{"id":"b78aae7b-10b4-4015-bcec-88036259401f","arxiv_id":"2411.12833","paper_version":1,"verdict":"REJECT","confidence":"LOW","novelty_score":3.0,"correctness_risk":"high","formal_verification":"none","parameter_count":3,"one_line_summary":"The paper claims that fine-tuning Real-ESRGAN for X-ray images improves resolution and reduces bandwidth, but it lacks any experimental substantiation.","lead":"The paper proposes a telemedicine pipeline that downscales X-ray images for transmission and restores them with Real-ESRGAN at the receiver. It reports a PSNR of 37.35 dB but provides no dataset, training details, or code, so the claim is unverifiable.","discovery_kind":"incremental","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central PSNR superiority claim is unevaluable: Table 1 lacks a defined test split, degradation parameters, and matched baseline protocol, so the 7 dB gap over Real-ESRGAN is unsupported.","rationale":"The reader's weakest assumption identifies the same load-bearing issue: the quantitative result depends on a properly held-out test set and a realistic degradation model, and neither is established. I would sharpen this even further: before any question of real-world realism, the paper fails to show that the 37.35 dB result and the 30.23 dB baseline were computed on the same images under the same degradation and evaluation protocol. This is the minimal condition for a comparative PSNR claim, and it is entirely absent. The proposed method is described as a fine-tuned Real-ESRGAN, so the large gap over the base model is not credible without matched evaluation. There is no independent support in the paper: no code, no data, no reproducible training configuration, and no error analysis. I therefore do not see a single fix that would make the central claim stand; the verdict should remain REJECT. My recommendation is UNCHANGED relative to the reader's verdict. The concept of applying and fine-tuning Real-ESRGAN for medical image super-resolution could be a useful incremental application, but the paper as written does not substantiate its headline performance.","tokens_in":6360,"tokens_out":3212,"duration_ms":34780,"concrete_test":"Ask the authors to release the trained model weights and an exact evaluation script, then run a controlled comparison on a fixed public chest X-ray set (e.g., 100 CheXpert frontal images). Use one scripted degradation from Eq. (4): 4x downscaling, Gaussian blur with sigma 1.0, Gaussian noise with sigma 0.01, and JPEG quality 60. Evaluate the fine-tuned Real-ESRGAN against the unmodified Real-ESRGAN and bicubic interpolation on the same 100 images, reporting mean PSNR and SSIM with 95% confidence intervals. If the fine-tuned model does not beat base Real-ESRGAN by a statistically significant margin, or if the released weights do not reproduce the reported 37.35 dB on this protocol, the claimed superiority is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"For the central claim to hold, the reported 37.35 dB PSNR must come from a held-out test set evaluated under the same degradation model and baseline protocol as the 30.23 dB Real-ESRGAN figure. The paper supplies none of the required protocol: Section 2.4 only says the dataset contains 'various medical images' resized to 256x256, with no source, split, or patient-level separation; Section 2.2 lists degradation equations but gives no parameter values for blur kernels, downsampling factors, noise levels, or JPEG quality; and Table 1 gives no error bars or baseline implementation details. Because the proposed method is itself a fine-tuned Real-ESRGAN, a 7.12 dB improvement over the base model is internally surprising and most plausibly reflects mismatched evaluation conditions rather than genuine superiority. The fact that the Results and Discussion sections are nearly verbatim duplicates further undermines confidence in the reported numbers. Without a matched evaluation protocol, the central quantitative claim cannot be checked and therefore fails as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a telemedicine pipeline in which X-ray images are downscaled and JPEG-compressed on the sender side, transmitted, and restored at the receiver side by a fine-tuned Real-ESRGAN. Section 2 describes classical and high-order degradation models, an RRDB-based generator, a U-Net discriminator with spectral normalization, and a composite loss. Section 2.4 reports training on \"various medical images\" resized to 256×256 using an Adam optimizer with learning rate 1e−4. Section 3 presents one PSNR table (Table 1) claiming 37.35 dB for the proposed method versus 30.23 dB for Real-ESRGAN, 30.48 dB for traditional CNN methods, and 27.25 dB for interpolation, together with qualitative figures and loss curves. The conclusion discusses broad clinical and telemedicine benefits. The paper does not provide a dataset, a test protocol, baseline implementations, error bars, or code.","tokens_in":6565,"tokens_out":3850,"duration_ms":35588,"significance":"If the central claim were established, a restoration pipeline that recovers diagnostic-quality X-ray images from low-bandwidth transmissions would be of genuine