Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Building a General SimCLR Self-Supervised Foundation Model Across Neurological Diseases to Advance 3D Brain MRI Diagnoses

T0 review · 2 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A SimCLR foundation model pre-trained on 45,000 3D brain MRIs outperforms supervised baselines on every tested diagnostic task, including out-of-distribution tasks and with only 20% of labels for Alzheimer's prediction.

desk verdict Useful released 3D brain MRI foundation model, but the 'outperforms all tasks' claim doesn't survive a look at the error bars on two of the four tasks. read the letter →

arxiv 2509.10620 v1 pith:4Y7TWIES submitted 2025-09-12 cs.CV cs.LG

Emily Kaczmarek , Justin Szeto , Brennan Nichyporuk , Tal Arbel This is my paper · ORCID
classification cs.CVcs.LG
keywords self-supervisedlearningSimCLR3DbrainMRIfoundationmodelAlzheimer'sdiseasestrokeoutcomecontrastivetransfer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that a single self-supervised SimCLR encoder, pre-trained at native 1 mm isotropic resolution on 44,958 T1-weighted brain MRI scans from 11 public datasets spanning healthy aging, Alzheimer's, Parkinson's, frontotemporal dementia, and stroke, learns representations general enough to beat fully supervised models on four downstream tasks: stroke-severity regression, Alzheimer's classification, sex classification, and age regression. The fine-tuned model wins on both in-distribution (SOOP stroke) and held-out out-of-distribution (AIBL, IXI) data. Crucially, for Alzheimer's classification, fine-tuning on only 20% of the training labels still beats supervised baselines trained on 100% of the labels. If true, this shows that diverse unlabeled brain MRI can substitute for task-specific labeled data, making high-accuracy diagnostic models feasible in low-label clinical settings.

What carries the argument

The load-bearing mechanism is the SimCLR contrastive objective (NT-Xent loss) applied to two random augmentations of each 3D volume, with a 3D ResNet-18 encoder and a 64-dimensional projection head. The augmentations—random spatial crop to 150×192×192, axial flip, 45° rotation, intensity shift, and contrast adjustment—force the encoder to become invariant to imaging nuisances while preserving anatomy. Pre-training uses one scan per patient per epoch across 18,759 patients, so frequent-scanner patients do not dominate. The same encoder is then either linearly probed or fine-tuned for each downstream task.

What would settle it

A concrete test: fine-tune the released SimCLR encoder on a held-out dataset from a different scanner vendor (e.g., a site not in the 11 used), with the same TurboPrep pipeline, and compare against a ResNet-18 trained from scratch on the same labeled data. If the SimCLR advantage vanishes or reverses on that unseen site, then the 'out-of-distribution' claim does not generalize beyond the tested sources. Also, an ablation that removes the intensity-shift and contrast augmentations could show whether anatomical invariance is the true driver.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that contrastive self-supervision (SimCLR) applied to a large, heterogeneous corpus of 3D brain MRIs yields a transferable encoder whose fine-tuned performance exceeds that of ResNet-18 and ViT-Tiny trained from scratch with full supervision across all four evaluated tasks (Table 2). The margin is largest for Alzheimer's classification (AUC 0.929 vs 0.869 for the next-best supervised model), and the model holds its advantage even when fine-tuned with only 20% of the labeled training data (Figure 4). The authors attribute the result to SimCLR's augmentation-invariant learning of anatomical features at high resolution (1 mm iso), which the ViT

Load-bearing premise

The paper assumes that after TurboPrep preprocessing (N4 bias correction, SynthStrip skull stripping, linear registration, intensity normalization), the 11 heterogeneous public datasets form a single learnable distribution, so the contrastively learned invariances reflect brain anatomy rather than scanner/site confounds, and that these features transfer to fully held-out datasets.

