# Distributionally Robust Segmentation of Abnormal Fetal Brain 3D MRI

arXiv: https://arxiv.org/abs/2108.04175 · HTML: https://mlsanity.com/paper/2108.04175/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 23 references that are in the map.
- First posted on arXiv 9 Aug 2021.

## Builds on (most cited first, 8 of 8)

- [Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging](https://mlsanity.com/paper/1909.12475/) (2019)
- [An automatic multi-tissue human fetal brain segmentation benchmark using the Fetal Tissue Annotation Dataset](https://mlsanity.com/paper/2010.15526/) (2020)
- [Generalised Wasserstein Dice Score for Imbalanced Multi-class Segmentation using Holistic Convolutional Networks](https://mlsanity.com/paper/1707.00478/) (2017)
- [Automatic brain tissue segmentation in fetal MRI using convolutional neural networks](https://mlsanity.com/paper/1906.04713/) (2019)
- [MONAIfbs: MONAI-based fetal brain MRI deep learning segmentation](https://mlsanity.com/paper/2103.13314/) (2021)
- [Label-set Loss Functions for Partial Supervision: Application to Fetal Brain 3D MRI Parcellation](https://mlsanity.com/paper/2107.03846/) (2021)
- [Distributionally Robust Deep Learning using Hardness Weighted Sampling](https://mlsanity.com/paper/2001.02658/) (2020)
- [Longitudinal analysis of fetal MRI in patients with prenatal spina bifida repair](https://mlsanity.com/paper/1911.06542/) (2019)

## Built on by (most cited first, 8 of 8)

- [A Dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation](https://mlsanity.com/paper/2204.02779/) (2022)
- [Learning to segment fetal brain tissue from noisy annotations](https://mlsanity.com/paper/2203.14962/) (2022)
- [Generalized Wasserstein Dice Loss, Test-time Augmentation, and Transformers for the BraTS 2021 challenge](https://mlsanity.com/paper/2112.13054/) (2021)
- [Partial supervision for the FeTA challenge 2021](https://mlsanity.com/paper/2111.02408/) (2021)
- [Adaptively Weighted Data Augmentation Consistency Regularization for Robust Optimization under Concept Shift](https://mlsanity.com/paper/2210.01891/) (2022)
- [Detailed delineation of the fetal brain in diffusion MRI via multi-task learning](https://mlsanity.com/paper/2409.06716/) (2024)
- [Anatomically Constrained Tractography of the Fetal Brain](https://mlsanity.com/paper/2403.02444/) (2024)
- [Enhancing Corpus Callosum Segmentation in Fetal MRI via Pathology-Informed Domain Randomization](https://mlsanity.com/paper/2508.20475/) (2025)

## Abstract (arXiv)

The performance of deep neural networks typically increases with the number of training images. However, not all images have the same importance towards improved performance and robustness. In fetal brain MRI, abnormalities exacerbate the variability of the developing brain anatomy compared to non-pathological cases. A small number of abnormal cases, as is typically available in clinical datasets used for training, are unlikely to fairly represent the rich variability of abnormal developing brains. This leads machine learning systems trained by maximizing the average performance to be biased toward non-pathological cases. This problem was recently referred to as hidden stratification. To be suited for clinical use, automatic segmentation methods need to reliably achieve high-quality segmentation outcomes also for pathological cases. In this paper, we show that the state-of-the-art deep learning pipeline nnU-Net has difficulties to generalize to unseen abnormal cases. To mitigate this problem, we propose to train a deep neural network to minimize a percentile of the distribution of per-volume loss over the dataset. We show that this can be achieved by using Distributionally Robust Optimization (DRO). DRO automatically reweights the training samples with lower performance, encouraging nnU-Net to perform more consistently on all cases. We validated our approach using a dataset of 368 fetal brain T2w MRIs, including 124 MRIs of open spina bifida cases and 51 MRIs of cases with other severe abnormalities of brain development.