Pith. sign in

REVIEW 5 major objections 5 minor 51 references

Skull stripping with purely synthetic data

T0 review · 5 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read PUMBA shows that a model trained exclusively on synthetic images of random ellipsoids can strip brains from real MRI scans of humans, marmosets, macaques, mice, and rats with accuracy comparable to established methods.

desk verdict PUMBA's synthetic-ellipsoid training is a genuinely new idea that seems to work across species, but the 'no anatomical priors' claim is overstated and the paper needs numeric metrics. read the letter →

arxiv 2505.07159 v1 pith:2VKJ6PQC submitted 2025-05-12 eess.IV cs.CV

Jong Sung Park , Juhyung Ha , Siddhesh Thakur , Alexandra Badea , Spyridon Bakas , Eleftherios Garyfallidis This is my paper · ORCID
classification eess.IVcs.CV
keywords skullstrippingbrainextractionpurelysynthetictrainingdatamulti-speciessegmentationmultimodalMRIellipsoidshapepriorunsupervisedgeneralizable
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

PUMBA asks how little information a skull-stripping model actually needs, and its answer is: none from real images. The authors train a standard 3D encoder-decoder segmentation network (a U-Net) on 3,000 procedurally generated volumes, each built from a filled ellipsoid inside a hollow ellipsoid shell, painted with random Gaussian intensity patches, studded with small ellipsoid artifacts, and warped by a smooth deformation field. On real MRI covering healthy and tumor-bearing human brains, marmosets, macaques, mice, and rats, this synthetic-only model delivers accuracy comparable to established skull-stripping tools. That matters because collecting and annotating real training data is the main obstacle to generalizable medical segmentation, and because it suggests a recipe that may transfer to other segmentation tasks.

What carries the argument

The load-bearing mechanism is the synthetic image generator rather than the network. Each training volume starts with two nested ellipsoids of random axis lengths — an outer hollow ellipsoid representing the head shell and an inner ellipsoid representing the brain — plus small ellipsoids and holes that stand in for imaging artifacts or anomalies. Each ellipsoid is split into patches with independent Gaussian intensity statistics, so the model cannot memorize any tissue-specific intensity pattern; a smooth random deformation field then breaks the perfect ellipsoid geometry. The three-class label scheme, brain, brain boundary, and background, lets a post-processing step keep only the largest central mass, which is the brain in normal and most pathological cases. This generator carries the argument because it is the only place where anatomical assumptions enter the pipeline.

What would settle it

Benchmark the same synthetic-only training on a collection of brain MRIs with large tumors, severe edema, or marked atrophy and compare Dice to manual labels; if accuracy falls substantially below the baselines on those cases, the ellipsoid shape prior is the limiting factor.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the prior knowledge needed for brain extraction can be reduced to three facts: a mammalian brain is a connected, roughly ellipsoidal mass; it sits inside a surrounding shell with an identifiable intensity boundary; and its axial extent is shorter than its other axes. From these assumptions alone, the authors train a simple 3D U-Net on 3,000 synthetic volumes, each an inner ellipsoid inside a hollow outer ellipsoid with random Gaussian intensities and smooth deformations, and report Dice, Jaccard, and Hausdorff performance comparable to established methods on healthy adult human T1, multi-modal human, glioblastoma and low-grade glioma, marmoset, macaque, mouse, and rat scans. They emphasize that the metrics were achieved without any image priors, supervised or unsupervised, during training. A dedicated brain boundary label is used at inference to select the largest central connected component, which reliably identifies the brain among other regions.

Load-bearing premise

The load-bearing assumption is that every mammalian brain, healthy or pathological, is well approximated for learning purposes by a filled ellipsoid with a contrast boundary against surrounding tissue; if a real brain departs from that shape enough, the synthetic-only training has nothing to fall back on.