Editorial extensions

If this is right

  • A single pre-trained encoder can replace task-specific training for at least four distinct 3D brain MRI tasks, reducing the need for separate specialized models.
  • In settings with scarce labeled data, fine-tuning the foundation model beats from-scratch supervision; for Alzheimer's classification, 20% of labels suffice to outperform supervised baselines.
  • The CNN-based SimCLR approach is more effective than the ViT-based MAE approach at this dataset scale, suggesting inductive bias matters for 3D medical imaging with modest data.
  • The model's success on held-out datasets (AIBL, IXI) suggests the learned features are not overfit to the pre-training acquisition sources.
  • Release of code and model allows the community to fine-tune for other tasks without repeating the costly pre-training.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the pre-training corpus includes datasets with diverse acquisition protocols, the contrastive invariances may generalize to scanner/site shifts beyond the two held-out datasets; a systematic study across many sites could quantify this.
  • The Alzheimer's result with 10 positive training samples hints that even smaller label sets (e.g., a handful per class) could suffice for some classification tasks, but the upper bound of this regime remains untested.
  • The failure of MAE relative to SimCLR may be specific to the dataset scale and patch size; a larger corpus or different masking ratio could reverse the comparison.
  • Since sex classification reached near-perfect AUC (0.991), the model seems to encode fine anatomical sex dimorphisms; this could be a confound for disease-disparity studies and should be checked when fine-tuning on unrelated tasks.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. This paper presents a 3D SimCLR-based self-supervised foundation model for T1-weighted brain MRI. The encoder is a 3D ResNet-18 pretrained on 44,958 scans from 18,759 patients across 11 public datasets, with the SOOP stroke dataset split before pretraining. The authors compare fine-tuned and linear-probed SimCLR against a 3D MAE (ViT-Tiny) and supervised ResNet-18/ViT-T baselines on four downstream tasks: NIH stroke-scale regression (in-distribution), AD vs healthy classification (AIBL, out-of-distribution), sex classification and age regression (IXI, out-of-distribution). The central claim is that the fine-tuned SimCLR model outperforms all other models across all tasks, and retains this advantage when fine-tuned on only 20% of labeled data for AD classification. The model and code are publicly released.

Significance. The practical contribution is substantial if the empirical claim holds: a high-resolution, publicly released 3D MRI encoder pretrained on a diverse multi-disease corpus, combined with a reproducible TurboPrep preprocessing pipeline and held-out evaluation on AIBL and IXI. The five-run reporting in Table 2 and the use of held-out datasets are good practices. However, the headline 'outperforms all models across all tasks' is only partially supported statistically; for two of the four tasks the margin over the strongest baseline is within sampling noise. The scientific significance of the paper therefore depends on tightening the statistical analysis and calibrating the claims.

major comments (2)
  1. [Section 5.2, Figures 4 and 5] The claim 'outperforms all other models across all tasks' (Abstract and Section 5.1) is not backed by any significance test or confidence interval. Using the reported mean±std over five runs, the stroke-scale difference between SimCLR-FT (5.37±0.24) and ResNet-18 (5.47±0.09) is 0.10, with approximate standard error of the difference sqrt(0.24^2/5 + 0.09^2/5) ≈ 0.11, so the observed difference is within one standard error. For sex classification, SimCLR-FT (0.991±0.004) vs ResNet-18 (0.988±0.005) differs by 0.003 with SE ≈ 0.003, again marginal. Thus two of four tasks do not provide statistically distinguishable improvements. Please add paired tests or bootstrap confidence intervals on the five runs, with multiple-comparison awareness, and temper the universal superiority claim if these comparisons are not significant.
  2. [Section 5.2, Figures 4 and 5] The data-efficiency experiments are reported from a single run at each data fraction ('each model is trained once using the same seed'). The claim that at 20% data the model 'still outperforms all supervised models' for Alzheimer's disease is therefore a single-seed observation with no variance estimate. Since this is one of the paper's headline contributions, please run at least three (preferably five) seeds per fraction, report error bars, and compare against supervised baselines trained on the same reduced data fractions as well as on full data. If the result is not stable across seeds, the claim should be softened accordingly.
minor comments (5)
  1. [Section 2.1, 2.2, 4.1] Units are inconsistently formatted: '1.4 mm 3', '1×1×1 mm 3', and '10 4' should use superscripts (mm^3, 10^4).
  2. [Table 1] The 'NA' entries for AIBL and IXI scan counts are ambiguous; clarify that these datasets were only used for testing and their scans were not part of pretraining.
  3. [Section 5.1] In the in-distribution paragraph, 'superior performance over all other models' is stronger than the numbers support; please add a qualifier such as 'numerically' or 'not statistically significant' until formal tests are added.
  4. [Section 4.2] The MAE model-size comparison ('the performance of all three models are similar') lacks the criterion and the numbers. Please state whether this refers to pretraining loss, downstream performance, or something else, and report the actual values for ViT-Tiny/Base/Large.
  5. [Throughout] The manuscript has no limitations section. A brief discussion of statistical power, the near-ceiling sex-classification task, and the T1-only scope would help calibrate the claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper reports empirical evaluations on held-out splits and does not derive predictions from fitted inputs or self-citations.

