Dataset-Specific Encoding of Image Content Underlies Cross-Dataset Failure in Brain MRI Foundation Models

arXiv:2610.13027 · cs.CV · Submitted 2026-10-09 · Read on arXiv

cs.CV

Submitted: 2026-10-09

Updated: 2026-10-09

Code: https://github.com/BrainFM/brainfm-profiler

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes

Sources

• A large-scale heterogeneous 3D magnetic resonance brain imaging dataset for self-supervised learning
• A Modality-agnostic Multi-task Foundation Model for Human Brain Imaging
• Copycats: the many lives of a publicly available medical imaging dataset
• Machine Learning with Multi-Site Imaging Data: An Empirical Study on the Impact of Scanner Effects
• Site Is Decodable Before Pretraining: Negative Controls for Probing Frozen Brain-MRI Foundation Models
• What EEG Foundation Models Encode: Dataset Identity and a Negative-Control Suite for Clinical Benchmarks
• Batch Effects In Brain Foundation Model Embeddings
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• Current Pathology Foundation Models are unrobust to Medical Center Differences
• Towards Robust Foundation Models for Digital Pathology
• Do 3D Medical Foundation Models See Through MRI Artifacts? A Controlled Study of Representation Robustness
• Do Histopathological Foundation Models Eliminate Batch Effects? A Comparative Study
• Do Medical Foundation Models Generalize on the African Brain?
• Metadata Supervised Imaging Representations for Modelling and Controlling Acquisition Variability
• Harmonization in Magnetic Resonance Imaging: A Survey of Acquisition, Image-level, and Feature-level Methods
• LEACE: Perfect linear concept erasure in closed form
• Representation Geometry as a Diagnostic for Out-of-Distribution Robustness
• Understanding intermediate layers using linear classifier probes
• The ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge 2023: Intracranial Meningioma
• Analysis of the BraTS 2023 Intracranial Meningioma Segmentation Challenge

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models