Editorial extensions

If this is right

  • Skull-stripping pipelines could be built for new species or scanners without any real training images or manual labels, as long as the target is roughly ellipsoidal and contrast-bounded.
  • A single synthetic-only model can replace separate human, non-human primate, and rodent brain-extraction tools, simplifying preprocessing pipelines.
  • Training with procedurally generated volumes sidesteps privacy and annotation cost entirely, since no patient data is used.
  • The boundary-label trick gives a general mechanism for selecting a central target object when the segmentation model lacks anatomical context.
  • If the shape prior holds for other organs, the same recipe may transfer to segmentation tasks beyond the brain.

Reading between the lines

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

  • The ellipsoid prior has an unquantified boundary: brains with large space-occupying lesions, severe midline shift, or atypical anatomy may fall outside the training distribution, and the paper does not test how far the shape model can bend.
  • The same generator could be adapted to other approximately convex organs by changing the outer shell and the contrast rule, but the largest-central-mass selection would need rethinking for organs that are not the dominant object in the field of view.
  • Because the paper reports training instability across fixed synthetic datasets, generating fresh synthetic volumes on every training step is a natural test of whether the approach's ceiling is higher than reported.
  • A quantitative sweep over ellipsoid axis ratios, deformation strength, and intensity contrast would reveal exactly which part of the synthetic distribution controls generalization to real scans.
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

5 major / 5 minor

Summary. The paper proposes PUMBA, a skull-stripping model trained exclusively on purely synthetic images consisting of nested ellipsoids with random intensities, deformations, and artifacts, with no real brain images or labels in the training set. The method is evaluated on human, marmoset, macaque, mouse, and rat MRI datasets, including tumor-bearing cases, and compared qualitatively and quantitatively with BET, Synthstrip, and Unet-studio. The central claim is that the model achieves comparable accuracy without real images or anatomical priors.

Significance. The idea of training a segmentation model on purely synthetic, non-anatomical primitives is original and, if rigorously supported, would be a meaningful step toward generalizable brain extraction. The paper provides publicly available code and evaluates on external public benchmarks, which are strengths. However, the headline claim that the method uses no anatomical priors is directly contradicted by the method description, and the results are presented qualitatively rather than through numeric metrics, making the central claims unverifiable as written. With corrected claims and a proper quantitative evaluation, the work could become a useful contribution to the field.

major comments (5)
  1. [Abstract and §4.1] The repeated claim that PUMBA works "without any real images or anatomical priors" is contradicted by §3.1, which introduces the prior that a mammalian brain "resembles a somewhat ellipsoid structure" and that the axial axis is shorter, and by §3.2, which builds every training image from nested ellipsoids and applies a post-processing rule that selects "the center most largest chunk." These are explicit anatomical/geometric priors. Please either rephrase the claim to say "without real images or species-specific anatomical labels" or, preferably, provide an ablation that removes the ellipsoid and central-chunk priors and shows that the model still performs adequately; otherwise the headline claim is not supported.
  2. [§4.1 and Fig. 3] The quantitative evaluation is presented only as plots: no numeric Dice, Jaccard, or Hausdorff values, standard deviations, or sample sizes are given in the text or tables. The statements that the method shows "comparable results" and "the highest accuracy in marmoset data" are therefore not verifiable. Please include a results table with per-dataset means and standard deviations, and ideally paired statistical comparisons against BET, Synthstrip, and Unet-studio.
  3. [§3 (Methods) and §5 (Discussion)] A grid search over image-generation parameters was used to select the final model, and §5 states that the intensity range had the largest effect on performance. The manuscript does not specify the data on which the grid search was evaluated. If real images were used to select the synthetic-data parameters, the claim of training without real data is compromised. Please state the validation protocol explicitly, and if real data informed parameter selection, re-run the evaluation with parameters fixed before seeing the test data.
  4. [§5 (Discussion)] The authors acknowledge "instability when training the same pipeline with different images generated by the same rules," but only a single trained model is reported, with no mention of random seeds, number of repeated runs, or variance of the resulting metrics. Because the training set is a fixed set of 3,000 synthetic samples (§3.3), the reported performance could be seed-dependent. Please report results across multiple training runs with different synthetic dataset draws to support the generalization claim.
  5. [§3.2] The post-processing step that selects the "center most largest chunk" is itself an anatomical prior about the brain being the central largest connected component. Its contribution to the reported performance should be quantified, for example by an ablation that removes the boundary label and the selection step, and it should be explicitly listed among the method's priors when describing the method's assumptions.
minor comments (5)
  1. [§3.1] The phrase "regarding of species" should read "regardless of species."
  2. [§3.1] The word "inifinite" is a typo and should be "infinite."
  3. [§5] The word "affect" in "have a noticeable affect" should be "effect."
  4. [§2.1] The tool name "Freeurfer" should be spelled "FreeSurfer."
  5. [§4.2] The dataset name "BORDEUX24" should likely be "BORDEAUX24" to match the NHP dataset reference.

