- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
The paper proposes HIDE, which corrects chest X-ray reports without updating the report generator. It extracts findings from a greedy draft and several stochastic samples, combining cross-sample support with local prototype-based visual evidence. Branch-specific predictors then authorize additions, deletions, or no change. The authors report improvements across six generators, two in-domain datasets, and transfer to IU-Xray.
Critical strengths
- The problem is important and well motivated. A finding omitted by one decode may still appear in other samples. The method also recognizes that repeated claims are not necessarily correct.
- Finding-level edits can combine useful content from different reports. Positive, normal, and confounder prototypes provide complementary evidence.
Critical weaknesses
W1. Key implementation details are missing
Predictor fitting. Section 3.4 and Appendices D/F specify the features and thresholds, but not the model used for . The target labels, fitting objective, fitting procedure, and exact definition of also need clarification. Appendix F confirms that the predictor has fitted parameters. The thresholds alone are therefore insufficient to reproduce it. Please report the actual size of the 0.5% subset and explain how fitting and threshold selection are organized.
Visual verification and missing prototype roles. Section 3.3 does not specify , , prototype counts, or how small and empty prototype groups are handled. Table B.2 states that no sufficient negative rule is defined for pleural other, so how are and in Eq. (4) computed for this category? Listing pleural effusion or atelectasis as alternatives does not establish that pleural other is absent. Please provide any additional label mapping and the fallback rule for groups with fewer than samples.
Text editing. Section 3.5 describes copying source sentences and deleting matched spans, but does not specify the editor. Does it use string operations, sentence-rewriting rules, or a language model? If a model is used, please provide its identity, prompt, and constraint checks. A source sentence may contain several findings, while only some are authorized. How does the editor avoid adding the others? Pseudocode and an edit trace would make the procedure much easier to assess.
W2. Baseline implementations and resource use need clarification
The baselines have different designs. Please document their full implementations and parameter settings. In particular, report CCD’s expert model and control parameters. For CCS, specify the utility, embedding checkpoint, candidate count, and sampling settings. Please also explain how DeCo, ICD, and VCD are adapted to the tested generators.
In addition, a matched-size or shared candidate pool would be a useful additional comparison for CCS. Reporting auxiliary data, pretrained components, and end-to-end costs would clarify the performance–resource trade-off. It is recommended to provide the above points as well.
Questions for authors
- See weaknesses.
- Suggestions: 1)The gains across models and datasets support the method's overall effectiveness. An edit-level analysis would help explain these gains. Useful statistics including action frequencies, correct and incorrect edits, and the rate of unchanged reports will help in understanding how overall benefits are generated and which cases remain at risk. 2)Please check Figure 4. There is a strange subfigure in it.
3 comments
Below is our continued response to [W1.3].
Pseudocode
For each authorized ADD or DEL:
Locate target clause c and sentence s.
If s contains only the target finding:
Copy s into the report for ADD, or remove s for DEL.
Otherwise:
ADD: retain authorized clauses and mark all other finding clauses for removal.
DEL: mark the target clause for removal and retain unrelated clauses.
Pass s and both clause lists to the frozen local editor.
Validate the returned sentence before applying the edit.
Edit trace example
| Step | Example |
|---|---|
| Sampled sentence | “Bibasilar atelectasis and a small left pleural effusion are present.” |
| Authorization | ADD bibasilar atelectasis only. |
| Editor input | ADD_SUBTRACTION · Retain “Bibasilar atelectasis” · Remove “a small left pleural effusion” |
| Sentence added | “Bibasilar atelectasis is present.” |
[W2.1] Baseline implementation details
Response: We have added the implementation to Appendix I (Lines 1316-1381). We followed the official implementations. The first table lists each method’s settings, and the second gives the candidate counts and generator-specific adaptations.
| Baseline | Implementation | Parameter settings |
|---|---|---|
| CCD | Uses the TorchXRayVision DenseNet expert. Symptom-grounded contrastive decoding (SCD) uses the expert’s symptom labels, and expert-informed contrastive decoding (ECD) adjusts token probabilities. | Expert threshold 0.5, SCD weight , ECD weight , diagnostic plausibility constraint . Fixed across generators. |
| CCS | Uses Image-Grounded Utility with the CCS authors’ CXR-adapted Qwen3-VL-Embed-2B checkpoint. CCS scores all candidate pairs and selects the report with the highest mean utility against the other reports. | Sampling temperature . The candidate count matches HIDE’s sampling budget for each generator, as listed below. |
| DeCo | At each decoding step, the generator’s native adapter selects an anchor layer from the specified range and uses its logits to calibrate the final-layer logits. | Top-, correction strength . Anchor-layer indices are listed below. |
| ICD | Each generator receives the same image with the original and negative prompts. ICD subtracts the weighted negative-prompt distribution, applies truncation, and samples the next token. | Contrast weight , truncation coefficient , top-, repetition penalty , beam size . |
| VCD | Contrasts the original image with a perturbed view generated by the official add_diffusion_noise function. Each generator’s visual interface is listed below. | Contrast strength , truncation threshold , noise step . |
CCS and HIDE use the same number of candidate reports for each generator. Both methods use the same fixed pool of reports for each study.
