Pith. sign in

REVIEW 4 major objections 6 minor 58 references

Residual Vision Transformer (ResViT) Based Self-Supervised Learning Model for Brain Tumor Classification

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that learning to synthesize one brain MRI sequence from another, then fine-tuning the same model for classification, outperforms ImageNet pretraining on small brain tumor datasets.

desk verdict Plausible SSL pipeline whose reported gains are likely inflated by slice-level leakage between pretext pretraining and the classification test set. read the letter →

arxiv 2411.12874 v1 pith:B3244JF2 submitted 2024-11-19 eess.IV cs.CVcs.LG

Meryem Altin Karagoz , O. Ufuk Nalbantoglu , Geoffrey C. Fox This is my paper · ORCID
classification eess.IVcs.CVcs.LG
keywords self-supervisedlearningbraintumorclassificationMRIsynthesisvisiontransformerconvolutionalneuralnetworkResViTtransfermedicalimaging
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 self-supervised pretext task—learning to synthesize one brain MRI sequence from another—can teach a deep model features that transfer to brain tumor classification better than pretraining on ImageNet. The model, ResViT, combines residual CNN layers with vision transformer blocks, so it captures local and global image structure at once; after synthesis pretraining, the same encoder and transformer blocks are fine-tuned with a classification head, and synthetic MRIs are added to balance the training set. On public benchmarks the paper reports 90.56% accuracy on BraTS (T1 sequence), 98.53% on Figshare, and 98.47% on Kaggle, with MRI-sequence pretraining beating all ImageNet-pretrained baselines compared. If this is right, label-free MRI data can substitute for large labeled datasets and for generic natural-image pretraining in small medical imaging settings.

What carries the argument

The central machinery is ResViT, a hybrid CNN-transformer generator whose information bottleneck is made of Aggregated Residual Transformer (ART) blocks: each block runs the feature map through a downsampler, splits it into patches, applies a transformer encoder with multi-head self-attention and MLP layers, then deflates and up-samples the result and concatenates it with the residual CNN path before channel compression and a residual CNN. This lets local convolutional features and global attention features be fused at every block. The same encoder and ART blocks are reused in the fine-tuning stage with the decoder replaced by an MLP classification head, so the pretraining and downstream tasks share the same feature extractor. The pretext objective is MRI sequence synthesis (for example T1 to T2, T2 to T1, Flair to T1) trained with a pixel-wise loss, a reconstruction loss, and a PatchGAN adversarial loss; the synthesized images are then also used to double the minority classes during classification training.

What would settle it

A decisive check is to train the identical ResViT classifier under three initialization conditions—MRI-synthesis pretraining, ImageNet pretraining, and random initialization—holding the slice-selection and augmentation protocol fixed, and compare test accuracy across multiple seeds. If the MRI-pretrained model's margin over the others shrinks to noise, the pretext task is not the cause of the reported gains; additionally, rerunning the full pipeline with all available slices rather than the manually chosen tumor-maximum and center-close healthy slices would reveal whether the reported accuracies depend on that selection.

Watch

Extended reading notes

Core claim

The central discovery claimed is that generative self-supervised learning on MRI synthesis is a better pretraining strategy for brain tumor classification than supervised pretraining on ImageNet. In the pretext stage, ResViT is trained to translate between MRI sequences (T1 to T2, T2 to T1, Flair to T1, T1 to Flair) using pixel-wise, reconstruction, and PatchGAN adversarial losses, learning the distribution of MRI data without tumor-type labels. In the downstream stage, the pretrained encoder and Aggregated Residual Transformer blocks are kept and the decoder is replaced by an MLP head; the model is fine-tuned on real plus synthesized images. The paper reports that this pipeline reaches 90.56% accuracy on the augmented BraTS T1 set, 98.53% on Figshare, and 98.47% on Kaggle, and that the MRI-pretrained model outperforms ConvNeXtTiny, ResNet-101, DenseNet-121, and several ImageNet-pretrained ViT variants on every dataset and sequence tested.

Load-bearing premise

The load-bearing premise is that learning to synthesize one MRI sequence from another teaches general MRI features that transfer to tumor classification, and that the manually chosen slices—the five with the largest tumor region and the healthy slices nearest the brain center—are representative enough that their selection does not inflate the accuracy numbers.

