Pith. sign in

REVIEW 3 major objections 6 minor 50 references

GIT-CXR: End-to-End Transformer for Chest X-Ray Report Generation

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read An end-to-end transformer with a report-length curriculum sets new state-of-the-art scores for chest X-ray report generation on METEOR and clinical F1 metrics.

desk verdict Solid internal ablation of a simple GIT-based report generator, but the SOTA claims rest on cross-paper metric comparisons that the paper's own documentation shows are not apples-to-apples. read the letter →

arxiv 2501.02598 v1 pith:XS54GAXT submitted 2025-01-05 cs.CL cs.CVcs.LG

Iustin Sîrbu , Iulia-Renata Sîrbu , Jasmina Bogojeska , Traian Rebedea This is my paper · ORCID
classification cs.CLcs.CVcs.LG
keywords radiologyreportgenerationchestX-rayend-to-endtransformercurriculumlearningMIMIC-CXR-JPGclinicalaccuracyMETEORimagecaptioning
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

This paper argues that a deliberately simple end-to-end transformer, built by fine-tuning the GIT image-to-text model, can generate chest X-ray reports that match or beat substantially more complex systems on the standard MIMIC-CXR-JPG benchmark. Its central innovation is a curriculum-learning schedule that sorts training reports by length and gradually shifts the model from short to long targets; the authors identify this schedule as the main driver of improved performance on long reports. If the claim is right, it matters because the best previous results came from systems with extra machinery such as object detectors, memory modules, and multi-stage classification, while GIT-CXR shows a plain transformer plus report-length scheduling can reach comparable or better scores. On top of this, the paper reports new top results on the NLG metric METEOR and on the clinical accuracy F1 metrics (macro, micro, and examples-averaged), with BLEU and ROUGE-L on par with earlier methods. The ablation also documents an incompatibility: adding an auxiliary classification head to the curriculum-trained model hurts performance, because length-based sampling distorts the pathology distribution the head sees.

What carries the argument

The load-bearing mechanism is the length-based curriculum. Reports are split into $b=10$ equal-size bins by token length; during each epoch the model samples without replacement a fraction $f=0.25$ of the dataset with per-bin weights $1/(1+|i-i_e|)$, so the effective training distribution peaks at bin $i_e$ and advances from short to long reports across training. This prevents the decoder from collapsing toward short generic output and is what the paper credits for the flattened long-report performance curves. Supporting pieces are a temporal embedding that distinguishes the two views, concatenation of the 'indication' and 'history' fields as context, and an auxiliary multi-label classification head with weighted cross-entropy over the 14 CheXbert labels, which helps only when the curriculum is not active.

What would settle it

Take the strongest prior systems listed in Table 1 and run them through the preprocessing and evaluation code used here (concatenated impression plus findings, CheXbert labels, official splits), then re-measure METEOR, F1-macro, and F1-micro. If the gaps shrink to the noise level of the three-run averages, the state-of-the-art claim collapses; if they persist, it stands. A cheaper partial check is to ablate only the curriculum from GIT-CXR (MV+C+CL) and compare the length-binned metric curves.

Watch

Extended reading notes

Core claim

The paper's central claim is that the configuration GIT-CXR (MV+C+CL) — a GIT transformer fine-tuned end-to-end on MIMIC-CXR-JPG using two-view images, concatenated indication/history context, and a length-based curriculum — sets a new state of the art on METEOR (0.369) and on the clinical accuracy metrics F1-macro (0.348), F1-micro (0.534), and F1 examples-averaged (0.458) on the full official test set, while matching the previous best on BLEU and ROUGE-L. The paper further claims that the curriculum is the main source of the gain: without it, generated reports run short, and METEOR, ROUGE-L, and F1-micro decline steeply as target reports exceed roughly 75 tokens, whereas with it the long-report curve flattens. The authors take this as evidence that simpler end-to-end transformer architectures can outperform more complex prior systems when paired with a training-order strategy that addresses the long-report problem directly.

Load-bearing premise

The state-of-the-art claims assume that the scores quoted from earlier papers were measured under the same protocol — the same target report sections, the same CheXbert labeler, the same official test split, and compatible metric implementations — so the gaps in Table 1 are true performance differences rather than evaluation-pipeline artifacts.