clinical value. The authors identify a relevant telemedicine bottleneck and sensibly propose to reuse Real-ESRGAN with fine-tuning for grayscale medical images, which is a reasonable research direction. However, the claimed 7.12 dB improvement over the base Real-ESRGAN is not backed by any reproducible evaluation: there is no dataset description, no train/test split, no degradation parameter values, no baseline implementation details, no error bars, and no code or training scripts. The results and discussion sections are near-verbatim duplicates, which further undermines confidence. In its current form the manuscript is a project description rather than a validated method, and the central quantitative claim is unevaluable.","major_comments":[{"comment":"The dataset is described only as \"various medical images, including chest X-rays, CT scans, and mammograms, all resized to 256 × 256 pixels\" with no source, dataset size, patient-level separation, train/test split, or evaluation protocol. Since the sole quantitative claim is the single PSNR per method in Table 1, the reported 7.12 dB advantage of the proposed method over Real-ESRGAN cannot be interpreted without this protocol.","section":"Section 2.4 and Table 1"},{"comment":"The degradation model parameters (Gaussian blur kernel, downsampling factor, noise variance τ², JPEG quality factor q, and sinc cutoff η) are never specified, and no evidence is provided that the synthetic degradation matches real telemedicine artifacts such as variable compression, sensor noise, or transmission errors. Without matched degradation parameters applied identically to every baseline, the reported PSNR gap could arise from mismatched evaluation conditions rather than from genuine method superiority.","section":"Section 2.2, Eqs. (4)–(7)"},{"comment":"The baseline entries are undefined: \"Real-ESRGAN (BW images)\", \"Traditional CNN-Based Methods\", and \"Conventional Interpolation Methods\" need concrete implementations, hyperparameters, and evaluation conditions, together with error bars, the number of test images, and significance tests. Because the proposed method is itself a fine-tuned Real-ESRGAN, a 7.12 dB improvement over the base model is internally surprising and demands a matched comparison to be credible.","section":"Table 1 and Section 3"},{"comment":"The fine-tuning procedure, which is the proposed method's core contribution, is not specified beyond the Adam optimizer and the learning rate. The loss weights λ_L1, λ_perc, and λ_GAN in Eq. (9) are not given, nor are the batch size, number of training iterations, initialization from the pretrained Real-ESRGAN, data augmentation, or early-stopping criteria. As a result, the method is not reproducible from the text.","section":"Sections 2.3 and 2.4"},{"comment":"The PSNR comparison paragraphs are near-verbatim duplicates, and Figures 2–4 are referenced without quantitative axes, scale bars, or clear descriptions of what is being plotted. The duplicated text does not provide independent evidence for the results, and the lack of axis labels and loss-value scales makes the loss curves unverifiable.","section":"Sections 3 and 4"}],"minor_comments":[{"comment":"The phrase \"is they key determinant\" should be \"is a key determinant\", and \"lightening compression algorithm\" should be \"lightweight compression algorithm\".","section":"Abstract and Section 1"},{"comment":"The sentence \"The discriminator employs a U-Net architecture with spectral normalization, providing pixel-wise feedback...\" appears twice; one occurrence should be removed.","section":"Section 2.3"},{"comment":"The heading \"Degradation Model of high-order\" should be \"High-order degradation model\", and the notation T^n in Eq. (5) is not formally defined.","section":"Section 2.2"},{"comment":"The term \"vertical image super-resolution methods\" is never defined and is inconsistent with the rest of the paper, which discusses X-ray and medical images; it should be replaced with a precise description of the evaluated task.","section":"Section 3"},{"comment":"Figure captions lack details such as image sources, zoom regions, and axis labels; Figures 3 and 4 do not report the scale or unit of the loss values.","section":"Figures 2–4"},{"comment":"The conclusion claims that Real-ESRGAN was trained on \"diverse datasets, including low-quality X-rays, CT scans, and MRIs\", but Section 2.4 only mentions chest X-rays, CT scans, and mammograms; the broader claim is unsupported.","section":"Section 5"}],"recommendation":"reject","confidential_remarks":"This manuscript is well below the standard for a serious journal: the sole quantitative result is a single PSNR table with no dataset, split, degradation parameters, baseline details, or error bars, and the results/discussion sections are near-duplicates. The missing evaluation protocol is not a local fix; the experimental section would need to be entirely redone with a matched, reproducible benchmark. I would recommend rejection without an invitation to revise."