{"id":"5a0ba9b7-5b89-429b-847b-31ce634f1ce1","arxiv_id":"2606.24317","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":1,"one_line_summary":"Unpaired two-step DL model (DDIM + domain-adversarial U-Net) synthesizes PCCT material-basis maps from EID CT images.","lead":"The paper presents a two-step deep learning framework using a diffusion model followed by a domain-adversarial U-Net to synthesize photon-counting CT water and iodine maps from standard energy-integrating CT scans without paired training data. This approach could enable generation of large synthetic PCCT datasets from abundant public EID CT collections for algorithm development.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Domain shift between DDIM-generated EID and real EID may not be small enough for the adversarial U-Net to produce accurate material maps on unpaired real scans","rationale":"The load-bearing assumption identified by the reader is exactly the point where the unpaired training pipeline is least secured; the abstract-only review already flagged it, and the method description confirms there is no direct paired real-data validation to test it.","tokens_in":1807,"tokens_out":388,"duration_ms":13885,"concrete_test":"Take the trained U-Net+DANN and run it on a set of real EID abdominal scans; derive 40 keV and 70 keV VMIs and compare their MTF curves and NPS to both the original EID images and to any available real PCCT VMIs from the same or matched patients. If the reported resolution gain disappears or HU errors in iodine/water exceed the values reported on generated data by >10 HU, the domain-shift assumption fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (feasibility shown by higher-resolution 40/70 keV VMIs that preserve EID structures/textures) depends on the second-stage U-Net+ DANN, trained exclusively on DDIM-generated EID images paired with PCCT-derived water/iodine maps, generalizing to real EID inputs. The DDIM step produces synthetic EID from PCCT; any residual mismatch in noise texture, spectral response, or beam-hardening not captured by the diffusion model creates a domain gap. The domain-adversarial loss is intended to close this gap, but without paired real EID-PCCT data the only evidence is indirect (MTF/NPS/HU on the generated domain or qualitative appearance). If the gap remains, material maps on real EID will be inaccurate even if the adversarial term is minimized on the training distribution.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript presents a two-step unpaired deep learning pipeline to synthesize photon-counting CT (PCCT) material-basis maps (water and iodine) from energy-integrating detector (EID) CT images. A DDIM first generates synthetic EID images from PCCT data; a domain-adversarial U-Net then predicts the material maps from these generated EID images. Virtual monoenergetic images (VMIs) at 40 and 70 keV are reconstructed from the predicted maps and evaluated via HU accuracy, MTF, NPS, and qualitative appearance, with the central claim that the resulting VMIs exhibit higher spatial resolution while preserving anatomical structures and textures of the original EID images.","tokens_in":2007,"tokens_out":538,"duration_ms":13830,"significance":"If the generalization to real unpaired EID data holds, the method could enable creation of large synthetic PCCT training sets from existing EID archives, supporting algorithm development where paired PCCT-EID data are unavailable. The two-stage design (diffusion-based domain translation followed by adversarial material decomposition) directly targets the unpaired-data constraint.","major_comments":[{"comment":"Results section (and abstract): All reported quantitative metrics (HU accuracy, MTF, NPS) and the qualitative preservation claim are demonstrated exclusively on DDIM-generated EID images paired with PCCT-derived material maps. No quantitative or qualitative results are supplied for the U-Net+DANN applied to real unpaired EID CT scans, so the load-bearing assumption that the domain-adversarial loss closes the gap between synthetic and real EID distributions remains untested on the target domain.","section":"Results/Evaluation"},{"comment":"Methods (second-stage network): The domain-adversarial loss is described as mitigating the domain shift, yet no ablation is presented that isolates its contribution (e.g., U-Net trained with vs. without the adversarial term) or that quantifies residual mismatch in noise texture or beam-hardening between DDIM outputs and real EID images; without such evidence the feasibility claim for real EID inputs rests on an unverified assumption.","section":"Methods"}],"minor_comments":[{"comment":"Abstract: 'clinc' is a typographical error for 'clinic'.","section":"Abstract"},{"comment":"Abstract: The evaluation metrics are listed but no numerical values, confidence intervals, or statistical comparisons are supplied, reducing the abstract's informativeness.