Editorial extensions

If this is right

  • MRI-specific self-supervised pretraining can replace ImageNet pretraining for brain tumor classification on small labeled sets; the proposed model surpasses every ImageNet-pretrained baseline tested on BraTS, Figshare, and Kaggle.
  • The pretrained model transfers across MRI datasets: a checkpoint pretrained on BraTS sequences fine-tunes to 98.53% on Figshare and 98.47% on Kaggle, which would mean the learned MRI features are not specific to one acquisition protocol.
  • The same synthesis model serves as a data augmentation tool: adding ResViT-synthesized MRIs to the BraTS training set raises accuracy from 83.44–88.89% to 88.89–90.56% depending on the MRI sequence.
  • T1-weighted images are the most informative sequence for the downstream classification among T1, T2, and FLAIR, with the highest reported accuracy on BraTS.
  • Combining residual CNN and vision transformer branches in one architecture outperforms using either branch alone, supporting the local-plus-global feature rationale.

Reading between the lines

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

  • Editorial inference: the largest confound is the manual slice-selection protocol; rerunning the comparison on all slices or on automatically selected slices would test whether the reported margins come from the pretraining signal or from picking the most informative slices.
  • Editorial inference: because the pretext task does not use tumor labels, the same ResViT checkpoint should transfer to other downstream tasks such as tumor segmentation or survival prediction on BraTS; a positive result would strengthen the representation-learning interpretation.
  • Editorial inference: the ImageNet baselines differ in architecture from ResViT, so the cleanest controlled test of the paper's core claim would be to pretrain the identical ResViT backbone on ImageNet (or on natural images) and compare fine-tuning from that checkpoint with fine-tuning from the MRI-synthesis checkpoint.
  • Editorial inference: the reported accuracy gains from synthetic MRI augmentation suggest that generation quality and classification transfer are coupled; a targeted ablation varying the number of synthetic images per class would reveal whether the benefit is from class balancing or from the diversity of synthesized examples.
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

4 major / 6 minor

Summary. The paper proposes a two-stage generative self-supervised learning (SSL) framework for brain tumor classification. In the first stage, a Residual Vision Transformer (ResViT) is pretrained for MRI sequence synthesis (e.g., T1 to T2, Flair to T1) as a pretext task on a selected subset of BraTS 2023 slices. In the second stage, a ResViT-based classifier is fine-tuned from the pretrained encoder, with synthetic MRI images used as data augmentation. The method is evaluated on BraTS (basic and augmented), Figshare, and Kaggle brain tumor MRI datasets, reporting accuracies of 90.56% on BraTS T1 (augmented), 98.53% on Figshare, and 98.47% on Kaggle, claiming superiority over ImageNet-pretrained baselines and prior work.

Significance. If the reported results are valid, the paper would provide a useful demonstration that MRI-specific generative self-supervised pretraining can outperform generic ImageNet transfer for small medical imaging datasets, and the hybrid CNN-transformer architecture is a sensible design choice. The manuscript includes comparisons across multiple datasets and MRI sequences, and the synthesis-quality metrics (PSNR, SSIM, MSE) are standard. However, the central claim rests on the integrity of the experimental protocol, especially the separation between pretext and downstream data, and on the statistical reliability of the accuracy numbers.