Circularity Check

1 steps flagged · score 4.0 of 10

Benchmark evaluation is self-contained, but the headline 'no anatomical priors' claim is true only by redefining the ellipsoid shape prior out of 'prior.'

  1. self definitional [Abstract; Section 3.1 'Assumptions'; Section 4.1 'Quantitative results']
    "Our results show that even without any real images or anatomical priors, the model achieves comparable accuracy... It resembles a somewhat ellipsoid structure, and wider around the axial plane. By using these, and only these information, we can remove the assumptions on existence of any complex structure... this metric was achievable without any image priors, supervised or unsupervised, when training the model."

    The paper's own method section identifies the ellipsoid shape and the shorter axial axis as the 'prior assumptions'/'information' used to generate every synthetic training image (Sections 3.1 and 3.2). The abstract and Section 4.1 then report the result as achieved 'without any... anatomical priors' and 'without any image priors.' The no-prior conclusion therefore holds only under a definition of 'prior' that excludes the method's sole input, the ellipsoid shape prior. The headline claim 'no priors are needed' is contained in the premise 'an ellipsoid is not a prior,' rather than being established by an experiment that varies or ablates the shape prior.

full rationale

The quantitative evaluation is not circular: the model is trained on synthetic ellipsoid images with generated labels and tested on external real MRI datasets (IXI, MINDS, CAMRI, LPBA40, TCGA, NHP, mouse) with independent ground truth; no test labels or test-set statistics enter training, and the grid search reported in Section 5 is over synthetic image-generation parameters only. The self-citations (EVAC+ [30] and BrainMaGe [4]) are related-work references and are not load-bearing for the method or results. The one substantive circularity is in the framing: the claim of working 'without any anatomical priors' is contradicted by the paper's own Section 3.1, which explicitly relies on the prior that a mammalian brain is a connected, somewhat ellipsoid structure with a shorter axial axis, and Section 3.2, which builds every training image from nested ellipsoids and selects the largest central connected component at post-processing. That makes the 'no priors' headline true only by definitional fiat, not by experimental demonstration. Because the actual benchmark comparison is externally grounded, the circularity is limited to the prior-free characterization rather than the accuracy claims, giving a score of 4 rather than 6 or higher.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim rests on hand-designed synthetic generation parameters: ellipsoid shapes, intensity distributions, shell thickness, and dataset size. These are design choices rather than physical free parameters, and the evaluation is external to them. No new physical entities are introduced.

free parameters (6)
  • Ellipsoid intensity ranges (inner and outer) = mean 0.4-1.0, std 0.0-0.4
    Set by grid search; Section 5 notes the largest performance difference came from choosing these ranges so brain and head intensities are clearly above background.
  • Background intensity mean and std = mean 0.1, std 0.1
    Chosen to simulate noise in the synthetic images; not fitted to test data.
  • Small artifact ellipsoid intensity = mean 1.0, std 0.4
    Assigned random intensities in a smaller range to simulate artifacts or anomalies.
  • Number of intensity subregions per ellipsoid = 4
    Each ellipsoid is divided into four Gaussian regions to create internal intensity variation.
  • Training dataset size = 3000
    A fixed pre-generated dataset was used; the authors note that online generation would be better but was too time-consuming.
  • Ellipsoid shape ranges and shell thickness = Random ranges, exact values only in code
    Ellipsoid axes and hollow thickness are sampled randomly; the exact distributions are not fully specified in the text.
assumptions (4)
  • domain assumption A mammalian brain is a connected, roughly ellipsoidal structure with a shorter axial axis.
    Stated in Section 3.1 and used to generate the ellipsoid training data.
  • domain assumption Brain and non-brain tissues have separable intensity ranges with some identifiable contrast.
    Used to justify the random intensity ranges in Section 3.1 and Appendix A.1.
  • domain assumption The largest connected component near the center of the image corresponds to the brain, so a boundary label can guide post-processing.
    Used in Section 3.2 and the post-processing step shown in Figure 2.
  • ad hoc to paper Training on 3,000 fixed synthetic samples is sufficient to achieve generalizable performance.
    The authors acknowledge in Section 5 that training with different images generated by the same rules can be unstable, so this assumption is not fully validated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Skull stripping with purely synthetic data." pith.science (2026). https://pith.science/paper/2VKJ6PQC

