnaga, title = "Shape-aware Segmentation of the Placenta in BOLD Fetal MRI Time Series", author = "Abulnaga, S. Mazdak and Dey, Neel and Young, Sean I. and Pan, Eileen and Hobgood, Katherine I. and Wang, Clinton J. and Grant, P. Ellen and Abaci Turk, Esra and Golland, Polina", journal = "Machine Learning for Biomedical Imaging", volume = "2", issue = "PIPPI 2022 special issue", year = "2023", pages = "527--546", issn = "2766-905X", doi = "", url = "" }
RISTY - JOUR AU - Abulnaga, S. Mazdak AU - Dey, Neel AU - Young, Sean I. AU - Pan, Eileen AU - Hobgood, Katherine I. AU - Wang, Clinton J. AU - Grant, P. Ellen AU - Abaci Turk, Esra AU - Golland, Polina PY - 2023 TI - Shape-aware Segmentation of the Placenta in BOLD Fetal MRI Time Series T2 - Machine Learning for Biomedical Imaging VL - 2 IS - PIPPI 2022 special issue SP - 527 EP - 546 SN - 2766-905X DO - UR - ER -

2023:017 cover