major comments (4)
  1. [Section 4.1] The pretraining and classification datasets are not explicitly disjoint at the subject or slice level. The pretext set uses the top five tumor-coverage slices per BraTS case plus center-close healthy slices, while the classification set uses a single 'most relevant' slice per case from the same case pool. Since the most relevant slice is likely among the top-coverage slices, the ResViT generator is plausibly trained on the exact 2D images that later appear in the classification test set. The paper never states that classification slices were excluded from pretraining or that the train/test split was patient-level. This information leakage would directly inflate the reported gains over ImageNet-pretrained baselines in Tables 4 and 5 and undermine the central claim. Please clarify the exact overlap and rerun the main comparisons with strictly disjoint subjects and with the classification slices excluded from all pretraining data.
  2. [Section 4.2 and Tables 4-6] All accuracy, precision, recall, and F1 results are reported from a single run on a single split, without error bars, repeated seeds, or significance tests. Given the modest test-set sizes (900 BraTS slices, 613 Figshare, 1311 Kaggle), the differences between the proposed model and the best baselines may not be statistically meaningful. The authors should report mean and standard deviation over multiple runs and provide a significance test (e.g., McNemar's test) for the key comparisons.
  3. [Table 2 versus Table 5] The test set used for the augmented BraTS classification appears inconsistent with the basic BraTS test set. Table 2 lists Glioma test as 250 for the basic dataset but 200 for the augmented dataset, while the total test is 900 in both columns (the augmented total 450+200+200=850). Since augmentation should affect only the training set, the test sets should be identical. Please correct the table and state which test set was used for each reported result.
  4. [Section 4.1 and Table 2] The construction of the 'no tumor' class in the pretraining stage is not clearly described. The text says healthy slices were selected from the same 2,251 glioma and meningioma cases, but those cases are tumor cases; it is unclear whether the no-tumor class in the pretraining set consists of healthy slices from tumor patients or from separate healthy subjects. This ambiguity affects the interpretation of the pretext task and the classification setup, and it should be clarified.
minor comments (6)
  1. [Abstract and Section 3.3] The abstract refers to 'DenseNet12' while the text and tables use 'DenseNet121'; please correct the typo.
  2. [Section 4.2] There is a typo in 'pre-trained ResV˙IT model'; it should read 'ResViT'.
  3. [Section 4.1 and Table 2] The naming 'Brats' and 'BraTS' is used inconsistently throughout the paper; please standardize.
  4. [Section 4.4 and Table 7] The comparison in Table 7 lists only a subset of prior Figshare results and omits several high-performing recent methods; please clarify the selection criteria for the comparison.
  5. [Section 4.3 and Table 3] The synthesis-quality differences between ResViT and the next-best method are small (e.g., PSNR 25.663 vs. 25.105 for T1-to-T2); please indicate whether these differences are statistically significant and report confidence intervals.
  6. [Figure 1] Figure 1 is very dense and the two stages are not visually separated clearly; adding explicit labels and larger fonts would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the SSL pipeline is empirical and self-contained; the only in-scope concern is a possible slice-level overlap between pretraining and classification, which is a data-hygiene risk, not a circular derivation.

full rationale

The paper is an empirical deep-learning study, not a derivation chain. The pretext task (MRI synthesis with ResViT) and the downstream task (tumor classification via fine-tuning) are measured on held-out test splits: Section 4.1 states that datasets are split 80:20 into train and test, and Tables 4-6 report accuracy, precision, recall, and F1 on those test sets. No equation in the paper defines the classification accuracy as a function of the synthesis loss, and no fitted parameter is renamed as a prediction. The architecture is taken from Dalmaz et al. [38], and the ImageNet-pretrained transformer backbones are explicitly stated in Section 3.1, so the comparison against ImageNet pretraining is a genuine empirical benchmark rather than a self-citation chain. The only in-scope concern is in Section 4.1, where the top-five tumor-coverage slices per case are used for pretraining and a single 'most relevant' slice per case is used for classification, without an explicit statement that pretraining and classification use disjoint subjects or disjoint slices. If the same 2D slices appear in both stages, the pretrained generator may have reconstructed those exact images, which could inflate reported BraTS accuracies. This is a data-leakage/validity risk and a legitimate weakness of the experimental protocol, but it is not circular reasoning: the classification results are empirical measurements on a test split, not a logical or algebraic consequence of the pretraining objective. No self-referential definition, no imported uniqueness theorem, and no ansatz smuggled in via citation was found, so the circularity score is 0.

Assumptions & free parameters 3 free parameters · 2 assumptions · 0 invented entities

The paper introduces no new theoretical entities or parameters. It relies on empirical assumptions about the benefit of synthesis-based pretraining and the validity of synthetic augmentation, plus several hand-chosen data selection rules. These are domain assumptions rather than formal axioms.