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this paper is not ready for peer review. It describes a sensible idea — downscale X-rays for transmission, then restore them with fine-tuned Real-ESRGAN — but it provides no experimental protocol that would let anyone check the central claim. The 37.35 dB PSNR in Table 1 is the entire quantitative case, and it comes with no dataset description, no train/test split, no error bars, and no baseline implementation details.\n\nWhat the paper does well is modest but real: it correctly identifies an application domain, cites Real-ESRGAN and ESRGAN, and sketches a plausible two-sided pipeline (patient-side compression, clinician-side restoration). The degradation model in Section 2.2 (Gaussian blur, downsampling, noise, JPEG, sinc ringing) is the right kind of thing to simulate, and the idea of fine-tuning Real-ESRGAN on B/W X-rays is reasonable.\n\nThe problems are load-bearing. First, the method is described as Real-ESRGAN fine-tuned for X-rays, yet the comparison baseline is also Real-ESRGAN ('BW images'). A 7.12 dB improvement over the base model is internally surprising, and the paper gives no details — data size, training epochs, augmentation, or degradation parameters — that would make it plausible. Second, the Results section is almost word-for-word identical to the Discussion section, which suggests the results were not actually written from measurements. Third, Table 1 labels the proposed method 'Your Vertical Image Super-Resolution Method' — a leftover placeholder that indicates a rough draft. Fourth, the Conclusion is about Real-ESRGAN in general, not the paper's own fine-tuned version, so the authors overclaim what they demonstrated.\n\nNone of this is a matter of taste. The central quantitative claim is unevaluable as stated. If the authors redo the evaluation with a defined protocol, a held-out test set, matched baselines, and error bars, the idea could become a useful incremental application. As it stands, there is no evidence that any training or testing actually happened.\n\nThis paper is not for a serious referee until those gaps are filled. I would desk reject, but I would send the authors a note explaining exactly what is missing.\n\nBest,","headline":"An application note on using Real-ESRGAN for X-ray transmission that has no checkable experimental support; desk reject.","tokens_in":7104,"tokens_out":2340,"would_cite":false,"duration_ms":23250,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims a compression–restoration pipeline for X-rays that reaches 37.35 dB PSNR, beating Real-ESRGAN, CNN, and interpolation baselines.","keywords":["Real-ESRGAN","X-ray image enhancement","medical imaging","noise reduction","super-resolution","GAN","telemedicine","image transmission"],"falsifier":"Compute the same PSNR comparison on a public chest X-ray dataset where the low-resolution inputs are produced by an actual telemedicine transmission chain, including real JPEG compression at varying quality factors and real sensor noise, and check whether the fine-tuned model still beats the Real-ESRGAN baseline; if the gap shrinks to near zero or the 37.35 dB figure is not reproducible on held-out data, the central claim collapses.","tokens_in":6131,"feed_emoji":"🩻","tokens_out":5754,"duration_ms":48531,"temperature":0.7,"pith_summary":"This paper proposes a telemedicine pipeline that compresses X-ray images on the sending side and restores them on the clinician's side using a fine-tuned Real-ESRGAN model. The central claim is that adapting Real-ESRGAN to the noise and contrast characteristics of black-and-white medical images yields higher resolution and lower noise than the base model or traditional super-resolution methods. The reported Peak Signal-to-Noise Ratio is 37.35 dB, compared with 30.23 dB for Real-ESRGAN, 30.48 dB for CNN-based methods, and 27.25 dB for interpolation. If correct, the approach would make diagnostic-quality imaging practical over low-bandwidth connections and in resource-constrained settings.","feed_headline":"GAN restores compressed X-rays to 37.35 dB PSNR","feed_subtitle":"Fine-tuned Real-ESRGAN restores low-bandwidth X-ray scans, beating CNN and ESRGAN baselines in paper tests.","key_machinery":"The central machinery is Real-ESRGAN itself: a generator with Residual-in-Residual Dense Blocks (RRDBs) that stabilizes training and reconstructs high-frequency textures, paired with a U-Net discriminator using spectral normalization for pixel-wise realism feedback. Training uses a composite loss $L = \\lambda_{L1} L_{L1} + \\lambda_{perc} L_{perc} + \\lambda_{GAN} L_{GAN}$, combining pixel-wise L1 loss, VGG-based perceptual loss, and adversarial loss. A synthetic degradation model, built from Gaussian blur, downsampling, noise injection, sinc-filter ringing artifacts, and JPEG compression, supplies low-resolution training pairs meant to mimic real telemedicine transmission conditions.","core_discovery":"The paper claims that a fine-tuned Real-ESRGAN model, adapted to the specific noise and contrast behavior of black-and-white X-ray images, can restore diagnostically useful detail from heavily compressed, downscaled medical images. The proposed two-sided pipeline downsamples and JPEG-compresses X-rays on the patient side to save bandwidth and server