Pith. sign in

REVIEW 2 major objections 2 minor 1 cited by

Self-supervised pretraining on large unlabeled brain MRI datasets produces models that generalize better to heterogeneous clinical data than supervised models trained directly on the target domain.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.3

2026-05-25 06:27 UTC pith:5CHYGABH

load-bearing objection FOMO25 delivers a clean multi-team benchmark on SSL pretraining for clinical brain MRI but the three tasks leave the broader clinic-deployment claim under-supported. the 2 major comments →

arxiv 2604.11679 v2 pith:5CHYGABH submitted 2026-04-13 cs.CV

Towards Brain MRI Foundation Models for the Clinic: Findings from the FOMO25 Challenge

Asbj{\o}rn Munk , Stefano Cerri , Vardan Nersesjan , Christian Hedeager Krag , Jakob Ambsdorf , Pablo Rocamora Garc\'ia , Julia Machnio , Peirong Liu
show 76 more authors
Suhyun Ahn Nasrin Akbari Yasmina Al Khalil Kimberly Amador Sina Amirrajab Tal Arbel Meritxell Bach Cuadra Ujjwal Baid Bhakti Baheti Jaume Banus Kamil Barbierik Christoph Brune Yansong Bu Baptiste Callard Yuhan Chen Cornelius Crijnen Corentin Dancette Peter Drotar Prasad Dutande Nils D. Forkert Saurabh Garg Jakub Gazda Matej Gazda Beno\^it G\'erin Partha Ghosh Weikang Gong Pedro M. Gordaliza Sam Hashemi Tobias Heimann Fucang Jia Jiexin Jiang Emily Kaczmarek Chris Kang Seung Kwan Kang Mohammad Khazaei Julien Khlaut Petros Koutsouvelis Jae Sung Lee Yuchong Li Mengye Lyu Mingchen Ma Anant Madabhushi Klaus H. Maier-Hein Pierre Manceron Andr\'es Mart\'inez Mora Moona Mazher Felix Meister Nataliia Molchanova Steven A. Niederer Leonard N\"urnberg Jinah Park Abdul Qayyum Jonas Richiardi Antoine Saporta Branislav Setlak Ning Shen Justin Szeto Constantin Ulrich Puru Vaish Vibujithan Vigneshwaran Leroy Volmer Zihao Wang Siqi Wei Anthony Winder Jelmer M. Wolterink Maxence Wynen Chang Yang Si Young Yie Mostafa Mehdipour Ghazi Akshay Pai Espen Jimenez Solem Sebastian N{\o}rgaard Llambias Mikael Boesen Michael Eriksen Benros Juan Eugenio Iglesias Mads Nielsen
This is my paper
classification cs.CV
keywords brain MRIself-supervised learningfoundation modelsdomain shiftclinical deploymentinfarct classificationmeningioma segmentationbrain age regression
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper organizes a challenge to test whether foundation models pretrained with self-supervised learning on unlabeled scans can handle the noise and domain shifts typical of real hospital brain MRI workflows. It supplies a 60,000-scan pretraining set and evaluates nineteen models on three clinical tasks—in infarct classification, meningioma segmentation, and brain age regression—using few-shot and out-of-domain test data. Results indicate that the best out-of-domain self-supervised models outperform in-domain supervised baselines, that different pretraining objectives suit different tasks, and that larger models or longer training do not reliably improve results. A sympathetic reader would care because clinical labels are expensive and data are messy; if the claim holds, hospitals could leverage their existing unlabeled archives to build deployable tools with far less annotation effort.

Core claim

Self-supervised pretraining improves generalization on clinical data under domain shift, with the strongest models trained out-of-domain surpassing supervised baselines trained in-domain. No single pretraining objective benefits all tasks: masked autoencoding favors segmentation while hybrid reconstruction-contrastive objectives favor classification. Strong performance was achieved by small pretrained models, and scaling model size or training duration did not yield reliable benefits across the evaluated tasks.

What carries the argument

The FOMO25 challenge protocol that compares self-supervised models pretrained on the FOMO60K unlabeled dataset against supervised baselines on out-of-domain clinical tasks covering infarct classification, meningioma segmentation, and brain age regression.

Load-bearing premise

The three chosen clinical tasks and their evaluation datasets serve as representative proxies for the heterogeneity and labeling costs encountered in actual hospital brain MRI deployment.