free parameters (3)
  • Synthetic augmentation factor = 2x for glioma and meningioma
    The training set is doubled for glioma and meningioma by adding ResViT-synthesized images; the exact factor is hand-chosen and not varied or justified.
  • Slice selection rule = top-5 slices (tumor coverage) or center-close healthy slices
    The rule for selecting pretraining and classification slices is manually defined in Section 4.1 and is a post hoc choice that can affect results.
  • Loss weights lambda = not reported
    Equation 9 defines the loss as a weighted combination, but the values of lambda_pix, lambda_rec, and lambda_adv are never given, which hinders exact reproduction.
assumptions (2)
  • domain assumption Self-supervised synthesis pretext improves downstream classification
    The paper assumes that learning to synthesize MRI sequences (T1 to T2, etc.) yields features useful for tumor classification; no ablation isolates the effect of the pretext task beyond fine-tuning from ImageNet.
  • domain assumption Synthetic images are valid augmentations
    Augmentation relies on ResViT-generated images being distributionally similar to real images; the performance collapse of ConvNeXtTiny on the augmented set (from 59.67 to 36.56) casts doubt on this assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Residual Vision Transformer (ResViT) Based Self-Supervised Learning Model for Brain Tumor Classification." pith.science (2026). https://pith.science/paper/B3244JF2

@misc{pith2026241112874,
  author       = {Pith},
  title        = {Pith review of: Residual Vision Transformer (ResViT) Based Self-Supervised Learning Model for Brain Tumor Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B3244JF2}},
  note         = {Machine review of arXiv:2411.12874}
}
read the original abstract

Deep learning has proven very promising for interpreting MRI in brain tumor diagnosis. However, deep learning models suffer from a scarcity of brain MRI datasets for effective training. Self-supervised learning (SSL) models provide data-efficient and remarkable solutions to limited dataset problems. Therefore, this paper introduces a generative SSL model for brain tumor classification in two stages. The first stage is designed to pre-train a Residual Vision Transformer (ResViT) model for MRI synthesis as a pretext task. The second stage includes fine-tuning a ResViT-based classifier model as a downstream task. Accordingly, we aim to leverage local features via CNN and global features via ViT, employing a hybrid CNN-transformer architecture for ResViT in pretext and downstream tasks. Moreover, synthetic MRI images are utilized to balance the training set. The proposed model performs on public BraTs 2023, Figshare, and Kaggle datasets. Furthermore, we compare the proposed model with various deep learning models, including A-UNet, ResNet-9, pix2pix, pGAN for MRI synthesis, and ConvNeXtTiny, ResNet101, DenseNet12, Residual CNN, ViT for classification. According to the results, the proposed model pretraining on the MRI dataset is superior compared to the pretraining on the ImageNet dataset. Overall, the proposed model attains the highest accuracy, achieving 90.56% on the BraTs dataset with T1 sequence, 98.53% on the Figshare, and 98.47% on the Kaggle brain tumor datasets. As a result, the proposed model demonstrates a robust, effective, and successful approach to handling insufficient dataset challenges in MRI analysis by incorporating SSL, fine-tuning, data augmentation, and combining CNN and ViT.

Figures

Figures reproduced from arXiv: 2411.12874 by the authors.

Figure 1
Figure 1. The proposed Residual Vision Transformer (ResViT) based Self-supervised learning model. [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. The several axial slices of MRI include labels and information of tumor region with coverage that all belong [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. The samples of synthesis MRI from T2 to T1, T2 to T1, Flair to T1, and T1 to Flair. [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 29 canonical work pages

  1. [1]

    An enhanced deep learning approach for brain cancer MRI images classification using residual networks,

    S. A. Abdelaziz Ismael, A. Mohammed, and H. Hefny, “An enhanced deep learning approach for brain cancer MRI images classification using residual networks,” Artif Intell Med, vol. 102, Jan. 2020, doi: 10.1016/j.artmed.2019.101779

    arXiv 2020

  2. [2]

    Content-based brain tumor retrieval for MR images using transfer learning,

    Z. N. K. Swati et al., “Content-based brain tumor retrieval for MR images using transfer learning,” IEEE Access, vol. 7, pp. 17809–17822, 2019

    work page 2019

  3. [3]

    A Lightweight Deep Learning Based Microwave Brain Image Network Model for Brain Tumor Classification Using,

    R. Microwave et al., “A Lightweight Deep Learning Based Microwave Brain Image Network Model for Brain Tumor Classification Using,” 2023

    work page 2023

  4. [4]

    Role of Ensemble Deep Learning for Brain Tumor Classification in Multiple Magnetic Resonance Imaging Sequence Data,

    G. S. Tandel, A. Tiwari, O. G. Kakde, N. Gupta, L. Saba, and J. S. Suri, “Role of Ensemble Deep Learning for Brain Tumor Classification in Multiple Magnetic Resonance Imaging Sequence Data,” Diagnostics, 2023

    work page 2023

  5. [5]