Editorial extensions

If this is right

  • Future radiology report generators can drop extra modules such as object detectors or memory banks and instead pair an end-to-end transformer with a report-length curriculum, lowering training and inference complexity.
  • The paper's long-report analysis implies that any method evaluated on MIMIC-CXR-JPG should report performance broken down by target length, since aggregate scores hide the systematic shortfall on the longest reports.
  • Because the curriculum and the classification head fight each other, a direct next step suggested by the paper is to find a mixing scheme that lets a model benefit from both techniques simultaneously.
  • The reported figures — METEOR 0.369, F1-macro 0.348, F1-micro 0.534, and F1 examples-averaged 0.458 — become the numbers a new system must beat under the official splits with CheXbert labeling.

Reading between the lines

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

  • One testable extension the paper does not run is a curriculum defined by clinical content, such as the number of pathologies mentioned, instead of raw report length; such a curriculum might avoid the pathology-distribution skew that makes the classification head incompatible with length-based training.
  • Because the paper truncates targets at 192 tokens and shows a performance decline with length, a natural follow-up is to raise the token cap and refine the bin schedule; the paper establishes the trend but does not test the boundary.
  • The clinical F1 numbers are computed by an automated labeler, so the paper does not establish that the improved F1 translates into expert-reader agreement; a reader study on the same test reports would be a meaningful next check.
  • The reported gaps over prior work rest on cross-paper metric comparability, which the paper itself notes is imperfect because prior work differs in report sections and labelers; re-running the leading baselines under this paper's pipeline would settle how much of the gap is real.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper presents GIT-CXR, an end-to-end transformer architecture for chest X-ray report generation, built by adapting the GIT model with a multi-label classification head, multi-view image inputs, patient context, and a length-based curriculum learning schedule. Experiments on the MIMIC-CXR-JPG dataset report that the best model (GIT-CXR MV+C+CL) achieves state-of-the-art results on METEOR and on F1-macro/F1-micro clinical accuracy metrics, while remaining comparable to prior work on BLEU and ROUGE-L. The ablation study shows that each added component improves the model on the same evaluation pipeline, and the paper argues that curriculum learning is particularly beneficial for generating long reports.

Significance. If the internal ablation is taken at face value, the paper makes a useful empirical contribution: a simple end-to-end transformer with a length-based curriculum consistently improves over its own baseline, and the aggregate curriculum gains are reported as statistically significant. The authors also provide a thorough ablation, per-pathology analysis, and use the official MIMIC-CXR splits, which is more rigorous than many prior works. However, the headline claims of new state-of-the-art on METEOR and F1-macro/F1-micro rest on cross-paper comparisons with documented protocol mismatches (target sections, labeler, splits, metric library). The significance of the contribution therefore hinges on whether those claims survive a controlled re-evaluation; the curriculum-learning idea is novel in this specific setting and could be of interest, but its claimed advantage for long reports needs stronger statistical support.