What would settle it

A new set of clinical brain MRI scans from a different hospital network on which in-domain supervised models trained from scratch achieve higher accuracy than the top FOMO25 self-supervised entries on the same three tasks.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • MAE-based pretraining should be prioritized when the downstream task is segmentation.
  • Hybrid reconstruction-contrastive objectives should be prioritized when the downstream task is classification.
  • Model scaling and extended pretraining can be deprioritized in favor of careful objective selection for brain MRI foundation models.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same pretraining strategy may transfer to other MRI contrasts or body regions if the unlabeled data distribution matches the target clinical workflow.
  • Deployment pipelines could start with a small pretrained model and add task-specific fine-tuning rather than training large models from scratch.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 2 minor

Summary. The manuscript reports findings from the FOMO25 MICCAI challenge, in which 19 foundation models submitted by 16 teams were evaluated via a standardized containerized pipeline on three clinical brain MRI tasks (infarct classification, meningioma segmentation, brain age regression) using the FOMO60K pretraining dataset. The central claims are that self-supervised pretraining improves out-of-domain generalization on clinical data, that the strongest out-of-domain SSL models surpass in-domain supervised baselines, that no single pretraining objective is optimal across tasks, and that scaling model size or training duration yields no reliable gains.

Significance. If the empirical comparisons hold, the work supplies a reproducible, multi-team benchmark demonstrating practical benefits of SSL for clinical brain MRI under domain shift and label scarcity. The containerized evaluation protocol and task-specific objective findings constitute concrete, falsifiable guidance for foundation model development.

major comments (2)
  1. [Results] Results section: the claim that out-of-domain SSL models surpass in-domain supervised baselines is presented without error bars, confidence intervals, or statistical significance tests on the performance differences, which is load-bearing for the central generalization claim given the known variability of medical imaging metrics.
  2. [Introduction/Evaluation] Introduction and Evaluation sections: no quantitative characterization of domain-shift severity (e.g., distribution distances between FOMO60K and the three clinical test sets, or in-domain vs. out-of-domain performance drops) is provided, leaving the representativeness of the chosen tasks for broader clinical deployment unverified and weakening the interpretation of the reported outperformance.
minor comments (2)
  1. [Abstract] Abstract: the description of the containerized pipeline would benefit from a brief statement of the exact metrics and few-shot protocols used for each task.
  2. [Tables/Figures] Tables/figures: performance tables should consistently report standard deviations across runs or folds to allow direct visual assessment of the surpassing claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive evaluation and recommendation for minor revision. The comments highlight important opportunities to strengthen the statistical presentation and domain-shift analysis. We address each point below.

read point-by-point responses

  1. Referee: [Results] Results section: the claim that out-of-domain SSL models surpass in-domain supervised baselines is presented without error bars, confidence intervals, or statistical significance tests on the performance differences, which is load-bearing for the central generalization claim given the known variability of medical imaging metrics.

    Authors: We agree that measures of variability and formal statistical tests are necessary to support the central claim. In the revised manuscript we will add error bars (standard deviation across repeated evaluations or bootstrap estimates) to all reported metrics and include paired statistical significance tests (Wilcoxon signed-rank or McNemar as appropriate) for the key comparisons between the strongest out-of-domain SSL models and the in-domain supervised baselines. revision: yes

  2. Referee: [Introduction/Evaluation] Introduction and Evaluation sections: no quantitative characterization of domain-shift severity (e.g., distribution distances between FOMO60K and the three clinical test sets, or in-domain vs. out-of-domain performance drops) is provided, leaving the representativeness of the chosen tasks for broader clinical deployment unverified and weakening the interpretation of the reported outperformance.