full rationale

The paper's central claims are empirical: a SimCLR-based SSL model is pre-trained on 44,958 scans, fine-tuned on downstream training splits, and evaluated on held-out in-distribution (SOOP) and out-of-distribution (AIBL, IXI) test sets. The reported comparisons in Table 2 and the data-efficiency experiments in Figures 4-5 are direct measurements, not derivations from the model's own definitions. No parameter is fitted to a test set and then called a prediction; hyperparameters such as batch size, temperature, mask ratio, and augmentation strengths are taken from prior public implementations or are design choices, not fitted to the evaluation targets. The only near-issue is that SOOP's training split is used both for SSL pre-training and downstream fine-tuning, but the test and validation splits are held out from both phases; this is a standard in-distribution evaluation design, not a circular reduction. The paper invokes no uniqueness theorem, imports no load-bearing result from the authors' prior work, and contains no self-citations that carry the argument. Concerns about statistical significance or the strength of the universality claim are about evidential robustness, not circularity, and do not affect the circularity score.

Assumptions & free parameters 9 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new theoretical entities; its contribution is a trained model resource. The central empirical claim rests on standard SSL objectives, a set of hand-chosen hyperparameters, and several domain assumptions about MRI data harmonization and the transferability of contrastively learned features. The free parameters listed are standard defaults or design choices, not quantities fitted to the evaluation test sets.