    Brain Tumor Detection and Classification Using Intelligence Techniques: An Overview,

    S. Solanki and U. P. Singh, “Brain Tumor Detection and Classification Using Intelligence Techniques: An Overview,” vol. 11, no. January, 2023

    work page 2023

  6. [6]

    Role of Deep Learning in Classification of Brain MRI Images for Prediction of Disorders: A Survey of Emerging Trends,

    P. Rani, V . Ashish, and K. Bhandari, “Role of Deep Learning in Classification of Brain MRI Images for Prediction of Disorders: A Survey of Emerging Trends,”Archives of Computational Methods in Engineering, no. 0123456789, 2023, doi: 10.1007/s11831-023-09967-0

    work page doi:10.1007/s11831-023-09967-0 2023

  7. [7]

    Deep CNN for Brain Tumor Classification,

    W. Ayadi, W. Elhamzi, I. Charfi, and M. Atri, “Deep CNN for Brain Tumor Classification,”Neural Process Lett, vol. 53, no. 1, pp. 671–700, 2021, doi: 10.1007/s11063-020-10398-2

    work page doi:10.1007/s11063-020-10398-2 2021

  8. [8]

    Self-supervised learning for medical image classification: a systematic review and implementation guidelines,

    S. C. Huang, A. Pareek, M. Jensen, M. P. Lungren, S. Yeung, and A. S. Chaudhari, “Self-supervised learning for medical image classification: a systematic review and implementation guidelines,” npj Digital Medicine, vol. 6, no. 1. Nature Research, Dec. 01, 2023. doi: 10.1038/s41746-023-00811-0

    doi:10.1038/s41746-023-00811-0 2023

Show all 58 references
  1. [9]

    An MRI-based deep learning approach for efficient classification of brain tumors,

    E. U. Haq, H. Jianjun, K. Li, H. U. Haq, and T. Zhang, “An MRI-based deep learning approach for efficient classification of brain tumors,” Journal of Ambient Intelligence and Humanized Computing, pp. 1–22, 2021

    2021

  2. [10]

    Brain Tumor/Mass Classification Framework Using Magnetic-Resonance-Imaging-Based Isolated and Developed Transfer Deep-Learning Model,

    M. F. Alanazi et al., “Brain Tumor/Mass Classification Framework Using Magnetic-Resonance-Imaging-Based Isolated and Developed Transfer Deep-Learning Model,” Sensors, vol. 22, no. 1, 2022, doi: 10.3390/s22010372. 13 A PREPRINT - NOVEMBER 21, 2024

    2022 doi

  3. [11]

    Brain Tumor Classification Using Fine-Tuned GoogLeNet Features and Machine Learning Algorithms: IoMT Enabled CAD System,

    A. Sekhar, S. Biswas, R. Hazra, A. K. Sunaniya, A. Mukherjee, and L. Yang, “Brain Tumor Classification Using Fine-Tuned GoogLeNet Features and Machine Learning Algorithms: IoMT Enabled CAD System,” IEEE journal of biomedical and health informatics, vol. 26, no. 3, pp. 983–991,...

    2022

  4. [12]

    Deep learning,

    Y . LeCun, Y . Bengio, and G. Hinton, “Deep learning,”Nature, vol. 521, no. 7553, pp. 436–444, 2015

    2015

  5. [13]

    Convolutional neural networks in medical image understanding: a survey.,

    D. R. Sarvamangala and R. V Kulkarni, “Convolutional neural networks in medical image understanding: a survey.,”Evolutionary intelligence, vol. 15, no. 1, pp. 1–22, 2022, doi: 10.1007/s12065-020-00540-3

    2022 doi

  6. [14]

    Brain tumor classification using deep CNN features via transfer learning,

    S. Deepak and P. M. Ameer, “Brain tumor classification using deep CNN features via transfer learning,”Computers in Biology and Medicine, vol. 111, Aug. 2019, doi: 10.1016/j.compbiomed.2019.103345

    2019

  7. [15]

    Deep neural network with generative adversarial networks pre-training for brain tumor classification based on MR images,

    N. Ghassemi, A. Shoeibi, and M. Rouhani, “Deep neural network with generative adversarial networks pre-training for brain tumor classification based on MR images,” Biomedical Signal Processing Control, vol. 57, p. 101678, 2020, doi: 10.1016/j.bspc.2019.101678