| Generator | DeCo anchor layers | VCD adaptation |
|---|---|---|
| Lingshu-7B | 20–28 | Native Qwen2.5-VL processor |
| MedGemma-4B-IT | 20–28 | Gemma3Processor, with noise applied directly to pixel_values |
| MedVLM-R1-2B | 20–28 | Native Qwen2-VL processor, preserving image_grid_thw alignment and RoPE state |
| Med-R1-2B | 20–28 | Native Qwen2-VL processor, preserving image_grid_thw alignment and RoPE state |
| LLaVA-RAD-7B | 20–28 | Native LLaVA wrapper with raw image tensors |
| CheXAgent-2-3B | 23–31 | CheXAgent visual mode with the Phi forward pass |
For CheXAgent, the DeCo layer range follows its hidden-state indexing, where hidden_states[0] contains the input embeddings. The ICD negative prompt used across generators was: “You are an unreliable medical image observer, only give a vague impression of this image and do not rely strictly on the visual content.”
[W2.2] CCS needs matched-pool comparison and resource documentation.
Response: We evaluated HIDE and CCS on MIMIC-CXR in the same controlled environment using Med-R1-2B, a shared fixed candidate pool, and the same sampling budget. HIDE uses MAIRA-2 for grounding, RAD-DINO for visual encoding, a prototype bank, and a fitted action predictor. CCS uses a pretrained RRG-Qwen-VL-Embedding selector adapted on MIMIC-CXR image–report pairs.
Appendix I reports runtime and FLOPs per study with all stages run sequentially. HIDE took 23.80 s for candidate generation, 6.32 s for parsing, 0.36 s for grounding, 0.09 s for prototype evidence, and less than 0.001 s for editing. CCS took 23.37 s for candidate generation and 1.55 s for embedding and selection. End-to-end costs were 30.57 s and FLOPs per study for HIDE, compared with 24.92 s and FLOPs for CCS.
Due to space limitations, we will continue our response in the next window.
[Q1(Suggestion 1)] Edit-level analysis is recommended to explain observed gains.
Response: To clarify the sources of HIDE’s gains, we directly matched the finding–polarity tuple of each authorized edit against the CheXbert-14 labels extracted from the physician-authored reference report.
| Dataset | Edit Type | Correct Additions | Incorrect Additions | Accuracy |
|---|---|---|---|---|
| MIMIC-CXR | Positive finding addition | 7,529 | 1,746 | 81.2% |
| MIMIC-CXR | Negative finding addition | 783 | 127 | 86.0% |
| CheXpert Plus | Positive finding addition | 158 | 27 | 85.4% |
| CheXpert Plus | Negative finding addition | 18 | 4 | 81.8% |
On MIMIC-CXR, 11,845 edits were evaluable. Positive finding additions accounted for 78.3% of them and achieved 81.2% label consistency (7,529/9,275); negative finding additions accounted for 7.7% and achieved 86.0% consistency (783/910).
On CheXpert Plus, 242 edits were evaluable. Positive finding additions accounted for 76.4% and achieved 85.4% label consistency (158/185); negative finding additions accounted for 9.1% and achieved 81.8% consistency (18/22).
These results identify positive finding additions as the largest source of label-consistent corrections and show that both positive and negative finding additions are generally reliable.
[Q2(Suggestion 2)] Check Figure 4.
Response: We confirmed that the values in this panel of Fig. 4 are correct. Med-R1-2B’s low baseline score makes a given absolute improvement larger in percentage terms.
We hope these clarifications and additional analyses address your concerns and better convey the scope and contribution of our work. We would be grateful if you would consider them in your final assessment.
We sincerely thank the reviewer for recognizing that our paper is well motivated, the importance of the research problem, the feasibility of recovering omitted findings from other samples, and the value of finding level edits and complementary prototype evidence.We have uploaded the revised paper.
The reviewer’s concerns focus on several implementation details, which we clarify and supplement below.
Weaknesses
[W1.1] Predictor Details.
Response:We have added details to Appendix D (Lines 1024-1040). We clarify this below.
[W1.2] Prototype Details.
Response: We have added details to Appendix B (Lines 901-917). We clarify this below.
[W1.3] Editing Module Details.
Response: We have added details to Appendix H (Lines 1242-1315). After locating the target clause and sentence (c,s), HIDE applies one of two rules:
Prompt
Due to space limitations, we will continue our response in the next window.