    Authors: We acknowledge the value of quantifying domain shift. Computing full high-dimensional distances (MMD, Wasserstein) on raw 3-D MRI volumes at the scale of FOMO60K is computationally prohibitive within the challenge setting. In the revision we will add quantitative proxies—differences in mean voxel intensity, intensity histogram overlap, and variance between FOMO60K and each clinical test set—together with explicit in-domain versus out-of-domain performance gaps for the supervised baselines where such data exist. These additions will improve interpretability without altering the core experimental design. revision: partial

Circularity Check

0 steps flagged

No circularity: empirical challenge results with external evaluation

full rationale

The paper reports outcomes from the FOMO25 challenge involving 19 models evaluated via standardized pipeline on clinical tasks (infarct classification, meningioma segmentation, brain age regression) using FOMO60K pretraining data. Central claims rest on direct empirical comparisons of SSL vs supervised performance under domain shift, with no equations, fitted parameters renamed as predictions, self-definitional constructs, or load-bearing self-citations. The derivation chain consists solely of observed performance metrics from independent teams and external test sets; no step reduces by construction to its inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

This is an empirical benchmark study relying on standard assumptions in medical imaging evaluation rather than new mathematical derivations or postulated entities.

axioms (1)
  • domain assumption The FOMO60K pretraining dataset and clinical test sets are representative of real clinical workflows and domain shifts.
    This assumption underpins the generalization claims and task evaluations described in the abstract.

pith-pipeline@v0.9.0 · 6236 in / 1214 out tokens · 38533 ms · 2026-05-25T06:27:12.262468+00:00 · methodology

0 comments
read the original abstract

Clinical deployment of automated brain MRI analysis faces a fundamental challenge: clinical data is heterogeneous and noisy, and high-quality labels are prohibitively costly to obtain. Self-supervised learning (SSL) can address this by leveraging the vast amounts of unlabeled data produced in clinical workflows to train robust \textit{foundation models} that adapt out-of-domain with minimal supervision. However, the development of foundation models for brain MRI has been limited by small pretraining datasets and in-domain benchmarking focused on high-quality, research-grade data. To address this gap, we organized the FOMO25 challenge as a satellite event at MICCAI 2025. FOMO25 provided participants with a large pretraining dataset, FOMO60K, and evaluated models on data sourced directly from clinical workflows in few-shot and out-of-domain settings. Tasks covered infarct classification, meningioma segmentation, and brain age regression, and considered both models trained on FOMO60K (method track) and any data (open track). Nineteen foundation models from sixteen teams were evaluated using a standardized containerized pipeline. Results show that (a) self-supervised pretraining improves generalization on clinical data under domain shift, with the strongest models trained \textit{out-of-domain} surpassing supervised baselines trained \textit{in-domain}. (b) No single pretraining objective benefits all tasks: MAE favors segmentation, hybrid reconstruction-contrastive objectives favor classification, and (c) strong performance was achieved by small pretrained models, and improvements from scaling model size and training duration did not yield reliable benefits.

Figures

Figures reproduced from arXiv: 2604.11679 by Abdul Qayyum, Akshay Pai, Anant Madabhushi, Andr\'es Mart\'inez Mora, Anthony Winder, Antoine Saporta, Asbj{\o}rn Munk, Baptiste Callard, Beno\^it G\'erin, Bhakti Baheti, Branislav Setlak, Chang Yang, Chris Kang, Christian Hedeager Krag, Christoph Brune, Constantin Ulrich, Corentin Dancette, Cornelius Crijnen, Emily Kaczmarek, Espen Jimenez Solem, Felix Meister, Fucang Jia, Jae Sung Lee, Jakob Ambsdorf, Jakub Gazda, Jaume Banus, Jelmer M. Wolterink, Jiexin Jiang, Jinah Park, Jonas Richiardi, Juan Eugenio Iglesias, Julia Machnio, Julien Khlaut, Justin Szeto, Kamil Barbierik, Kimberly Amador, Klaus H. Maier-Hein, Leonard N\"urnberg, Leroy Volmer, Mads Nielsen, Matej Gazda, Maxence Wynen, Mengye Lyu, Meritxell Bach Cuadra, Michael Eriksen Benros, Mikael Boesen, Mingchen Ma, Mohammad Khazaei, Moona Mazher, Mostafa Mehdipour Ghazi, Nasrin Akbari, Nataliia Molchanova, Nils D. Forkert, Ning Shen, Pablo Rocamora Garc\'ia, Partha Ghosh, Pedro M. Gordaliza, Peirong Liu, Peter Drotar, Petros Koutsouvelis, Pierre Manceron, Prasad Dutande, Puru Vaish, Sam Hashemi, Saurabh Garg, Sebastian N{\o}rgaard Llambias, Seung Kwan Kang, Sina Amirrajab, Siqi Wei, Si Young Yie, Stefano Cerri, Steven A. Niederer, Suhyun Ahn, Tal Arbel, Tobias Heimann, Ujjwal Baid, Vardan Nersesjan, Vibujithan Vigneshwaran, Weikang Gong, Yansong Bu, Yasmina Al Khalil, Yuchong Li, Yuhan Chen, Zihao Wang.