major comments (3)
  1. [Section 4.1, Table 1, Appendix B] The headline SOTA claim is not supported by a controlled comparison. The paper uses 'findings' plus 'impression' as the target report, while the baselines of Nguyen et al. (2021), Lovelace and Mortazavi (2020), and Miura et al. (2020) use only 'findings'; Table 1 also marks AGA, LOVE, MMTN, CXR-R EPAIR, and M2TR as not using the official splits. These differences change the length and content distribution of targets, which directly affects BLEU/ROUGE/METEOR scores and the derived clinical labels. To substantiate the claim of new state-of-the-art, the authors should re-evaluate at least the most competitive baselines under the same target-section, labeler, split, and metric pipeline, or restrict the SOTA claim to methods evaluated under identical conditions.
  2. [Section 4.2, Appendix C] The clinical and NLG metrics are not directly comparable across papers. The authors compute NLG metrics with the Huggingface library, whereas prior work predominantly uses pycocoevalcap; METEOR and BLEU are known to differ by implementation and tokenization. For clinical accuracy, the authors use CheXbert labels, while Chen et al. (2020), Lovelace and Mortazavi (2020), and Nguyen et al. (2021) use CheXpert. Labeler choice alone can shift per-pathology F1 by several points. The reported margins over baselines are large, but without a shared-pipeline comparison there is no evidence that the ordering holds; this undermines the F1-macro/F1-micro SOTA claims.
  3. [Section 5.1, Figure 3] The claim that curriculum learning is the main driver of improved performance on long reports is supported only by visual inspection of curves without error bars or per-bucket significance tests. The paper states that all curriculum improvements are statistically significant (p<0.01) for aggregate scores, but not for the length-stratified analysis in Figure 3. To support the load-bearing claim about long-report generation, the authors should report confidence intervals or significance tests on the per-length-bucket metrics, and clarify how the length bins are defined and how many samples fall in each bin.
minor comments (6)
  1. [Throughout] There are several typos and grammatical errors: 'it's efficacy' should be 'its efficacy', 'weather' in Section 4.2 should be 'whether', 'descent accuracy' in Section 5.2 should be 'decent accuracy', and 'We address this on the fact' in Section 5.1 should be 'We attribute this to the fact'.
  2. [Section 3.3, Equation (1)] The summation notation 'i≤DX' in Equation (1) is malformed; it should be written as a sum over i=1 to D.
  3. [Tables 1 and 2] The paper reports averages of three runs but does not provide standard deviations or confidence intervals; the claim of statistical significance (p<0.01) would be better supported by reporting variance alongside the mean values.
  4. [Figure 3] The curves in Figure 3 would be more informative with shaded confidence intervals or error bars, especially since the text states that short sequences are rare and the metrics are less reliable there.
  5. [Table 3] The column 'SUPPORT' in Table 3 is not defined in the caption; it appears to be the number of positive samples per pathology, but this should be stated explicitly.
  6. [Appendix A.2] The paper states 'We will make our code publicly available' in two places; consider including a URL or repository reference in the final version.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: performance is measured on a held-out test set after standard validation-based checkpoint selection; the SOTA-comparison caveats are protocol-comparability concerns, not circular reasoning.

full rationale

The paper's derivation chain is empirical and self-contained. The proposed model fine-tunes GIT on MIMIC-CXR-JPG with multi-view inputs, patient context, an auxiliary multi-label classification loss, and a length-based curriculum schedule. None of these components define a target metric in terms of the model's own outputs or fit the reported test metrics by construction. The curriculum weighting formula (1/(1+|i-ie|)) and binning are fixed before training, and the best checkpoints are selected on the validation set using a weighted average of NLG metrics (AVG_NLG), which is standard model selection rather than a fitted input renamed as a prediction. All final results are reported on the held-out MIMIC-CXR-JPG test set and averaged over three runs, with a t-test for the curriculum comparisons. The clinical F1 metrics use CheXbert labels both for the auxiliary classification loss and for evaluating generated reports; this is a methodological choice that may affect comparability with prior work, but it is not a definitional equivalence between the training objective and the reported metric—the model can still fail to reproduce CheXbert's labels. The paper's SOTA claims rely on baseline numbers taken from previous papers with differing target sections, labelers, dataset splits, and metric implementations, as clearly documented in Sections 4.1, 4.2, and Appendices B and C. That is a threat to the comparability of the headline comparison, but it is not circularity under the strict definition: the paper does not derive its own scores from those baselines, and its ablation results stand independently of the cross-paper comparison. There are no load-bearing self-citations, no imported uniqueness theorems, and no ansatz smuggled in solely through the authors' prior work. Consequently, the central claims are not circular.

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

The central empirical contribution rests on a handful of tuned hyperparameters and several assumptions about evaluation comparability. The model introduces no new entities, but the SOTA claims inherit the assumption that metrics from different papers and pipelines are comparable.