    2020

  8. [16]

    Classification of brain tumors from mri images using a convolutional neural network,

    M. M. Badža and M. C. Barjaktarovi´c, “Classification of brain tumors from mri images using a convolutional neural network,” Applied Sciences (Switzerland), vol. 10, no. 6, Mar. 2020, doi: 10.3390/app10061999

    2020 doi

  9. [17]

    Enhanced brain tumor classification using an optimized multi-layered convolutional neural network architecture,

    M. Alshayeji, J. Al-Buloushi, A. Ashkanani, and S. Abed, “Enhanced brain tumor classification using an optimized multi-layered convolutional neural network architecture,” Multimedia Tools and Applications, vol. 80, no. 19, pp. 28897–28917, Aug. 2021, doi: 10.1007/s11042-021-10927-8

    2021 doi

  10. [18]

    Three-class classification of brain magnetic resonance images using average-pooling convolutional neural network,

    J. Kakarla, B. V . Isunuri, K. S. Doppalapudi, and K. S. R. Bylapudi, “Three-class classification of brain magnetic resonance images using average-pooling convolutional neural network,” International Journal of Imaging Systems and Technology, vol. 31, no. 3, pp. 1731–1740, Sep...

    2021 doi

  11. [19]

    Multi-class brain tumor classification using residual network and global average pooling,

    R. L. Kumar, J. Kakarla, B. V . Isunuri, and M. Singh, “Multi-class brain tumor classification using residual network and global average pooling,” Multimedia Tools and Applications, vol. 80, no. 9, pp. 13429–13438, Apr. 2021, doi: 10.1007/s11042-020-10335-4

    2021 doi

  12. [20]

    BTC-fCNN: Fast Convolution Neural Network for Multi-class Brain Tumor Classification,

    B.S. Abd El-Wahab, M. E. Nasr, S. Khamis, and A. S. Ashour, “BTC-fCNN: Fast Convolution Neural Network for Multi-class Brain Tumor Classification,” Health Information Science and Systems, 2023, doi: 10.1007/s13755- 022-00203-w

    2023 doi

  13. [21]

    MedViT: a robust vision transformer for generalized medical image classification,

    O. N. Manzari, H. Ahmadabadi, H. Kashiani, S. B. Shokouhi, and A. Ayatollahi, “MedViT: a robust vision transformer for generalized medical image classification,” Computers in Biology and Medicine, vol. 157, p. 106791, 2023

    2023

  14. [22]

    A survey on vision transformer,

    K. Han et al., “A survey on vision transformer,” IEEE Trans Pattern Anal Mach Intell, vol. 45, no. 1, pp. 87–110, 2022

    2022

  15. [23]

    Pyramid vision transformer: A versatile backbone for dense prediction without convolutions,

    W. Wanget al., “Pyramid vision transformer: A versatile backbone for dense prediction without convolutions,” in Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 568–578

    2021

  16. [24]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy et al., “ An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2020

    2010 arXiv

  17. [25]

    Transmed: Transformers advance multi-modal medical image classification,

    Y . Dai, Y . Gao, and F. Liu, “Transmed: Transformers advance multi-modal medical image classification,”Diag- nostics, vol. 11, no. 8, Aug. 2021, doi: 10.3390/diagnostics11081384

    2021 doi

  18. [26]

    Combining the Transformer and Convolution for Effective Brain Tumor Classification Using MRI Images,

    M. Aloraini, A. Khan, S. Aladhadh, S. Habib, M. F. Alsharekh, and M. Islam, “Combining the Transformer and Convolution for Effective Brain Tumor Classification Using MRI Images,”Applied Sciences, 2023

    2023

  19. [27]

    Attention transformer mechanism and fusion-based deep learning architecture for MRI brain tumor classification system,

    S. Tabatabaei, K. Rezaee, and M. Zhu, “Attention transformer mechanism and fusion-based deep learning architecture for MRI brain tumor classification system,” Biomedical Signal Processing Control, vol. 86, Sep. 2023, doi: 10.1016/j.bspc.2023.105119

    2023

  20. [28]

    LCDEiT: A Linear Complexity Data-Efficient Image Transformer for MRI Brain Tumor Classification,