Figure 1
Figure 1. Figure 1: Self-supervised pretraining boosts generalization. Across tasks, the top pretraining-based model from the method-track outperforms both from-scratch out-of-domain and in-domain supervised baselines, demonstrating that SSL can effectively leverage heterogeneous MRI data. Baselines are nnU-Net (segmentation) and Asparagus (classification/regression). The best performing method for classification was ashash, … view at source ↗
Figure 2
Figure 2. Figure 2: Effect of pretraining choices on downstream performance. (A–C) Pairwise rank differences between tasks, grouped by SSL objective category (global, hybrid, local): classification vs. segmentation (A), segmentation vs. regression (B), and classification vs. regression (C). Teams with Dice or NSD < 0.01 were excluded from (A) and (B). Positive values indicate better relative performance on the task shown at t… view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of Method and Open Track Submissions. (A) Distribution of pretraining dataset sizes in the Open track (Method track teams all used FOMO60K). (B) Dimensionality of the input representation (2D vs. 3D). (C–F) Share of submissions within each track by SSL objective category (global, hybrid, local) (C), encoder size (D), backbone architecture (E), and number of available GPUs (F) three task-specific… view at source ↗
Figure 4
Figure 4. Figure 4: Effect of augmentation strategies on task performance. (A) Task-specific rank as a function of the total number of pretraining augmentations. (B) Task-specific rank as a function of the number of spatial augmentations only. Marker shapes denote track, and colors indicate overall rank within each track. Lower rank indicates better performance. Trend lines are shown for each task. cash prize of $2000, with $… view at source ↗
Figure 5
Figure 5. Figure 5: Impact of hyperparameter tuning on final rank. Categorization of the extent of hyperparameter tuning done for pretraining hyperparameters compared to the final rank on each task. between ranks over task types grouped by SSL objective type. We find that local objectives favor segmentation over classifi￾cation, with no notable difference between either segmentation and regression or classification and regres… view at source ↗
Figure 6
Figure 6. Figure 6: Rankings with supervised baseline models. (A, B) Bootstrap rank distributions (mean ± 95% CI) for the Method and Open tracks, with teams grouped into statistically defined performance tiers (colours). (C-E) Task-level performance for the best out-of-domain method across both tracks alongside in-domain and out-of-domain supervised baselines: (C) ROC curves for the infarct classification task, (D) Dice and N… view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Beyond Instance-Level Self-Supervision in 3D Multi-Modal Medical Imaging

    cs.CV 2026-05 unverdicted novelty 5.0

    A self-supervised approach uses consistent spatial relationships of anatomical structures across patients to improve 3D multi-modal medical image representations, yielding modest gains on segmentation and classificati...

Reference graph

Works this paper leans on

54 extracted references · 54 canonical work pages · cited by 1 Pith paper · 9 internal anchors

  1. [1]

    Isensee, P

    F. Isensee, P. F. Jaeger, S. A. Kohl, J. Petersen, K. H. Maier-Hein, nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation, Nature methods 18 (2021) 203–211

    work page 2021

  2. [2]

    Isensee, T

    F. Isensee, T. Wald, C. Ulrich, M. Baumgartner, S. Roy, K. Maier-Hein, P. F. Jaeger, nnu-net revisited: A call for rigorous validation in 3d medical image segmentation, in: International Conference on Medical Image Comput- ing and Computer-Assisted Intervention, Springer, 2024, pp. 488–498

    work page 2024

  3. [3]

    Mårtensson, D

    G. Mårtensson, D. Ferreira, T. Granberg, L. Cavallin, K. Oppedal, A. Padovani, I. Rektorova, L. Bonanni, M. Pardini, M. G. Kramberger, J.-P. Taylor, J. Hort, J. Snædal, J. Kulisevsky, F. Blanc, A. Antonini, P. Mecocci, B. Vellas, M. Tsolaki, I. Kłoszewska, H. Soininen, S. Lovestone, A. Simmons, D. Aarsland, E. Westman, The reliability of a deep learning m...