@misc{pith2026250507159,
  author       = {Pith},
  title        = {Pith review of: Skull stripping with purely synthetic data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2VKJ6PQC}},
  note         = {Machine review of arXiv:2505.07159}
}
read the original abstract

While many skull stripping algorithms have been developed for multi-modal and multi-species cases, there is still a lack of a fundamentally generalizable approach. We present PUMBA(PUrely synthetic Multimodal/species invariant Brain extrAction), a strategy to train a model for brain extraction with no real brain images or labels. Our results show that even without any real images or anatomical priors, the model achieves comparable accuracy in multi-modal, multi-species and pathological cases. This work presents a new direction of research for any generalizable medical image segmentation task.

Figures

Figures reproduced from arXiv: 2505.07159 by the authors.

Figure 1
Figure 1. A visual example of how the synthetic training data is generated. The steps [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. The effect of the post processing step is shown on a T1 weighted image from the [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. The quantitative metrics are shown on the TCGA (top row), MINDS (middle [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The predicted mask of each method on multiple modalities of the TCGA (tumor) [PITH_FULL_IMAGE:figures/full_fig_p019_4.png]
Figure 5
Figure 5. Figure 5: The segmentation results on the MINDS (Marmoset) dataset and CAMRI (Ro [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]
Figure 6
Figure 6. Figure 6: The segmentation results on the privately gathered Mouse T1 and T2 images [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 44 canonical work pages

  1. [1]

    S. M. Smith, Fast robust automated brain extraction, Human brain mapping 17 (2002) 143–155

    work page 2002

  2. [2]

    Kirillov, E

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y. Lo, et al., Segment anything, in: Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 4015–4026

    2023

  3. [3]

    Isensee, M

    F. Isensee, M. Schell, I. Pflueger, G. Brugnara, D. Bonekamp, U. Neu- berger, A. Wick, H.-P. Schlemmer, S. Heiland, W. Wick, et al., Au- tomated brain extraction of multisequence mri using artificial neural networks, Human brain mapping 40 (2019) 4952–4964

    work page 2019

  4. [4]

    Thakur, J

    S. Thakur, J. Doshi, S. Pati, S. Rathore, C. Sako, M. Bilello, S. M. Ha, G. Shukla, A. Flanders, A. Kotrotsou, et al., Brain extraction on mri scans in presenceof diffuse glioma: Multi-institutionalperformance eval- uation of deep learning methods and robust modality-agnostic training, Neuroimage 220 (2020) 117081

    work page 2020

  5. [5]

    Mohapatra, A

    S. Mohapatra, A. Gosai, G. Schlaug, Sam vs bet: A comparative study for brain extraction and segmentation of magnetic resonance images using deep learning, arXiv preprint arXiv:2304.04738 (2023)

    arXiv 2023

  6. [6]

    S. Roy, T. Wald, G. Koehler, M. R. Rokuss, N. Disch, J. Holzschuh, D. Zimmerer, K. H. Maier-Hein, Sam. md: Zero-shot medical image segmentation capabilities of the segment anything model, arXiv preprint arXiv:2304.05396 (2023)

    arXiv 2023

  7. [7]

    J. Ma, Y. He, F. Li, L. Han, C. You, B. Wang, Segment anything in medical images, Nature Communications 15 (2024) 654

    2024

  8. [8]