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive and detailed comments. We address each major point below and indicate where revisions will be made to strengthen the manuscript.","responses":[{"response":"We agree that the reported quantitative metrics rely on DDIM-generated EID images paired with PCCT-derived material maps as ground truth, which is feasible only in this controlled setting. Real unpaired EID scans lack corresponding material maps, precluding the same quantitative evaluation. The domain-adversarial loss is intended to bridge the distribution gap, but we acknowledge that direct evidence on real EID inputs would better support the generalization claim. In the revised manuscript we will add qualitative examples of material-basis maps and VMIs produced from real unpaired EID CT scans to illustrate applicability to the target domain.","revision_made":"yes","referee_comment":"[Results/Evaluation] Results section (and abstract): All reported quantitative metrics (HU accuracy, MTF, NPS) and the qualitative preservation claim are demonstrated exclusively on DDIM-generated EID images paired with PCCT-derived material maps. No quantitative or qualitative results are supplied for the U-Net+DANN applied to real unpaired EID CT scans, so the load-bearing assumption that the domain-adversarial loss closes the gap between synthetic and real EID distributions remains untested on the target domain."},{"response":"We agree that an ablation isolating the domain-adversarial term and additional quantification of residual domain mismatch would strengthen the methods section. In the revised manuscript we will include an ablation study comparing the U-Net trained with and without the adversarial loss, together with analysis of noise texture and beam-hardening characteristics between DDIM outputs and real EID images.","revision_made":"yes","referee_comment":"[Methods] Methods (second-stage network): The domain-adversarial loss is described as mitigating the domain shift, yet no ablation is presented that isolates its contribution (e.g., U-Net trained with vs. without the adversarial term) or that quantifies residual mismatch in noise texture or beam-hardening between DDIM outputs and real EID images; without such evidence the feasibility claim for real EID inputs rests on an unverified assumption."}],"tokens_in":1538,"tokens_out":469,"duration_ms":22334,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The core idea is to bootstrap PCCT training data by first using DDIM to turn real PCCT scans into synthetic EID images, then training a U-Net with DANN on those pairs to predict water and iodine maps that can later be applied to real EID scans. The target is virtual monoenergetic images at 40 and 70 keV. This is a reasonable response to the lack of public PCCT datasets.\n\nWhat is actually new is the specific chaining of DDIM unpaired translation with domain-adversarial training for material decomposition in abdominal CT. Prior work on diffusion models or DANN exists, but the medical application to PCCT basis maps from unpaired EID appears fresh.\n\nThe paper does a clean job laying out the pipeline and naming the right evaluation axes (HU error, MTF, NPS, visual texture). The motivation is practical and the unpaired constraint is respected.\n\nThe main weakness is the absence of any quantitative results in the abstract. It states that the reconstructed VMIs show higher resolution while keeping EID structures, yet gives no values, error bars, or direct comparison to real paired PCCT. The stress-test concern about residual domain shift between DDIM-generated EID and real EID is therefore unaddressed by numbers; if the adversarial term does not fully close the gap in noise texture or beam hardening, the material maps on real EID could still be off. Without paired validation data the claim rests on indirect evidence.\n\nThis is for groups building spectral CT algorithms who need synthetic data now. A reader already working on diffusion-based medical translation might pick up the two-stage trick, but anyone expecting a validated method will be disappointed. The work shows clear thinking about the data scarcity problem and honest use of existing tools, so it deserves a serious referee even if heavy revision is likely.","headline":"The paper outlines a two-step unpaired pipeline (DDIM then domain-adversarial U-Net) to make synthetic PCCT material maps from EID CT, but supplies no numbers to show the maps are accurate on real unpaired scans.","tokens_in":2459,"tokens_out":461,"would_cite":false,"duration_ms":10265,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"A two-step deep learning pipeline can synthesize photon-counting CT material maps from