    work page doi:10.1016/j.media.2020.101714 2020

  4. [4]

    E. A. AlBadawy, A. Saha, M. A. Mazurowski, Deep learn- ing for segmentation of brain tumors: Impact of cross- institutional training and testing, Medical physics 45 (2018) 1150–1158

    work page 2018

  5. [5]

    Nørgaard Llambias, M

    S. Nørgaard Llambias, M. Nielsen, M. Mehdipour Ghazi, Data augmentation-based unsupervised domain adapta- tion in medical imaging, in: Scandinavian Conference on Image Analysis, Springer, 2025, pp. 177–186

    work page 2025

  6. [6]

    Smith-Bindman, D

    R. Smith-Bindman, D. L. Miglioretti, E. Johnson, C. Lee, H. S. Feigelson, M. Flynn, R. T. Greenlee, R. L. Kruger, M. C. Hornbrook, D. Roblin, et al., Use of diagnos- tic imaging studies and associated radiation exposure for patients enrolled in large integrated health care systems, 1996-2010, Jama 307 (2012) 2400–2409

    work page 1996

  7. [7]

    Smith-Bindman, M

    R. Smith-Bindman, M. L. Kwan, E. C. Marlow, M. K. Theis, W. Bolch, S. Y . Cheng, E. J. Bowles, J. R. Duncan, R. T. Greenlee, L. H. Kushi, et al., Trends in use of med- ical imaging in US health care systems and in Ontario, Canada, 2000-2016, Jama 322 (2019) 843–856. 13

    work page 2000

  8. [8]

    Devlin, M.-W

    J. Devlin, M.-W. Chang, K. Lee, K. Toutanova, Bert: Pre- training of deep bidirectional transformers for language understanding, in: Proceedings of the 2019 conference of the North American chapter of the association for com- putational linguistics: human language technologies, vol- ume 1 (long and short papers), 2019, pp. 4171–4186

    work page 2019

  9. [9]

    T. Chen, S. Kornblith, M. Norouzi, G. Hinton, A simple framework for contrastive learning of visual representa- tions, in: International conference on machine learning, PmLR, 2020, pp. 1597–1607

    work page 2020

  10. [10]

    DINOv2: Learning Robust Visual Features without Supervision

    M. Oquab, T. Darcet, T. Moutakanni, H. V o, M. Szafraniec, V . Khalidov, P. Fernandez, D. Haz- iza, F. Massa, A. El-Nouby, et al., DINOv2: Learning Robust Visual Features without Supervision, arXiv preprint arXiv:2304.07193 (2023)

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  11. [11]

    Y . Shi, I. Daunhawer, J. E. V ogt, P. Torr, A. Sanyal, How robust is unsupervised representation learning to distri- bution shift?, in: The Eleventh International Confer- ence on Learning Representations, 2023. URL:https: //openreview.net/forum?id=LiXDW7CF94J

    work page 2023

  12. [12]

    K. He, X. Chen, S. Xie, Y . Li, P. Dollár, R. Girshick, Masked autoencoders are scalable vision learners, in: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 16000–16009

    work page 2022

  13. [13]

    Brown, B

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Ka- plan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, et al., Language models are few-shot learn- ers, Advances in neural information processing systems 33 (2020) 1877–1901

    work page 2020

  14. [14]

    Y . Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, V . Stoyanov, Roberta: A robustly optimized bert pretraining approach, arXiv preprint arXiv:1907.11692 (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 1907

  15. [15]

    Y . Tang, D. Yang, W. Li, H. R. Roth, B. Landman, D. Xu, V . Nath, A. Hatamizadeh, Self-supervised pre-training of swin transformers for 3d medical image analysis, in: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 20730–20740

    work page 2022

  16. [16]

    Z. Chen, D. Agarwal, K. Aggarwal, W. Safta, M. M. Balan, K. Brown, Masked image modeling advances 3d medical image analysis, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2023, pp. 1970–1980

    work page 2023

  17. [17]

    L. Wu, J. Zhuang, H. Chen, V oco: A simple-yet-effective volume contrastive learning framework for 3d medical im- age analysis, in: Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition, 2024, pp. 22873–22882

    work page 2024

  18. [18]