    Zoetmulder, E

    R. Zoetmulder, E. Gavves, M. Caan, H. Marquering, Domain-and task- specific transfer learning for medical segmentation tasks, Computer Methods and Programs in Biomedicine 214 (2022) 106539. 13

    work page 2022

Show all 51 references
  1. [9]

    Thambawita, P

    V. Thambawita, P. Salehi, S. A. Sheshkal, S. A. Hicks, H. L. Hammer, S. Parasa, T. d. Lange, P. Halvorsen, M. A. Riegler, Singan-seg: Syn- thetic training data generation for medical image segmentation, PloS one 17 (2022) e0267976

    2022

  2. [10]

    Yeh, Brain mri segmentation using template-based training and visual perception augmentation, ArXiv (2023) arXiv–2308

    F.-C. Yeh, Brain mri segmentation using template-based training and visual perception augmentation, ArXiv (2023) arXiv–2308

    2023

  3. [11]

    Gopinath, A

    K. Gopinath, A. Hoopes, D. C. Alexander, S. E. Arnold, Y. Balbastre, B. Billot, A. Casamitjana, Y. Cheng, R. Y. Z. Chua, B. L. Edlow, et al., Synthetic data in generalizable, learning-based neuroimaging, Imaging Neuroscience 2 (2024) 1–22

    2024

  4. [12]

    URL:https: //brain-development.org/ixi-dataset/, [Online]

    brain-development.org team, IXI dataset, accessed 2022. URL:https: //brain-development.org/ixi-dataset/, [Online]

    2022

  5. [13]

    J. Hata, K. Nakae, H. Tsukada, A. Woodward, Y. Haga, M. Iida, A. Ue- matsu, F.Seki, N.Ichinohe, R.Gong, etal., Multi-modalbrainmagnetic resonance imaging database covering marmosets with a wide age range, Scientific Data 10 (2023) 221

    2023

  6. [14]

    camri mouse brain mri data

    L.-M. Hsu, W. Ban, T.-H. Chao, S. Song, D. H. Cerri, L. Walton, M. Broadwater, S.-H. Lee, Y. I. Shih, "camri mouse brain mri data",

  7. [15]

    camri rat brain mri data

    L.-M. Hsu, W. Ban, T.-H. Chao, S. Song, D. H. Cerri, L. Walton, M. Broadwater, S.-H. Lee, Y. I. Shih, "camri rat brain mri data", 2021. doi:10.18112/openneuro.ds002870.v1.0.1

    2021 doi

  8. [16]

    D. W. Shattuck, M. Mirza, V. Adisetiyo, C. Hojatkashani, G. Salamon, K. L. Narr, R. A. Poldrack, R. M. Bilder, A. W. Toga, Construction of a 3d probabilistic atlas of human cortical structures, Neuroimage 39 (2008) 1064–1080

    2008

  9. [17]

    Pedano, A

    N. Pedano, A. E. Flanders, L. Scarpace, T. Mikkelsen, J. M. Eschbacher, B. Hermes, V. Sisneros, J. Barnholtz-Sloan, Q. Ostrom, The cancer genome atlas low grade glioma collection (TCGA-LGG), 2016

    2016

  10. [18]

    Scarpace, T

    L. Scarpace, T. Mikkelsen, S. Cha, S. Rao, S. Tekchandani, D. Gutman, J. H. Saltz, B. J. Erickson, N. Pedano, A. E. Flanders, J. Barnholtz- Sloan, Q. Ostrom, D. Barboriak, L. J. Pierce, The cancer genome atlas glioblastoma multiforme collection (TCGA-GBM), 2016. 14

    2016

  11. [19]

    Jenkinson, C

    M. Jenkinson, C. F. Beckmann, T. E. Behrens, M. W. Woolrich, S. M. Smith, Fsl, Neuroimage 62 (2012) 782–790

    2012

  12. [20]