unpaired standard EID CT scans.","keywords":["photon-counting CT","material basis maps","unpaired learning","DDIM","domain-adversarial U-Net","virtual monoenergetic images","energy-integrating detector CT","spectral CT synthesis"],"falsifier":"Quantitative comparison on any set of real paired PCCT and EID scans showing large errors in Hounsfield unit values for the predicted water or iodine maps or failure to recover the claimed spatial-resolution gain in the 40 and 70 keV reconstructions.","tokens_in":2716,"feed_emoji":"🩻","tokens_out":724,"duration_ms":13833,"temperature":0.7,"pith_summary":"Photon-counting CT provides better spatial resolution and spectral information than conventional energy-integrating detector CT, but public PCCT datasets are scarce because the technology is new. This paper presents a method that first uses a denoising diffusion implicit model to translate between PCCT and EID image domains, then trains a domain-adversarial U-Net on the generated EID images to predict water and iodine material maps. The resulting maps are used to reconstruct virtual monoenergetic images at 40 and 70 keV. Evaluation shows these images maintain anatomical detail from the original EID scans while gaining resolution, indicating the approach works without any paired PCCT-EID training examples.","feed_headline":"Unpaired model turns standard CT into photon-counting material maps","feed_subtitle":"DDIM plus domain-adversarial U-Net predicts water and iodine maps from public EID scans without any paired training examples.","key_machinery":"Two-step unpaired pipeline: DDIM generation of EID-style images from PCCT followed by domain-adversarial U-Net material decomposition into water and iodine maps.","core_discovery":"The proposed framework provides a feasible approach for synthesizing PCCT spectral material-basis images from conventional EID CT without requiring paired images by using DDIM to generate EID CT images from PCCT followed by a domain-adversarial U-Net to predict water and iodine maps, with reconstructed 40 and 70 keV images showing higher spatial resolution while preserving anatomical structures and textures of the original EID CT images.","pith_inferences":["The same unpaired translation strategy could be tested on other emerging CT modalities that also lack paired reference data.","Once real paired PCCT-EID cases appear in limited numbers they could serve as a direct validation set for the synthetic maps.","Downstream PCCT reconstruction algorithms could be pretrained on the generated material maps before fine-tuning on scarce real data."],"forward_implications":["Large synthetic PCCT material-map datasets become available for algorithm development without waiting for paired clinical acquisitions.","Reconstructed virtual monoenergetic images at 40 and 70 keV retain original EID textures while exhibiting measurably higher spatial resolution.","The method supplies training and evaluation data for PCCT-specific tasks in environments where real photon-counting scanners remain limited.","Water and iodine basis maps can be obtained directly from existing public EID CT collections."],"fun_headline_variants":["Unpaired DL synthesizes PCCT material maps from EID CT","DDIM plus domain-adversarial U-Net predicts water and iodine maps","Synthesizes photon-counting CT maps from unpaired EID abdominal CT","Two-step model generates PCCT basis images without paired examples"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The distribution shift between DDIM-generated EID images and real EID images is small enough that a network trained on the generated images will still produce accurate material maps on real unpaired EID scans.","fun_headline_variants_meta":{"raw":{"variants":["Unpaired DL synthesizes PCCT material maps from EID CT","DDIM plus domain-adversarial U-Net predicts water and iodine maps","Synthesizes photon-counting CT maps from unpaired EID abdominal CT","Two-step model generates PCCT basis images without paired examples"]},"model":"grok-4.3","cost_usd":0.011968,"raw_usage":{"total_tokens":5198,"prompt_tokens":768,"num_sources_used":0,"completion_tokens":73,"cost_in_usd_ticks":119678000,"prompt_tokens_details":{"text_tokens":768,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":4357,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":768,"tokens_out":73,"duration_ms":30762,"temperature":1.0,"reasoning_tokens":4357,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-25T21:38:56.030609+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Quantitative comparison on any set of real paired PCCT and EID scans showing large errors in Hounsfield unit values for the predicted water or iodine maps or failure to recover the claimed spatial-resolution gain in the 40 and 70 keV reconstructions.","supporting_citations":[],"review_version":1}