# Understanding Sources of Demographic Predictability in Brain MRI via Disentangling Anatomy and Contrast

arXiv: https://arxiv.org/abs/2603.04113 · HTML: https://mlsanity.com/paper/2603.04113/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 27 references that are in the map.
- First posted on arXiv 4 Mar 2026.

## Builds on (most cited first, 14 of 14)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining](https://mlsanity.com/paper/2107.09559/) (2021)
- [Reading Race: AI Recognises Patient's Racial Identity In Medical Images](https://mlsanity.com/paper/2107.10356/) (2021)
- [SynthStrip: Skull-Stripping for Any Brain Image](https://mlsanity.com/paper/2203.09974/) (2022)
- [Fairness in Cardiac MR Image Analysis: An Investigation of Bias Due to Data Imbalance in Deep Learning Based Segmentation](https://mlsanity.com/paper/2106.12387/) (2021)
- [Diagnosing failures of fairness transfer across distribution shift in real-world medical settings](https://mlsanity.com/paper/2202.01034/) (2022)
- [Feature robustness and sex differences in medical imaging: a case study in MRI-based Alzheimer's disease detection](https://mlsanity.com/paper/2204.01737/) (2022)
- [3D Convolutional Neural Networks for Stalled Brain Capillary Detection](https://mlsanity.com/paper/2104.01687/) (2021)
- [MR-CLIP: Efficient Metadata-Guided Learning of MRI Contrast Representations](https://mlsanity.com/paper/2507.00043/) (2025)
- [Investigating Demographic Bias in Brain MRI Segmentation: A Comparative Study of Deep-Learning and Non-Deep-Learning Methods](https://mlsanity.com/paper/2510.17999/) (2025)
- [Metadata-Aligned 3D MRI Representations for Contrast Understanding and Quality Control](https://mlsanity.com/paper/2511.00681/) (2025)
- [Invisible Attributes, Visible Biases: Exploring Demographic Shortcuts in MRI-based Alzheimer's Disease Classification](https://mlsanity.com/paper/2509.09558/) (2025)
- [DIST-CLIP: Arbitrary Metadata and Image Guided MRI Harmonization via Disentangled Anatomy-Contrast Representations](https://mlsanity.com/paper/2512.07674/) (2025)
- [Understanding-informed Bias Mitigation for Fair CMR Segmentation](https://mlsanity.com/paper/2503.17089/) (2025)

## Built on by (most cited first, 1 of 1)

- [Metadata Supervised Imaging Representations for Modelling and Controlling Acquisition Variability](https://mlsanity.com/paper/2607.11295/) (2026)

## Abstract (arXiv)

Demographic attributes can be predicted from medical images, raising concerns about bias in clinical AI systems. In X-ray imaging, acquisition characteristics have been shown to contribute substantially to this predictability. Whether the same holds in brain MRI remains unclear, as anatomical variation and acquisition-dependent contrast are deeply entangled in the image formation process, obscuring the origins of demographic signal. To address this, we propose a controlled framework based on disentangled representation learning, decomposing brain MRI into anatomy-focused representations that suppress acquisition influence and contrast embeddings that capture acquisition-dependent characteristics. Training predictive models for age, sex, and race on full images, anatomical representations, and contrast embeddings allows us to quantify the relative contributions of structure and acquisition to the demographic signal. Across three datasets and multiple MRI sequences, demographic predictability is found to be driven primarily by anatomical variation, with anatomy-focused representations largely preserving the performance of models trained on raw images. Contrast embeddings retain a weaker signal that is dataset-specific and does not generalize across sites. These findings suggest that effective mitigation must explicitly account for the primarily anatomical and secondarily acquisition-dependent origins of demographic signal, ensuring that any bias reduction generalizes robustly across domains.