free parameters (9)
  • SimCLR temperature tau = 0.5 (not reported; SimCLR default)
    Temperature in the NT-Xent loss; controls the sharpness of the contrastive distribution; inherited from the public SimCLR implementation, not tuned to downstream tasks.
  • SimCLR projection head dimension = 64
    Dimension of the projection head output; chosen by hand from the reference implementation; affects the learned embedding geometry.
  • SimCLR effective batch size = 72
    Contrastive methods are batch-size-sensitive; 72 was the largest feasible on 12 H100 GPUs; no batch-size sweep is reported.
  • Pre-training epochs (SimCLR and MAE) = 150
    All SSL models trained for 150 epochs; chosen by the authors, no early-stopping or epoch sweep reported.
  • MAE mask ratio = 75%
    Kept from the original MAE paper; determines the reconstruction difficulty.
  • MAE patch size = 25x16x16
    Chosen for 3D volumes; yields 864 patches; affects the tokenization and reconstruction task.
  • MAE ViT size = ViT-Tiny
    Selected after observing similar performance across ViT-Tiny/Base/Large; choice affects downstream results.
  • Downstream fine-tuning epochs = 100 for SSL, 300 for supervised
    The reported comparison uses different epoch budgets; a design choice that could affect the relative outcomes.
  • SimCLR augmentation suite = crop min 30x40x40, flip p=0.5, rotate 45 deg p=0.5, intensity shift 0.5 p=0.8, contrast gamma 0.5-1.5 p=0.8
    Strong augmentations central to the contrastive objective; chosen by hand with MONAI; if too strong, can destroy pathology-relevant signal.
assumptions (6)
  • domain assumption T1-weighted MRI is the appropriate common sequence across the 11 datasets for learning general brain representations
    Section 4.1 selects T1 because it is the most common sequence; other sequences may carry complementary disease information.
  • domain assumption The TurboPrep pipeline (N4 bias correction, SynthStrip skull stripping, linear registration, SynthSeg segmentation, intensity normalization) harmonizes multi-site scans sufficiently for shared representation learning
    Section 4.1 Preprocessing; cross-site scanner differences are assumed removed by registration and normalization.
  • standard math NT-Xent contrastive loss and MAE MSE reconstruction are appropriate objectives for learning clinically useful 3D brain representations
    Equations (1) and (2); both objectives are taken from prior literature and assumed to transfer to medical volumes.
  • domain assumption The diagnostic and demographic labels of the public datasets (AD, MCI, PD, stroke, healthy) are correct as provided by their sources
    Table 1 lists conditions per dataset; erroneous labels would propagate through evaluation.
  • domain assumption Linear registration to a template and Z-score normalization preserve the subtle pathological differences (e.g., ventricle size, lesions) relevant to the downstream tasks
    Section 4.1 Preprocessing; registration could wash out some volumetric disease signatures.
  • domain assumption Mean +/- std over five runs is a sufficient characterization of model variability
    Section 5.1 reports five independent runs; no significance tests are performed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Building a General SimCLR Self-Supervised Foundation Model Across Neurological Diseases to Advance 3D Brain MRI Diagnoses." pith.science (2026). https://pith.science/paper/4Y7TWIES

@misc{pith2026250910620,
  author       = {Pith},
  title        = {Pith review of: Building a General SimCLR Self-Supervised Foundation Model Across Neurological Diseases to Advance 3D Brain MRI Diagnoses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4Y7TWIES}},
  note         = {Machine review of arXiv:2509.10620}
}
read the original abstract

3D structural Magnetic Resonance Imaging (MRI) brain scans are commonly acquired in clinical settings to monitor a wide range of neurological conditions, including neurodegenerative disorders and stroke. While deep learning models have shown promising results analyzing 3D MRI across a number of brain imaging tasks, most are highly tailored for specific tasks with limited labeled data, and are not able to generalize across tasks and/or populations. The development of self-supervised learning (SSL) has enabled the creation of large medical foundation models that leverage diverse, unlabeled datasets ranging from healthy to diseased data, showing significant success in 2D medical imaging applications. However, even the very few foundation models for 3D brain MRI that have been developed remain limited in resolution, scope, or accessibility. In this work, we present a general, high-resolution SimCLR-based SSL foundation model for 3D brain structural MRI, pre-trained on 18,759 patients (44,958 scans) from 11 publicly available datasets spanning diverse neurological diseases. We compare our model to Masked Autoencoders (MAE), as well as two supervised baselines, on four diverse downstream prediction tasks in both in-distribution and out-of-distribution settings. Our fine-tuned SimCLR model outperforms all other models across all tasks. Notably, our model still achieves superior performance when fine-tuned using only 20% of labeled training samples for predicting Alzheimer's disease. We use publicly available code and data, and release our trained model at https://github.com/emilykaczmarek/3D-Neuro-SimCLR, contributing a broadly applicable and accessible foundation model for clinical brain MRI analysis.

Figures

Figures reproduced from arXiv: 2509.10620 by the authors.

Figure 1
Figure 1. Brain MRI slices of patients with Alzheimer’s dis [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. SimCLR Architecture. Two different views of a single [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Classification performance (AUC) of two downstream [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Performance (Mean Absolute Error) of age (blue, out [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. A continually expandable foundation model for brain MRI

    cs.CV 2026-08 conditional novelty 6.0 of 10

    Graph-Blueprint Pruning lets a 425,000-scan brain MRI foundation model expand across clinical domains with less forgetting than sequential tuning or EWC.