    G. J. Ferdous, K. A. Sathi, M. A. Hossain, M. M. Hoque, and M. Ali Akber Dewan, “LCDEiT: A Linear Complexity Data-Efficient Image Transformer for MRI Brain Tumor Classification,”IEEE Access, vol. 11, pp. 20337–20350, 2023, doi: 10.1109/ACCESS.2023.3244228

    2023

  21. [29]

    DiRA: Discriminative, Restorative, and Adversarial Learning for Self-supervised Medical Image Analysis,

    F. Haghighi, M. Reza, H. Taher, M. B. Gotway, and J. Liang, “DiRA: Discriminative, Restorative, and Adversarial Learning for Self-supervised Medical Image Analysis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 20824-20834

    2022

  22. [30]

    3D Self-Supervised Methods for Medical Imaging,

    A. Taleb et al., “3D Self-Supervised Methods for Medical Imaging,” Advances in neural information processing systems, vol. 33, pp. 18158-18172, 2020

    2020

  23. [31]

    Models Genesis,

    Z. Zhou, V . Sodha, J. Pang, M. B. Gotway, and J. Liang, “Models Genesis,”Medical Image Analysis, vol. 67, Jan. 2021, doi: 10.1016/j.media.2020.101840. 14 A PREPRINT - NOVEMBER 21, 2024

    2021

  24. [32]

    Self-supervised driven consistency training for annotation efficient histopathology image analysis,

    C. L. Srinidhi, S. W. Kim, F. Der Chen, and A. L. Martel, “Self-supervised driven consistency training for annotation efficient histopathology image analysis,” Medical Image Analysis, vol. 75, Jan. 2022, doi: 10.1016/j.media.2021.102256

    2022

  25. [33]

    Transformer-based unsupervised contrastive learning for histopathological image classification,

    X. Wang et al., “Transformer-based unsupervised contrastive learning for histopathological image classification,” Medical Image Analysis, vol. 81, Oct. 2022, doi: 10.1016/j.media.2022.102559

    2022

  26. [34]

    SAM: Self-Supervised Learning of Pixel-Wise Anatomical Embeddings in Radiologi- cal Images,

    K. Yan et al., “ SAM: Self-Supervised Learning of Pixel-Wise Anatomical Embeddings in Radiologi- cal Images,” IEEE Transactions on Medical Imaging , vol. 41, no. 10, pp. 2658–2669, Oct. 2022., doi: 10.1109/TMI.2022.3169003

    2022

  27. [35]

    Attention De-sparsification Matters: Inducing diversity in digital pathology representation learning,

    S. Kapse et al., “ Attention De-sparsification Matters: Inducing diversity in digital pathology representation learning,” Medical Image Analysis, vol. 93, pp. 103070, 2024

    2024

  28. [36]

    A General-Purpose Self-Supervised Model for Computational Pathology,

    R. J. Chen et al., “A General-Purpose Self-Supervised Model for Computational Pathology,” Aug. 2023, [Online]. Available: http://arxiv.org/abs/2308.15474

    2023 arXiv

  29. [37]

    Virchow: A Million-Slide Digital Pathology Foundation Model,

    E. V orontsovet al., “ Virchow: A Million-Slide Digital Pathology Foundation Model,” Sep. 2023, [Online]. Available: http://arxiv.org/abs/2309.07778

    2023 arXiv

  30. [38]

    ResViT: Residual Vision Transformers for Multimodal Medical Im- age Synthesis,

    O. Dalmaz, M. Yurt, and T. Cukur, “ResViT: Residual Vision Transformers for Multimodal Medical Im- age Synthesis,” IEEE Transactions on Medical Imaging , vol. 41, no. 10, pp. 2598–2614, 2022, doi: 10.1109/TMI.2022.3167808

    2022

  31. [39]

    Image-to-image translation with conditional adversarial networks,

    P. Isola, J.-Y . Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 1125–1134

    2017

  32. [40]

    Image synthesis in multi-contrast MRI with conditional generative adversarial networks,

    S. U. H. Dar, M. Yurt, L. Karacan, A. Erdem, E. Erdem, and T. Çukur, “Image synthesis in multi-contrast MRI with conditional generative adversarial networks,”IEEE Transactions on Medical Imaging, vol. 38, no. 10, pp. 2375–2388, Oct. 2019

    2019

  33. [41]