    M. S. Atkins, K. Siu, B. Law, J. J. Orchard, W. L. Rosenbaum, Dif- ficulties of t1 brain mri segmentation techniques, in: Medical Imaging 2002: Image Processing, volume 4684, SPIE, 2002, pp. 1837–1844

    2002

  13. [21]

    R. W. Cox, Afni: software for analysis and visualization of functional magnetic resonance neuroimages, Computers and Biomedical research 29 (1996) 162–173

    1996

  14. [22]

    Ségonne, A

    F. Ségonne, A. M. Dale, E. Busa, M. Glessner, D. Salat, H. K. Hahn, B. Fischl, A hybrid approach to the skull stripping problem in mri, Neuroimage 22 (2004) 1060–1075

    2004

  15. [23]

    J. E. Iglesias, C.-Y. Liu, P. M. Thompson, Z. Tu, Robust brain extrac- tion across datasets and comparison with publicly available methods, IEEE transactions on medical imaging 30 (2011) 1617–1634

    2011

  16. [24]

    D. W. Shattuck, R. M. Leahy, Brainsuite: an automated cortical surface identification tool, Medical image analysis 6 (2002) 129–142

    2002

  17. [25]

    S. F. Eskildsen, P. Coupé, V. Fonov, J. V. Manjón, K. K. Leung, N. Guizard, S. N. Wassef, L. R. Østergaard, D. L. Collins, A. D. N. Initiative, et al., Beast: brain extraction based on nonlocal segmenta- tion technique, NeuroImage 59 (2012) 2362–2373

    2012

  18. [26]

    Macnaught, S

    A.Serag, M.Blesa, E.J.Moore, R.Pataky, S.A.Sparrow, A.Wilkinson, G. Macnaught, S. I. Semple, J. P. Boardman, Accurate learning with few atlases (alfa): an algorithm for mri neonatal brain extraction and comparison with 11 publicly available methods, Scientific Reports 6 (2016) 1–15

    2016

  19. [27]

    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

    2021

  20. [28]

    J. Ma, F. Li, B. Wang, U-mamba: Enhancing long-range dependency for biomedical image segmentation, arXiv preprint arXiv:2401.04722 (2024). 15

    2024 arXiv

  21. [29]

    H. Gong, L. Kang, Y. Wang, Y. Wang, X. Wan, X. Wu, H. Li, nn- mamba: 3d biomedical image segmentation, classification and landmark detection with state space model, in: ISBI, 2025

    2025

  22. [30]

    J. S. Park, S. Fadnavis, E. Garyfallidis, Multi-scale v-net architec- ture with deep feature crf layers for brain extraction, Communications Medicine 4 (2024) 29

    2024

  23. [31]

    Arnab, S

    A. Arnab, S. Zheng, S. Jayasumana, B. Romera-Paredes, M. Larsson, A. Kirillov, B. Savchynskyy, C. Rother, F. Kahl, P. H. Torr, Condi- tional random fields meet deep neural networks for semantic segmen- tation: Combining probabilistic graphical models with deep learning for str...

    2018

  24. [32]

    K. Kim, A. Duman, E. Spezi, Rgu-net: Computationally efficient u-net for automated brain extraction of mpmri with presence of glioblastoma, in: Research Conference 2024, 2024, p. 28

    2024

  25. [33]

    K. Han, Y. Wang, Q. Tian, J. Guo, C. Xu, C. Xu, Ghostnet: More fea- tures from cheap operations, in: Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition, 2020, pp. 1580–1589

    2020

  26. [34]

    Moazami, D

    S. Moazami, D. Ray, D. Pelletier, A. A. Oberai, Probabilistic brain extraction in mr images via conditional generative adversarial networks, IEEE Transactions on Medical Imaging 43 (2023) 1071–1088

    2023

  27. [35]

    A. V. Dalca, J. Guttag, M. R. Sabuncu, Anatomical priors in convolu- tional networks for unsupervised biomedical segmentation, in: Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, 2018, pp. 9290–9299

    2018

  28. [36]

    Hoopes, J

    A. Hoopes, J. S. Mora, A. V. Dalca, B. Fischl, M. Hoffmann, Synthstrip: skull-stripping for any brain image, NeuroImage 260 (2022) 119474