Reference graph

Works this paper leans on

40 extracted references · 2 linked inside Pith · cited by 1 Pith paper

  1. [1]

    The stroke outcome optimization project: Acute ischemic strokes from a comprehensive stroke center

    John Absher, Sarah Goncher, Roger Newman-Norlund, Nicholas Perkins, Grigori Yourganov, Jan Vargas, San- jeev Sivakumar, Naveen Parti, Shannon Sternberg, Alex Teghipco, et al. The stroke outcome optimization project: Acute ischemic strokes from a comprehensive stroke center. Scientific Data, 11(1):839, 2024. 5

    2024

  2. [2]

    Self-supervised learning from images with a joint-embedding predictive architecture.arXiv preprint arXiv:2301.08243, 2023

    Mahmoud Assran, Quentin Duval, Ishan Misra, Piotr Bo- janowski, Pascal Vincent, Michael Rabbat, Yann LeCun, and Nicolas Ballas. Self-supervised learning from images with a joint-embedding predictive architecture.arXiv preprint arXiv:2301.08243, 2023. 2

    arXiv 2023

  3. [3]

    Avants, C.L

    B.B. Avants, C.L. Epstein, M. Grossman, and J.C. Gee. Symmetric diffeomorphic image registration with cross- correlation: Evaluating automated labeling of elderly and neurodegenerative brain.Medical Image Analysis, 12(1):26– 41, 2008. Special Issue on The Third International Workshop on Biomedical Image Registration – WBIR 2006. 5

    2008

  4. [4]

    Big Self-Supervised Models Advance Medical Image Classifica- tion

    Shekoofeh Azizi, Basil Mustafa, Fiona Ryan, Zachary Beaver, Jan Freyberg, Jonathan Deaton, Aaron Loh, Alan Karthikesalingam, Simon Kornblith, Ting Chen, et al. Big Self-Supervised Models Advance Medical Image Classifica- tion. In2021 IEEE/CVF International Conference on Com- puter Vision (ICCV), pages 3458–3468. IEEE Computer So- ciety, 2021. 2

    2021

  5. [5]

    Greve, Oula Puonti, Axel Thielscher, Koen Van Leemput, Bruce Fischl, Adrian V

    Benjamin Billot, Douglas N. Greve, Oula Puonti, Axel Thielscher, Koen Van Leemput, Bruce Fischl, Adrian V . Dalca, and Juan Eugenio Iglesias. Synthseg: Segmentation of brain mri scans of any contrast and resolution without re- training.Medical Image Analysis, 86:102789, 2023. 5

    2023

  6. [6]

    Monai: An open-source frame- work for deep learning in healthcare.arXiv:2211.02701,

    M Jorge Cardoso, Wenqi Li, Richard Brown, Nic Ma, Eric Kerfoot, Yiheng Wang, et al. Monai: An open-source frame- work for deep learning in healthcare.arXiv:2211.02701,

    arXiv

  7. [7]

    Emerg- ing properties in self-supervised vision transformers

    Mathilde Caron, Hugo Touvron, Ishan Misra, Herv ´e J´egou, Julien Mairal, Piotr Bojanowski, and Armand Joulin. Emerg- ing properties in self-supervised vision transformers. InPro- ceedings of the International Conference on Computer Vi- sion (ICCV), 2021. 2

    2021

  8. [8]

    Tsaftaris

    Agisilaos Chartsias, Thomas Joyce, Mario Valerio Giuffrida, and Sotirios A. Tsaftaris. Multimodal mr synthesis via modality-invariant latent representation.IEEE Transactions on Medical Imaging, 37(3):803–814, 2018. 1

    2018

Show all 40 references
  1. [9]