free parameters (6)
  • Learning rate = 5e-5
    Tuned on validation A VG_NLG for GIT-CXR (SV+C); used for all models.
  • Classification loss weight = 0.1
    Tuned on validation for GIT-CXR-CLS (SV+C) from {0.1, 1}.
  • Curriculum sampling fraction f = 0.25
    Tuned on validation for GIT-CXR (SV+C+CL) from {0.1, 0.25}.
  • Number of curriculum bins b = 10
    Set without reported tuning; controls difficulty schedule.
  • Class weights for weighted cross entropy = not specified
    Computed per pathology, but exact values are not reported; changes classification loss balance.
  • Checkpoint selection weights A VG_NLG = METEOR 0.25, ROUGE-L 0.25, BLEU1-4 0.125 each
    Chosen to maximize all NLG metrics; influences which epoch is selected.
assumptions (4)
  • domain assumption GIT base weights pretrained on MSRVTT-QA provide a suitable initialization for medical report generation.
    The authors use GITb pretrained on video QA because it has temporal embeddings; no evidence this is better than other initializations for chest X-ray.
  • domain assumption CheXbert labeler produces accurate enough labels for evaluating generated reports.
    CheXbert has near-radiologist performance, but is itself an imperfect model; errors propagate to F1 scores.
  • domain assumption Huggingface implementations of BLEU, ROUGE-L, and METEOR are directly comparable to implementations used in prior baseline papers.
    SOTA comparisons in Tables 1 and 2 assume metric scores from different libraries and preprocessing are equivalent.
  • ad hoc to paper Report length is a valid proxy for generation difficulty in curriculum learning.
    The curriculum uses target length as the difficulty measure; if length does not track difficulty, the method's rationale weakens.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GIT-CXR: End-to-End Transformer for Chest X-Ray Report Generation." pith.science (2026). https://pith.science/paper/XS54GAXT

@misc{pith2026250102598,
  author       = {Pith},
  title        = {Pith review of: GIT-CXR: End-to-End Transformer for Chest X-Ray Report Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XS54GAXT}},
  note         = {Machine review of arXiv:2501.02598}
}
read the original abstract

Medical imaging is crucial for diagnosing, monitoring, and treating medical conditions. The medical reports of radiology images are the primary medium through which medical professionals attest their findings, but their writing is time consuming and requires specialized clinical expertise. The automated generation of radiography reports has thus the potential to improve and standardize patient care and significantly reduce clinicians workload. Through our work, we have designed and evaluated an end-to-end transformer-based method to generate accurate and factually complete radiology reports for X-ray images. Additionally, we are the first to introduce curriculum learning for end-to-end transformers in medical imaging and demonstrate its impact in obtaining improved performance. The experiments have been conducted using the MIMIC-CXR-JPG database, the largest available chest X-ray dataset. The results obtained are comparable with the current state-of-the-art on the natural language generation (NLG) metrics BLEU and ROUGE-L, while setting new state-of-the-art results on F1 examples-averaged, F1-macro and F1-micro metrics for clinical accuracy and on the METEOR metric widely used for NLG.

Figures

Figures reproduced from arXiv: 2501.02598 by the authors.

Figure 1
Figure 1. Example of a study of one patient. riculum learning in other NLP tasks. Spitkovsky et al. (2009) shows the importance of an easy-to￾hard training strategy for unsupervised grammar induction, while Chang et al. (2021) applies cur￾riculum learning for data-to-text generation and shows that it improves both generation quality and convergence speed. In the field of medical imaging, curriculum learn￾ing has been mostly u… view at source ↗
Figure 2
Figure 2. Our proposed methods with their setups. MODEL BL1 BL2 BL3 BL4 RGL M F1MA F1MI F1MI5 F1EX GIT-CXR (MV+C+CL) 0.403 0.286 0.215 0.168 0.312 0.369 0.348 0.534 0.565 0.458 GIT-CXR (SV+C+CL) 0.393 0.278 0.208 0.162 0.305 0.359 0.327 0.505 0.538 0.428 ARR TR (2022B) 0.351 0.223 0.157 0.118 0.287 − − − − − RGRG (2023) 0.373 0.249 0.175 0.126 0.264 0.168 − − 0.547 0.447 EKAGEN (2024) 0.419 0.258 0.170 0.119 0.287 0.157 − 0.4… view at source ↗
Figure 3
Figure 3. The metrics score evolution with report length. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Distributions for the length of the generated [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 25 canonical work pages

