# The role of MRI physics in brain segmentation CNNs: achieving acquisition invariance and instructive uncertainties

arXiv: https://arxiv.org/abs/2111.02771 · HTML: https://mlsanity.com/paper/2111.02771/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 20 references that are in the map.
- First posted on arXiv 4 Nov 2021.

## Builds on (most cited first, 9 of 9)

- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](https://mlsanity.com/paper/1506.02142/) (2015)
- [What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?](https://mlsanity.com/paper/1703.04977/) (2017)
- [TorchIO: A Python library for efficient loading, preprocessing, augmentation and patch-based sampling of medical images in deep learning](https://mlsanity.com/paper/2003.04696/) (2020)
- [No New-Net](https://mlsanity.com/paper/1809.10483/) (2018)
- [A Learning Strategy for Contrast-agnostic MRI Segmentation](https://mlsanity.com/paper/2003.01995/) (2020)
- [Towards safe deep learning: accurately quantifying biomarker uncertainty in neural network predictions](https://mlsanity.com/paper/1806.08640/) (2018)
- [PSACNN: Pulse Sequence Adaptive Fast Whole Brain Segmentation](https://mlsanity.com/paper/1901.05992/) (2019)
- [Physics-informed brain MRI segmentation](https://mlsanity.com/paper/2001.10767/) (2020)
- [Contrast Adaptive Tissue Classification by Alternating Segmentation and Synthesis](https://mlsanity.com/paper/2103.02767/) (2021)

## Abstract (arXiv)

Being able to adequately process and combine data arising from different sites is crucial in neuroimaging, but is difficult, owing to site, sequence and acquisition-parameter dependent biases. It is important therefore to design algorithms that are not only robust to images of differing contrasts, but also be able to generalise well to unseen ones, with a quantifiable measure of uncertainty. In this paper we demonstrate the efficacy of a physics-informed, uncertainty-aware, segmentation network that employs augmentation-time MR simulations and homogeneous batch feature stratification to achieve acquisition invariance. We show that the proposed approach also accurately extrapolates to out-of-distribution sequence samples, providing well calibrated volumetric bounds on these. We demonstrate a significant improvement in terms of coefficients of variation, backed by uncertainty based volumetric validation.