    Medical image foundation models in assisting diagnosis of brain tumors: a pilot study.European Radiology, 34, 2024

    Mengyao Chen, Meng Zhang, Lijuan Yin, Lu Ma, Renxing Ding, Tao Zheng, Qiang Yue, Su Lui, and Huaiqiang Sun. Medical image foundation models in assisting diagnosis of brain tumors: a pilot study.European Radiology, 34, 2024. 2, 3, 7

    2024

  2. [10]

    A simple framework for contrastive learning of visual representations

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Ge- offrey Hinton. A simple framework for contrastive learning of visual representations. InProceedings of the 37th Inter- national Conference on Machine Learning, 2020. 2, 4

    2020

  3. [11]

    Cole, Rudra P.K

    James H. Cole, Rudra P.K. Poudel, Dimosthenis Tsagkra- soulis, Matthan W.A. Caan, Claire Steves, Tim D. Spector, and Giovanni Montana. Predicting brain age with deep learn- ing from raw imaging data results in a reliable and heritable biomarker.NeuroImage, 163:115–124, 2017. 1

    2017

  4. [12]

    Stolte, Yunchao Yang, Kang Liu, Kyle B

    Joseph Cox, Peng Liu, Skylar E. Stolte, Yunchao Yang, Kang Liu, Kyle B. See, Huiwen Ju, and Ruogu Fang. Brainseg- founder: Towards 3d foundation models for neuroimage seg- mentation.Medical Image Analysis, 97:103301, 2024. 2, 3, 7

    2024

  5. [13]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, et al. An image is worth 16x16 words: Trans- formers for image recognition at scale. InInternational ...

    2021

  6. [14]

    Con- trastive learning with continuous proxy meta-data for 3d mri classification

    Benoit Dufumier, Pietro Gori, Julie Victor, Antoine Grigis, Michele Wessa, Paolo Brambilla, Pauline Favre, Mircea Polosan, Colm McDonald, Camille Marie Piguet, et al. Con- trastive learning with continuous proxy meta-data for 3d mri classification. InMedical Image Computing an...

    2021

  7. [15]

    Kathryn A Ellis, Ashley I Bush, David Darby, Daniela De Fazio, Jonathan Foster, Peter Hudson, Nicola T Laut- enschlager, Nat Lenzo, Ralph N Martins, Paul Maruff, et al. The australian imaging, biomarkers and lifestyle (aibl) study of aging: methodology and baseline characteris...

    2009

  8. [16]

    Brain tumor seg- mentation with deep neural networks.Medical Image Anal- ysis, 35:18–31, 2017

    Mohammad Havaei, Axel Davy, David Warde-Farley, An- toine Biard, Aaron Courville, Yoshua Bengio, Chris Pal, Pierre-Marc Jodoin, and Hugo Larochelle. Brain tumor seg- mentation with deep neural networks.Medical Image Anal- ysis, 35:18–31, 2017. 1

    2017

  9. [17]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016. 6

    2016

  10. [18]

    Momentum contrast for unsupervised visual repre- sentation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual repre- sentation learning. In2020 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 9726– 9735, 2020. 2

    2020

  11. [19]

    Masked autoencoders are scal- able vision learners

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross Girshick. Masked autoencoders are scal- able vision learners. In2022 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 15979– 15988, 2022. 2, 4

    2022

  12. [20]

    Brain genomics superstruct project initial data release with structural, functional, and behavioral mea- sures.Scientific data, 2(1):1–16, 2015

    Avram J Holmes, Marisa O Hollinshead, Timothy M O’keefe, Victor I Petrov, Gabriele R Fariello, Lawrence L Wald, Bruce Fischl, Bruce R Rosen, Ross W Mair, Joshua L Roffman, et al. Brain genomics superstruct project initial data release with structural, functional, and behaviora...