    T. Wald, C. Ulrich, S. Lukyanenko, A. Goncharov, A. Paderno, M. Miller, L. Maerkisch, P. Jaeger, K. Maier- Hein, Revisiting mae pre-training for 3d medical im- age segmentation, in: Proceedings of the Computer Vi- sion and Pattern Recognition Conference, 2025, pp. 5186– 5196

    work page 2025

  19. [19]

    A large-scale heterogeneous 3D magnetic resonance brain imaging dataset for self-supervised learning

    S. Cerri, A. Munk, S. N. Llambias, J. Ambsdorf, J. Machnio, V . Nersesjan, C. Hedeager Krag, P. Liu, P. Rocamora García, M. Mehdipour Ghazi, M. Boesen, M. E. Benros, J. E. Iglesias, M. Nielsen, A large- scale heterogeneous 3D magnetic resonance brain imag- ing dataset for self-supervised learning, arXiv preprint arXiv:2506.14432 (2026). URL:https://arxiv....

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  20. [20]

    Ulrich, T

    C. Ulrich, T. Wald, Y . Kirchhoff, M. Knopp, R. Peret- zke, M. Fischer, P. Ghosh, F. Isensee, A. Hilbert, P. Naser, L. Wessel, M. Foltyn-Dumitru, G. Brug- nara, J. B. Fiebach, J. O. Neumann, L. König, P. V ollmuth, K. Maier-Hein, SSL3D, 2025.https:// ssl3d-challenge.dkfz.de/home

    work page 2025

  21. [21]

    T. Wald, C. Ulrich, J. Suprijadi, S. Ziegler, M. Nohel, R. Peretzke, G. Kohler, K. Maier-Hein, An OpenMind for 3D medical vision self-supervised learning, in: Pro- ceedings of the IEEE/CVF International Conference on Computer Vision, 2025, pp. 23839–23879

    work page 2025

  22. [22]

    J. Ma, Y . Zhou, B. Wang, S. Kim, Z. Marinov, C. Xing, F. Li, Y . He, W. Li, F. Isensee, M. Rokuss, L. Krämer, K. Maier-Hein, Y . Du, B. Zhao, H. Wang, J. He, Y . Qiao, M. Zhang, H. Zhang, G.-Z. Yang, Y . Gu, L. Lumetti, F. Bolelli, C. Grana, Y . Chen, A. Erturk, T. Kuestner, S. Gatidis, M. Ingrisch, R. Graf, H. Möller, J. Kirschke, Z. Lin, T. Tan, H. Qu,...

    work page 2025

  23. [23]

    J. E. Iglesias, B. Billot, Y . Balbastre, C. Magdamo, S. E. Arnold, S. Das, B. L. Edlow, D. C. Alexander, P. Gol- land, B. Fischl, SynthSR: A public AI tool to turn het- erogeneous clinical brain scans into high-resolution T1- weighted images for 3D morphometry, Science advances 9 (2023) eadd3607

    work page 2023

  24. [24]

    Cerri, V

    S. Cerri, V . Nersesjan, K. V . Klein, E. C. Cóppulo, S. N. Llambias, M. M. Ghazi, M. Nielsen, M. E. Benros, Cross- disorder comparison of brain structures among 4,836 indi- viduals with mental disorders and controls utilizing danish population-based clinical mri scans, Molecular Psychiatry (2026). 14

    work page 2026

  25. [25]

    A. Munk, J. Ambsdorf, S. Llambias, M. Nielsen, Amaes: Augmented masked autoencoder pretraining on public brain mri data for 3d-native segmentation, MICCAI Workshop on Advancing Data Solutions in Medical Imag- ing AI (ADSMI 2024), MICCAI 2024 (2024)

    work page 2024

  26. [26]

    S. N. Llambias, J. Machnio, A. Munk, J. Ambsdorf, M. Nielsen, M. M. Ghazi, Yucca: A deep learning framework for medical image analysis, arXiv preprint arXiv:2407.19888 (2024)

    work page arXiv 2024

  27. [27]

    Maier-Hein, A

    L. Maier-Hein, A. Reinke, P. Godau, M. D. Tizabi, F. Buettner, E. Christodoulou, B. Glocker, F. Isensee, J. Kleesiek, M. Kozubek, et al., Metrics reloaded: recom- mendations for image analysis validation, Nature methods 21 (2024) 195–212

    work page 2024

  28. [28]