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Mohammad Alsharid, Rasheed El-Bouri, Harshita Sharma, Lior Drukker, Aris T Papageorghiou, and J Alison Noble. 2020. A curriculum learning based approach to captioning ultrasound images. In Medical Ultrasound, and Preterm, Perinatal and Paediatric Image Analysis: First International Workshop, ASMUS 2020, and 5th International Workshop, PIPPI 2020, Held in ...

    work page 2020

  4. [4]

    Satanjeev Banerjee and Alon Lavie. 2005. Meteor: An automatic metric for mt evaluation with improved correlation with human judgments. In Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization, pages 65--72

    2005

  5. [5]

    Yoshua Bengio, J \'e r \^o me Louradour, Ronan Collobert, and Jason Weston. 2009. Curriculum learning. In Proceedings of the 26th annual international conference on machine learning, pages 41--48

    2009

  6. [6]

    William Boag, Tzu-Ming Harry Hsu, Matthew McDermott, Gabriela Berner, Emily Alesentzer, and Peter Szolovits. 2020. Baselines for chest x-ray report generation. In Machine learning for health workshop, pages 126--140. PMLR

    work page 2020

  7. [7]

    Shenshen Bu, Taiji Li, Yuedong Yang, and Zhiming Dai. 2024. Instance-level expert knowledge and aggregate discriminative attention for radiology report generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14194--14204

    2024

  8. [8]

    Yiming Cao, Lizhen Cui, Lei Zhang, Fuqiang Yu, Zhen Li, and Yonghui Xu. 2023. Mmtn: multi-modal memory transformer network for image-report consistent medical report generation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 277--285

    work page 2023

Show all 50 references
  1. [9]

    Ernie Chang, Hui-Syuan Yeh, and Vera Demberg. 2021. Does the order of training samples matter? improving neural data-to-text generation with curriculum learning. arXiv preprint arXiv:2102.03554

    2021 arXiv

  2. [10]

    Zhihong Chen, Yan Song, Tsung-Hui Chang, and Xiang Wan. 2020. Generating radiology reports via memory-driven transformer. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1439--1449

    2020

  3. [11]

    Marcella Cornia, Matteo Stefanini, Lorenzo Baraldi, and Rita Cucchiara. 2020. Meshed-memory transformer for image captioning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10578--10587

    2020

  4. [12]

    Louke Delrue, Robert Gosselin, Bart Ilsen, An Van Landeghem, Johan de Mey, and Philippe Duyck. 2011. Difficulties in the interpretation of chest radiography. Comparative interpretation of CT and standard radiography of the chest, pages 27--49

    2011

  5. [13]

    Dina Demner-Fushman, Marc D Kohli, Marc B Rosenman, Sonya E Shooshan, Laritza Rodriguez, Sameer Antani, George R Thoma, and Clement J McDonald. 2016. Preparing a collection of radiology examinations for distribution and retrieval. Journal of the American Medical Informatics As...

    2016

  6. [14]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805

    2018 arXiv

  7. [15]

    Mark Endo, Rayan Krishnan, Viswesh Krishna, Andrew Y Ng, and Pranav Rajpurkar. 2021. Retrieval-based chest x-ray report generation using a pre-trained contrastive language-image model. In Machine Learning for Health, pages 209--219. PMLR

    2021

  8. [16]

    Sepp Hochreiter and J \"u rgen Schmidhuber. 1997. Long short-term memory. Neural computation, 9(8):1735--1780

    1997

  9. [17]

    Jeremy Irvin, Pranav Rajpurkar, Michael Ko, Yifan Yu, Silviana Ciurea-Ilcus, Chris Chute, Henrik Marklund, Behzad Haghgoo, Robyn Ball, Katie Shpanskaya, et al. 2019. Chexpert: A large chest radiograph dataset with uncertainty labels and expert comparison. In Proceedings of the...

    2019

  10. [18]

    Amelia Jim \'e nez-S \'a nchez, Diana Mateus, Sonja Kirchhoff, Chlodwig Kirchhoff, Peter Biberthaler, Nassir Navab, Miguel A Gonz \'a lez Ballester, and Gemma Piella. 2022. Curriculum learning for improved femur fracture classification: Scheduling data with prior knowledge and...