    2015

  13. [21]

    Mora, Adrian V

    Andrew Hoopes, Jocelyn S. Mora, Adrian V . Dalca, Bruce Fischl, and Malte Hoffmann. Synthstrip: skull-stripping for any brain image.NeuroImage, 260:119474, 2022. 5

    2022

  14. [22]

    Information extraction from im- ages (ixi) dataset, 2025

    IXI Dataset Contributors. Information extraction from im- ages (ixi) dataset, 2025. 2, 5

    2025

  15. [23]

    An approach to building foundation models for brain image analysis

    Davood Karimi. An approach to building foundation models for brain image analysis. InMedical Image Computing and Computer Assisted Intervention – MICCAI 2024, pages 421– 431, 2024. 2, 3

    2024

  16. [24]

    Residual and plain convolutional neural networks for 3d brain mri classification

    Sergey Korolev, Amir Safiullin, Mikhail Belyaev, and Yulia Dodonova. Residual and plain convolutional neural networks for 3d brain mri classification. In2017 IEEE 14th Interna- tional Symposium on Biomedical Imaging (ISBI 2017), pages 835–838, 2017. 1

    2017

  17. [25]

    Segment anything in medical images.Nature Communications, 15:654, 2024

    Jun Ma, Yuting He, Feifei Li, Lin Han, Chenyu You, and Bo Wang. Segment anything in medical images.Nature Communications, 15:654, 2024. 2

    2024

  18. [26]

    The parkinson’s progres- sion markers initiative (ppmi)–establishing a pd biomarker cohort.Annals of clinical and translational neurology, 5(12): 1460–1477, 2018

    Kenneth Marek, Sohini Chowdhury, Andrew Siderowf, Shirley Lasch, Christopher S Coffey, Chelsea Caspell- Garcia, Tanya Simuni, Danna Jennings, Caroline M Tan- ner, John Q Trojanowski, et al. The parkinson’s progres- sion markers initiative (ppmi)–establishing a pd biomarker coh...

    2018

  19. [27]

    Sid E O’Bryant, Leigh A Johnson, Robert C Barber, Mered- ith N Braskie, Bradley Christian, James R Hall, Nalini Hazra, Kevin King, Deydeep Kothapalli, Stephanie Large, et al. The health & aging brain among latino elders (hable) study meth- ods and participant characteristics.A...

    2021

  20. [28]

    The dal- las lifespan brain study: A comprehensive adult lifespan data set of brain and cognitive aging.Scientific Data, 12(1):1–14,

    Denise C Park, Joseph P Hennessee, Evan T Smith, Mi- caela Y Chan, Xi Chen, Marianna Dakanali, Michelle E Far- rell, Peiying Liu, Hanzhang Lu, Neil Rofsky, et al. The dal- las lifespan brain study: A comprehensive adult lifespan data set of brain and cognitive aging.Scientific...

  21. [29]

    Castro, Anton Schwaighofer, Matthew P

    Fernando P ´erez-Garc´ıa, Harshita Sharma, Sam Bond-Taylor, Kenza Bouzid, Valentina Salvatelli, Maximilian Ilse, Shruthi Bannur, Daniel C. Castro, Anton Schwaighofer, Matthew P. Lungren, et al. Exploring scalable medical image encoders beyond text supervision.Nature Machine In...

    2025

  22. [30]

    Alzheimer’s disease neuroimaging initiative (adni) clinical characterization.Neurology, 74(3): 201–209, 2010

    Ronald Carl Petersen, Paul S Aisen, Laurel A Beckett, Michael C Donohue, Anthony Collins Gamst, Danielle J Harvey, CR Jack Jr, William J Jagust, Leslie M Shaw, Arthur W Toga, et al. Alzheimer’s disease neuroimaging initiative (adni) clinical characterization.Neurology, 74(3): ...