    Phipson, G

    B. Phipson, G. K. Smyth, Permutation P-values should never be zero: calculating exact P-values when permutations are randomly drawn, Statistical appli- cations in genetics and molecular biology 9 (2010). URL:https://doi.org/10.2202/1544-6115.1585. doi:10.2202/1544-6115.1585

    work page doi:10.2202/1544-6115.1585 2010

  29. [29]

    Maier-Hein, M

    L. Maier-Hein, M. Eisenmann, A. Reinke, S. Onogur, M. Stankovic, P. Scholz, T. Arbel, H. Bogunovic, A. P. Bradley, A. Carass, et al., Why rankings of biomedical im- age analysis competitions should be interpreted with care, Nature communications 9 (2018) 5217

    work page 2018

  30. [30]

    DINOv3

    O. Siméoni, H. V . V o, M. Seitzer, F. Baldassarre, M. Oquab, C. Jose, V . Khalidov, M. Szafraniec, S. Yi, M. Ramamonjisoa, et al., Dinov3, arXiv preprint arXiv:2508.10104 (2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  31. [31]

    S. Woo, S. Debnath, R. Hu, X. Chen, Z. Liu, I. S. Kweon, S. Xie, Convnext v2: Co-designing and scaling con- vnets with masked autoencoders, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 16133–16142

    work page 2023

  32. [32]

    Caron, H

    M. Caron, H. Touvron, I. Misra, H. Jégou, J. Mairal, P. Bojanowski, A. Joulin, Emerging properties in self- supervised vision transformers, in: Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 9650–9660

    work page 2021

  33. [33]

    J. Zhou, C. Wei, H. Wang, W. Shen, C. Xie, A. Yuille, T. Kong, ibot: Image bert pre-training with online tok- enizer, arXiv preprint arXiv:2111.07832 (2021)

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  34. [34]

    E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, W. Chen, et al., LoRA: Low-Rank Adaptation of Large Language Models., Iclr 1 (2022) 3

    work page 2022

  35. [35]

    Billot, D

    B. Billot, D. N. Greve, O. Puonti, A. Thielscher, K. Van Leemput, B. Fischl, A. V . Dalca, J. E. Iglesias, et al., SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining, Medical image analysis 86 (2023) 102789

    work page 2023

  36. [36]

    Fischl, FreeSurfer, Neuroimage 62 (2012) 774–781

    B. Fischl, FreeSurfer, Neuroimage 62 (2012) 774–781

    work page 2012

  37. [37]

    LaBella, O

    D. LaBella, O. Khanna, S. McBurney-Lin, R. Mclean, P. Nedelec, A. S. Rashid, N. H. Tahon, T. Altes, U. Baid, R. Bhalerao, et al., A multi-institutional meningioma MRI dataset for automated multi-sequence image segmenta- tion, Scientific data 11 (2024) 496

    work page 2024

  38. [38]

    D. P. Kingma, M. Welling, Auto-Encoding Variational Bayes, arXiv preprint arXiv:1312.6114 (2013)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  39. [39]

    P. M. Gordaliza, J. Banus, B. Gérin, M. Wynen, N. Molchanova, J. Richiardi, M. B. Cuadra, From 100,000+images to winning the first brain mri foun- dation model challenges: Sharing lessons and models (2026). URL:https://arxiv.org/abs/2601.13166. arXiv:2601.13166

    work page arXiv 2026

  40. [40]

    Decoupled Weight Decay Regularization

    I. Loshchilov, F. Hutter, Decoupled Weight Decay Regu- larization, arXiv preprint arXiv:1711.05101 (2017)

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  41. [41]

    M. Beck, K. Pöppel, M. Spanring, A. Auer, O. Prud- nikova, M. Kopp, G. Klambauer, J. Brandstetter, S. Hochreiter, xLSTM: Extended Long Short-Term Mem- ory, Advances in Neural Information Processing Systems 37 (2024) 107547–107603

    work page 2024

  42. [42]

    D. P. Kingma, J. Ba, Adam: A Method for Stochastic Optimization, arXiv preprint arXiv:1412.6980 (2014)

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  43. [43]