load, then reconstructs the full resolution at the clinician side using a generator built from Residual-in-Residual Dense Blocks, a spectrally normalized U-Net discriminator, and a composite loss of L1, perceptual, and adversarial terms. The quantitative claim is that this method reaches 37.35 dB PSNR, compared with 30.23 dB for the base Real-ESRGAN, 30.48 dB for CNN-based super-resolution, and 27.25 dB for conventional interpolation, which the paper takes as evidence of superior noise reduction and detail preservation.","pith_inferences":["The paper leaves implicit that the reported 7 dB gain over the base Real-ESRGAN is unusually large for a fine-tuning step; a plausible but untested explanation is that the comparison baseline was not exposed to the same degradation distribution during training.","The viability of the pipeline depends on the synthetic degradation model matching real transmission artifacts; a natural extension, which the paper does not perform, is to test on actual compressed clinical transmissions to see whether the PSNR gap persists.","The paper claims qualitative superiority but reports no radiologist reader study; a blinded diagnostic study would be the decisive test of whether the restored images are truly 'diagnostic quality.'","The same fine-tuning recipe could be transferred to other modalities such as ultrasound or MRI as a check of generality, since their noise statistics differ from X-rays."],"forward_implications":["If the 37.35 dB result holds, telemedicine systems could transmit quarter-resolution JPEG X-rays and restore them to diagnostic quality, cutting bandwidth and server load.","Fine-tuning a general super-resolution GAN to monochrome medical noise characteristics may be enough to beat generic CNN and ESRGAN baselines in PSNR.","The pipeline is claimed to extend beyond chest X-rays to mammograms and CT scans, making it a potential general tool for medical image transmission.","The composite loss with L1, perceptual, and adversarial terms is the mechanism that preserves high-frequency detail while suppressing noise, which is the key to diagnostic usefulness."],"supporting_citations":[{"why":"Supplies the Real-ESRGAN architecture and synthetic degradation model that the paper fine-tunes for black-and-white medical images.","marker":"[1]"},{"why":"Provides the ESRGAN baseline with RRDB blocks and relativistic discriminator, which the paper compares against.","marker":"[2]"},{"why":"Introduces the perceptual and adversarial loss formulation used in the composite training objective.","marker":"[17]"},{"why":"Establishes the GAN adversarial training framework on which the generator-discriminator setup depends.","marker":"[19]"},{"why":"Defines the VGG feature-based perceptual loss used to preserve high-frequency texture in the restored images.","marker":"[6]"}],"fun_headline_variants":["GAN upscales compressed X-rays to 37.35 dB PSNR","Fine-tuned Real-ESRGAN beats CNN on X-ray recovery","Lower bandwidth, sharp X-rays via GAN super-resolution","X-ray GAN hits 37.35 dB after heavy JPEG compression"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's performance claim rests on the assumption that its synthetic degradation model, built from Gaussian blur, downsampling, noise, and JPEG compression, faithfully reproduces what actually happens to X-ray images in real telemedicine transmission, and that the reported PSNR was measured on a properly held-out test set.","fun_headline_variants_meta":{"raw":{"variants":["GAN upscales compressed X-rays to 37.35 dB PSNR","Fine-tuned Real-ESRGAN beats CNN on X-ray recovery","Lower bandwidth, sharp X-rays via GAN super-resolution","X-ray GAN hits 37.35 dB after heavy JPEG compression"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00027,"raw_usage":{"total_tokens":1671,"prompt_tokens":1035,"completion_tokens":636,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":651,"completion_tokens_details":{"reasoning_tokens":558}},"tokens_in":651,"tokens_out":636,"duration_ms":6107,"temperature":1.0,"reasoning_tokens":558,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T17:08:15.090379+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same PSNR comparison on a public chest X-ray dataset where the low-resolution inputs are produced by an actual telemedicine transmission chain, including real JPEG compression at varying quality factors and real sensor noise, and check whether the fine-tuned model still beats the Real-ESRGAN baseline; if the gap shrinks to near zero or the 37.35 dB figure is not reproducible on held-out data, the central claim collapses.","supporting_citations":[{"cited_title":"Ledig, L","cited_arxiv_id":null,"evidence_quote":"Introduces the perceptual and adversarial loss formulation used in the composite training objective."},{"cited_title":"Goodfellow, J","cited_arxiv_id":null,"evidence_quote":"Establishes the GAN adversarial training framework on which the generator-discriminator setup depends."},{"cited_title":"Johnson, A","cited_arxiv_id":null,"evidence_quote":"Defines the VGG feature-based perceptual loss used to preserve high-frequency texture in the restored images."}],"review_version":1}