    2010

  23. [31]

    The mayo clinic study of aging: design and sampling, participation, baseline measures and sample characteristics

    Rosebud O Roberts, Yonas E Geda, David S Knopman, Ruth H Cha, V Shane Pankratz, Bradley F Boeve, Robert J Ivnik, Eric G Tangalos, Ronald C Petersen, and Walter A Rocca. The mayo clinic study of aging: design and sampling, participation, baseline measures and sample characteris...

    2008

  24. [32]

    Rosen and the FTLDNI Study Group

    Howard J. Rosen and the FTLDNI Study Group. Frontotem- poral lobar degeneration neuroimaging initiative (ftldni). Funded by the National Institute on Aging; coordinated by UCSF and collaborating sites in North America, 2010. Data used in this study were obtained from the FTLDN...

    2010

  25. [33]

    Shinohara, Elizabeth M

    Russell T. Shinohara, Elizabeth M. Sweeney, Jeff Goldsmith, Navid Shiee, Farrah J. Mateen, Peter A. Calabresi, Sam- son Jarso, Dzung L. Pham, Daniel S. Reich, and Ciprian M. Crainiceanu. Statistical normalization techniques for mag- netic resonance imaging.NeuroImage: Clinical...

  26. [34]

    The amsterdam open mri collection, a set of multimodal mri datasets for individual difference analyses

    Lukas Snoek, Maite M van der Miesen, Tinka Beemster- boer, Andries Van Der Leij, Annemarie Eigenhuis, and H Steven Scholte. The amsterdam open mri collection, a set of multimodal mri datasets for individual difference analyses. Scientific data, 8(1):85, 2021. 5

    2021

  27. [35]

    Revisiting unreasonable effectiveness of data in deep learning era

    Chen Sun, Abhinav Shrivastava, Saurabh Singh, and Abhi- nav Gupta. Revisiting unreasonable effectiveness of data in deep learning era. In2017 IEEE International Conference on Computer Vision (ICCV), pages 843–852, 2017. 3

    2017

  28. [36]

    Tustison, Brian B

    Nicholas J. Tustison, Brian B. Avants, Philip A. Cook, Yuanjie Zheng, Alexander Egan, Paul A. Yushkevich, and James C. Gee. N4itk: Improved n3 bias correction.IEEE Transactions on Medical Imaging, 29(6):1310–1320, 2010. 5

    2010

  29. [37]

    Structural and functional brain scans from the cross-sectional southwest university adult lifespan dataset

    Dongtao Wei, Kaixiang Zhuang, Lei Ai, Qunlin Chen, Wen- jing Yang, Wei Liu, Kangcheng Wang, Jiangzhou Sun, and Jiang Qiu. Structural and functional brain scans from the cross-sectional southwest university adult lifespan dataset. Scientific data, 5(1):1–10, 2018. 5

    2018

  30. [38]

    David A Wood, Matthew Townend, Emily Guilhem, Sina Kafiabadi, Ahmed Hammam, Yiran Wei, Ayisha Al Busaidi, Asif Mazumder, Peter Sasieni, Gareth J Barker, et al. Op- timising brain age estimation through transfer learning: A suite of pre-trained foundation models for improved pe...

    2024

  31. [39]

    Matthews, Chuyang Ye, and Wenjia Bai

    Xinru Zhang, Ni Ou, Berke Doga Basaran, Marco Visentin, Mengyun Qiao, Renyang Gu, Cheng Ouyang, Yaou Liu, Paul M. Matthews, Chuyang Ye, and Wenjia Bai. A foun- dation model for brain lesion segmentation with mixture of modality experts. InMedical Image Computing and Com- puter...

    2024

  32. [40]

    An open science resource for establishing reliability and reproducibility in functional connectomics.Scientific data, 1(1):1–13, 2014

    Xi-Nian Zuo, Jeffrey S Anderson, Pierre Bellec, Rasmus M Birn, Bharat B Biswal, Janusch Blautzik, John Breitner, Randy L Buckner, Vince D Calhoun, F Xavier Castellanos, et al. An open science resource for establishing reliability and reproducibility in functional connectomics....

    2014

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.