    2022

  11. [19]

    Baoyu Jing, Pengtao Xie, and Eric Xing. 2017. On the automatic generation of medical imaging reports. arXiv preprint arXiv:1711.08195

    2017 arXiv

  12. [20]

    Alistair EW Johnson, Tom J Pollard, Seth J Berkowitz, Nathaniel R Greenbaum, Matthew P Lungren, Chih-ying Deng, Roger G Mark, and Steven Horng. 2019 a . Mimic-cxr, a de-identified publicly available database of chest radiographs with free-text reports. Scientific data, 6(1):317

    2019

  13. [21]

    Alistair EW Johnson, Tom J Pollard, Nathaniel R Greenbaum, Matthew P Lungren, Chih-ying Deng, Yifan Peng, Zhiyong Lu, Roger G Mark, Seth J Berkowitz, and Steven Horng. 2019 b . Mimic-cxr-jpg, a large publicly available database of labeled chest radiographs. arXiv preprint arXi...

    2019 arXiv

  14. [22]

    Christy Y Li, Xiaodan Liang, Zhiting Hu, and Eric P Xing. 2019. Knowledge-driven encode, retrieve, paraphrase for medical image report generation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, pages 6666--6673

    2019

  15. [23]

    Yuan Li, Xiaodan Liang, Zhiting Hu, and Eric P Xing. 2018. Hybrid retrieval-generation reinforced agent for medical image report generation. Advances in neural information processing systems, 31

    2018

  16. [24]

    Chin-Yew Lin. 2004. Rouge: A package for automatic evaluation of summaries. In Text summarization branches out, pages 74--81

    2004

  17. [25]

    Fenglin Liu, Shen Ge, Yuexian Zou, and Xian Wu. 2022. Competence-based multimodal curriculum learning for medical report generation. arXiv preprint arXiv:2206.14579

    2022 arXiv

  18. [26]

    William Lotter, Greg Sorensen, and David Cox. 2017. A multi-scale cnn and curriculum learning strategy for mammogram classification. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: Third International Workshop, DLMIA 2017, and ...

    2017

  19. [27]

    Justin Lovelace and Bobak Mortazavi. 2020. Learning to generate clinically coherent chest x-ray reports. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 1235--1243

    2020

  20. [28]

    Yasuhide Miura, Yuhao Zhang, Emily Bao Tsai, Curtis P Langlotz, and Dan Jurafsky. 2020. Improving factual completeness and consistency of image-to-text radiology report generation. arXiv preprint arXiv:2010.10042

    2020 arXiv

  21. [29]

    Hoang TN Nguyen, Dong Nie, Taivanbat Badamdorj, Yujie Liu, Yingying Zhu, Jason Truong, and Li Cheng. 2021. Automated generation of accurate & fluent medical x-ray reports. arXiv preprint arXiv:2108.12126

    2021 arXiv

  22. [30]

    Van-Quang Nguyen, Masanori Suganuma, and Takayuki Okatani. 2022. Grit: Faster and better image captioning transformer using dual visual features. In European Conference on Computer Vision, pages 167--184. Springer

    2022

  23. [31]

    Aaron Nicolson, Jason Dowling, and Bevan Koopman. 2023. Improving chest x-ray report generation by leveraging warm starting. Artificial intelligence in medicine, 144:102633

    2023

  24. [32]

    Ilkay Oksuz, Bram Ruijsink, Esther Puyol-Ant \'o n, James R Clough, Gastao Cruz, Aurelien Bustin, Claudia Prieto, Rene Botnar, Daniel Rueckert, Julia A Schnabel, et al. 2019. Automatic cnn-based detection of cardiac mr motion artefacts using k-space data augmentation and curri...