    2022

  29. [37]

    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. 16

    2023

  30. [38]

    Y. Wang, J. Nie, P.-T. Yap, G. Li, F. Shi, X. Geng, L. Guo, D. Shen, A. D. N. Initiative, Knowledge-guided robust mri brain extraction for diverse large-scale neuroimaging studies on humans and non-human pri- mates, PloS one 9 (2014) e77810

    2014

  31. [39]

    Zhong, X

    T. Zhong, X. Wu, S. Liang, Z. Ning, L. Wang, Y. Niu, S. Yang, Z. Kang, Q. Feng, G. Li, et al., nbest: Deep-learning-based non-human primates brain extraction and segmentation toolbox across ages, sites and species, NeuroImage 295 (2024) 120652

    2024

  32. [40]

    Cerri, L

    L.-M.Hsu, S.Wang, P.Ranadive, W.Ban, T.-H.H.Chao, S.Song, D.H. Cerri, L. R. Walton, M. A. Broadwater, S.-H. Lee, et al., Automatic skull stripping of rat and mouse brain mri data using u-net, Frontiers in neuroscience 14 (2020) 568614

    2020

  33. [41]

    X.Wang, X.-H.Li, J.W.Cho, B.E.Russ, N.Rajamani, A.Omelchenko, L. Ai, A. Korchmaros, S. Sawiak, R. A. Benn, et al., U-net model for brain extraction: Trained on humans for transfer to non-human pri- mates, Neuroimage 235 (2021) 118001

    2021

  34. [42]

    Lohmeier, T

    J. Lohmeier, T. Kaneko, B. Hamm, M. R. Makowski, H. Okano, at- lasbrex: Automated template-derived brain extraction in animal mri, Scientific reports 9 (2019) 12219

    2019

  35. [43]

    M. P. Milham, L. Ai, B. Koo, T. Xu, C. Amiez, F. Balezeau, M. G. Baxter, E. L. Blezer, T. Brochier, A. Chen, et al., An open resource for non-human primate imaging, Neuron 100 (2018) 61–74

    2018

  36. [44]

    Zhang, Z

    Y. Zhang, Z. Shen, R. Jiao, Segment anything model for medical image segmentation: Current applications and future directions, Computers in Biology and Medicine (2024) 108238

    2024

  37. [45]

    T. H. Rafi, R. Mahjabin, E. Ghosh, Y.-W. Ko, J.-G. Lee, Domain gen- eralization for semantic segmentation: a survey, Artificial Intelligence Review 57 (2024) 247

    2024

  38. [46]

    Kascenas, N

    A. Kascenas, N. Pugeault, A. Q. O’Neil, Denoising autoencoders for un- supervised anomaly detection in brain mri, in: International Conference on Medical Imaging with Deep Learning, PMLR, 2022, pp. 653–664. 17

    2022

  39. [47]

    G. Luo, W. Xie, R. Gao, T. Zheng, L. Chen, H. Sun, Unsupervised anomaly detection in brain mri: Learning abstract distribution from massive healthy brains, Computers in biology and medicine 154 (2023) 106610

    2023

  40. [48]

    Omidi, A

    A. Omidi, A. Mohammadshahi, N. Gianchandani, R. King, L. Lei- jser, R. Souza, Unsupervised domain adaptation of mri skull-stripping trained on adult data to newborns, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024, pp. 7718– 7727

    2024

  41. [49]

    Ronneberger, P

    O. Ronneberger, P. Fischer, T. Brox, U-net: Convolutional networks for biomedical image segmentation, in: Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international con- ference, Munich, Germany, October 5-9, 2015, proceedings, part III 18, Spri...

    2015

  42. [50]

    Abadi, A

    M. Abadi, A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G. S. Corrado, A. Davis, J. Dean, M. Devin, et al., Tensorflow: Large-scale machine learning on heterogeneous systems, software available from ten- sorflow. org (2015), URL https://www. tensorflow. org (2015). 18 F...

    2015

  43. [2021]

    doi:10.18112/openneuro.ds002868.v1.0.1

    doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.