    Hatamizadeh, V

    A. Hatamizadeh, V . Nath, Y . Tang, D. Yang, H. R. Roth, D. Xu, Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images, in: Inter- national MICCAI brainlesion workshop, Springer, 2021, pp. 272–284

    work page 2021

  44. [44]

    X. Chen, H. Fan, R. Girshick, K. He, Improved base- lines with momentum contrastive learning, arXiv preprint arXiv:2003.04297 (2020)

    work page internal anchor Pith review Pith/arXiv arXiv 2003

  45. [45]

    Y . He, V . Nath, D. Yang, Y . Tang, A. Myronenko, D. Xu, Swinunetr-v2: Stronger swin transformers with stagewise convolutions for 3d medical image segmentation, in: In- ternational Conference on Medical Image Computing and Computer-Assisted Intervention, Springer, 2023, pp. 416– 426

    work page 2023

  46. [46]

    Z. Xie, Z. Zhang, Y . Cao, Y . Lin, J. Bao, Z. Yao, Q. Dai, H. Hu, SimMIM: A Simple Framework for Masked Im- age Modeling, in: Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition, 2022, pp. 9653–9663

    work page 2022

  47. [47]

    Vaish, F

    P. Vaish, F. Meister, T. Heimann, C. Brune, J. M. Wolterink, Consistent View Alignment Improves Founda- tion Models for 3D Medical Image Segmentation, arXiv preprint arXiv:2509.13846 (2025)

    work page arXiv 2025

  48. [48]

    La Rosa, J

    F. La Rosa, J. Dos Santos Silva, E. Dereskewicz, A. In- vernizzi, N. Cahan, J. Galasso, N. Garcia, R. Graney, S. Levy, G. Verma, et al., BrainAgeNeXt: advancing 15 brain age modeling for individuals with multiple sclero- sis, Imaging Neuroscience 3 (2025) imag_a_00487

    work page 2025

  49. [49]

    S. Roy, G. Koehler, C. Ulrich, M. Baumgartner, J. Pe- tersen, F. Isensee, P. F. Jaeger, K. H. Maier-Hein, Med- next: transformer-driven scaling of convnets for medical image segmentation, in: International conference on med- ical image computing and computer-assisted intervention, Springer, 2023, pp. 405–415

    work page 2023

  50. [50]

    arXiv preprint arXiv:2304.06716 (2023) 12, 17, 19, 26, 27, 28, 29, 30, 31 18 Q

    Z. Huang, H. Wang, Z. Deng, J. Ye, Y . Su, H. Sun, J. He, Y . Gu, L. Gu, S. Zhang, Y . Qiao, STU-Net: Scalable and Transferable Medical Image Segmentation Models Em- powered by Large-Scale Supervised Pre-training, arXiv preprint arXiv:2304.06716 (2023)

    work page arXiv 2023

  51. [51]

    Wasserthal, H.-C

    J. Wasserthal, H.-C. Breit, M. T. Meyer, M. Pradella, D. Hinck, A. W. Sauter, T. Heye, D. T. Boll, J. Cyriac, S. Yang, et al., TotalSegmentator: robust segmentation of 104 anatomic structures in CT images, Radiology: Artifi- cial Intelligence 5 (2023) e230024

    work page 2023

  52. [52]

    Curia: A multi- modal foundation model for radiology.arXiv preprint arXiv:2509.06830, 2025

    C. Dancette, J. Khlaut, A. Saporta, H. Philippe, E. Fer- reres, B. Callard, T. Danielou, L. Alberge, L. Machado, D. Tordjman, et al., Curia: A multi-modal foundation model for radiology, arXiv preprint arXiv:2509.06830 (2025)

    work page arXiv 2025

  53. [53]

    K. Li, Y . Wang, J. Zhang, P. Gao, G. Song, Y . Liu, H. Li, Y . Qiao, Uniformer: Unifying convolution and self-attention for visual recognition, IEEE Transactions on Pattern Analysis and Machine Intelligence 45 (2023) 12581–12600

    work page 2023

  54. [54]

    S. Rui, L. Chen, Z. Tang, L. Wang, M. Liu, S. Zhang, X. Wang, Multi-modal vision pre-training for medical im- age analysis, in: Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 5164–5174. Appendix A. Dataset Details This appendix details subject demographics, MRI sequences, preprocessing, and labeling protocols for all FOMO...

    work page 2025