    2019

  25. [33]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting of the Association for Computational Linguistics, pages 311--318

    2002

  26. [34]

    Akshay Smit, Saahil Jain, Pranav Rajpurkar, Anuj Pareek, Andrew Y Ng, and Matthew P Lungren. 2020. Chexbert: combining automatic labelers and expert annotations for accurate radiology report labeling using bert. arXiv preprint arXiv:2004.09167

    2020 arXiv

  27. [35]

    Karen Sparck Jones. 1972. A statistical interpretation of term specificity and its application in retrieval. Journal of documentation, 28(1):11--21

    1972

  28. [36]

    less is more

    Valentin I Spitkovsky, Hiyan Alshawi, and Daniel Jurafsky. 2009. Baby steps: How “less is more” in unsupervised dependency parsing. NIPS: Grammar Induction, Representation of Language and Language Learning, pages 1--10

    2009

  29. [37]

    Preethi Srinivasan, Daksh Thapar, Arnav Bhavsar, and Aditya Nigam. 2020. Hierarchical x-ray report generation via pathology tags and multi head attention. In Proceedings of the Asian Conference on Computer Vision

    2020

  30. [38]

    Sandeep Subramanian, Sai Rajeswar, Francis Dutil, Christopher Pal, and Aaron Courville. 2017. Adversarial generation of natural language. In Proceedings of the 2nd Workshop on Representation Learning for NLP, pages 241--251

    2017

  31. [39]

    Tim Tanida, Philip M \"u ller, Georgios Kaissis, and Daniel Rueckert. 2023. Interactive and explainable region-guided radiology report generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7433--7442

    2023

  32. [40]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, ukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  33. [41]

    Oriol Vinyals, Alexander Toshev, Samy Bengio, and Dumitru Erhan. 2015. Show and tell: A neural image caption generator. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3156--3164

    2015

  34. [42]

    Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, and Lijuan Wang. 2022 a . Git: A generative image-to-text transformer for vision and language. arXiv preprint arXiv:2205.14100

    2022 arXiv

  35. [43]

    Zhanyu Wang, Hongwei Han, Lei Wang, Xiu Li, and Luping Zhou. 2022 b . Automated radiographic report generation purely on transformer: A multicriteria supervised approach. IEEE Transactions on Medical Imaging, 41(10):2803--2813

    2022

  36. [44]

    Jerry Wei, Arief Suriawinata, Bing Ren, Xiaoying Liu, Mikhail Lisovsky, Louis Vaickus, Charles Brown, Michael Baker, Mustafa Nasir-Moin, Naofumi Tomita, et al. 2021. Learn like a pathologist: curriculum learning by annotator agreement for histopathology image classification. I...

    2021

  37. [45]

    Dejing Xu, Zhou Zhao, Jun Xiao, Fei Wu, Hanwang Zhang, Xiangnan He, and Yueting Zhuang. 2017. Video question answering via gradually refined attention over appearance and motion. In Proceedings of the 25th ACM international conference on Multimedia, pages 1645--1653

    2017

  38. [46]

    Kelvin Xu, Jimmy Ba, Ryan Kiros, Kyunghyun Cho, Aaron Courville, Ruslan Salakhudinov, Rich Zemel, and Yoshua Bengio. 2015. Show, attend and tell: Neural image caption generation with visual attention. In International conference on machine learning, pages 2048--2057. PMLR

    2015

  39. [47]

    Changchang Yin, Buyue Qian, Jishang Wei, Xiaoyu Li, Xianli Zhang, Yang Li, and Qinghua Zheng. 2019. Automatic generation of medical imaging diagnostic report with hierarchical recurrent neural network. In 2019 IEEE international conference on data mining (ICDM), pages 728--737. IEEE

    2019

  40. [48]

    Lu Yuan, Dongdong Chen, Yi-Ling Chen, Noel Codella, Xiyang Dai, Jianfeng Gao, Houdong Hu, Xuedong Huang, Boxin Li, Chunyuan Li, et al. 2021. Florence: A new foundation model for computer vision. arXiv preprint arXiv:2111.11432

    2021 arXiv

  41. [49]

    Xuying Zhang, Xiaoshuai Sun, Yunpeng Luo, Jiayi Ji, Yiyi Zhou, Yongjian Wu, Feiyue Huang, and Rongrong Ji. 2021. Rstnet: Captioning with adaptive attention on visual and non-visual words. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pag...

    2021

  42. [50]

    Guosheng Zhao, Zijian Zhao, Wuxian Gong, and Feng Li. 2023. Radiology report generation with medical knowledge and multilevel image-report alignment: A new method and its verification. Artificial Intelligence in Medicine, 146:102714

    2023

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.