    Attention is all you need,

    A. Vaswani et al., “Attention is all you need,” Adv Neural Inf Process Syst , vol. 30, 2017

    2017

  34. [42]

    Haykin, Neural networks: a comprehensive foundation

    S. Haykin, Neural networks: a comprehensive foundation . Prentice Hall PTR, 1998

    1998

  35. [43]

    Attention u-net: Learning where to look for the pancreas,

    O. Oktay et al., “Attention u-net: Learning where to look for the pancreas,” arXiv preprint arXiv:1804.03999, 2018

    2018 arXiv

  36. [44]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778

    2016

  37. [45]

    A ConvNet for the 2020s,

    Z. Liu, H. Mao, C.-Y . Wu, C. Feichtenhofer, T. Darrell, and S. Xie, “A ConvNet for the 2020s,”arXiv preprint arXiv:2201.03545, 2022

    2022 arXiv

  38. [46]

    Identity mappings in deep residual networks,

    K. He, X. Zhang, S. Ren, and J. Sun, “Identity mappings in deep residual networks,” in Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part IV 14 , Springer, 2016, pp. 630–645

    2016

  39. [47]

    Densely connected convolutional networks,

    G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 4700–4708

    2017

  40. [48]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in 2009 IEEE conference on computer vision and pattern recognition, Jun. 2009

    2009

  41. [49]

    Brain Tumor MRI Dataset,

    M. Nickparvar, “Brain Tumor MRI Dataset,” Kaggle, 2021. [Online]. Available: https://doi.org/10.34740/KAGGLE/DSV/2645886

    2021

  42. [50]

    brain tumor dataset

    J. Cheng, “brain tumor dataset”. figshare, 2017, [Online]. Available: https://doi.org/10.6084/m9.figshare.1512427.v5 Accessed on: Apr. 2, 2017

    2017 doi

  43. [51]

    The cancer imaging archive (TCIA): maintaining and operating a public information repository,

    K. Clark, B. Vendt, K. Smith, J. Freymann, J. Kirby, P. Koppel, and L.Tarbox, “ The cancer imaging archive (TCIA): maintaining and operating a public information repository,” Journal of digital imaging, vol. 26, no. 6, pp. 1045–1057, Jul. 2013

    2013

  44. [52]

    The rsna-asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification,

    U. Baid et al., “ The rsna-asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification,” arXiv preprint arXiv:2107.02314, 2021

    2021 arXiv

  45. [53]

    The multimodal brain tumor image segmentation benchmark (BRATS),

    B. H. Menze et al., “The multimodal brain tumor image segmentation benchmark (BRATS),”IEEE Transactions on Medical Imaging 2014, vol. 34, no. 10, pp. 1993–2024,

    2014

  46. [54]

    Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features,

    S. Bakas et al., “Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features,” Sci Data, vol. 4, no. 1, pp. 1–13, 2017. 15 A PREPRINT - NOVEMBER 21, 2024

    2017

  47. [55]

    Segmentation Labels and Radiomic Features for the Pre-operative Scans of the TCGA-GBM collection

    S. Bakas, H. Akbari, A. Sotiras, M. Bilello, M. Rozycki, J. Kirby, et al., “Segmentation Labels and Radiomic Features for the Pre-operative Scans of the TCGA-GBM collection", The Cancer Imaging Archive, 2017. DOI: 10.7937/K9/TCIA.2017.KLXWJJ1Q

    2017 doi

  48. [56]

    Segmentation Labels and Radiomic Features for the Pre-operative Scans of the TCGA-LGG collection

    S. Bakas, H. Akbari, A. Sotiras, M. Bilello, M. Rozycki, J. Kirby, et al., “ Segmentation Labels and Radiomic Features for the Pre-operative Scans of the TCGA-LGG collection", The Cancer Imaging Archive, 2017. DOI: 10.7937/K9/TCIA.2017.GJQ7R0EF

    2017 doi

  49. [57]

    The ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge 2023: Intracranial Meningioma,

    D. LaBella et al., “ The ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge 2023: Intracranial Meningioma,” arXiv preprint arXiv:2305.07642, 2023

    2023 arXiv

  50. [58]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H.R. Sheikh, E. P. Simoncelli EP. “Image quality assessment: from error visibility to structural similarity,” IEEE transactions on image processing, vol. 13, no. 4, pp. 600–612, Apr. 2004. 16

    2004

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.