{"id":"14fc0721-533b-4053-b267-0a57b1b35c3c","arxiv_id":"2605.29195","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Proposes TOF-decomp ADMM that splits fast- and slow-CTR log-likelihood terms under a constraint to balance contributions and enable improved contrast-noise trade-offs via early stopping in multi-kernel TOF-PET reconstruction.","lead":"The paper proposes an ADMM-based algorithm called TOF-decomp ADMM for image reconstruction in multi-kernel TOF-PET systems that mix fast and slow timing resolutions. A smart generalist might read it because it addresses how to better use emerging detector technologies that combine different photon timing signals in medical imaging.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Central claim rests on perfect per-event kernel labeling, feasible only in simulation","rationale":"Reader correctly flagged the labeling assumption as weakest; with full text the same assumption remains the single point whose violation would nullify the balancing mechanism and the reported early-stopping benefit. No other internal inconsistency appears in the abstract-level description of the ADMM split.","tokens_in":1701,"tokens_out":313,"duration_ms":17207,"concrete_test":"Re-run the brain and IQ phantom simulations with 5 % and 15 % random kernel-label flips; recompute the contrast-noise curves at the iteration chosen by the paper’s early-stopping rule. If the reported improvement over conventional ADMM shrinks by more than 30 % at either error rate, the headline claim does not survive realistic labeling uncertainty.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The TOF-decomp ADMM splits fast- and slow-CTR log-likelihood terms under an explicit constraint, but this construction presupposes that every coincidence event has already been assigned its correct kernel (fast vs. slow CTR). The abstract states the method is proposed “assuming that the events are labeled with the appropriate kernels,” and the simulations are generated with oracle labels. No section demonstrates how such labels would be obtained from real detector data, nor quantifies degradation when labels contain even modest error. Because the claimed balancing of convergence speeds and the early-stopping contrast-noise gain both derive directly from the split, any failure of the labeling step renders the balancing inoperative.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes TOF-decomp ADMM, an alternating direction method of multipliers algorithm for multi-kernel TOF-PET reconstruction. Assuming per-event labels assigning each coincidence to either a fast or slow CTR kernel are available, the method decomposes the log-likelihood into separate fast- and slow-CTR terms that are optimized independently under an explicit consistency constraint. This decomposition is claimed to balance the differing convergence rates induced by the two CTR components, permitting early stopping at iterations that improve the contrast-noise trade-off relative to standard methods. Validation consists of brain and image-quality phantom simulations that reportedly demonstrate more stable convergence and better contrast-noise characteristics.","tokens_in":1810,"tokens_out":521,"duration_ms":19550,"significance":"If the kernel-labeling step can be performed reliably on measured data, the approach would provide a practical way to exploit detectors that mix Cherenkov and scintillation timing information. The simulation evidence of stabilized convergence supplies an initial indication that the split formulation can mitigate the convergence imbalance inherent to multi-kernel data; however, the absence of quantitative metrics and real-data experiments limits the immediate clinical or technical impact.","major_comments":[{"comment":"Abstract: The central claim that the split enables 'early stopping at iterations that yield improved contrast-noise trade-offs' rests entirely on the assumption that 'events are labeled with the appropriate kernels.' The manuscript supplies neither an algorithm for obtaining these labels from real detector signals nor any sensitivity analysis showing how label errors degrade the balancing property. Because the simulations use oracle labels, the reported stabilization is not shown to survive the labeling step that would be required in practice.","section":"Abstract"},{"comment":"Validation description (Abstract and Results): The abstract asserts 'improved contrast-noise characteristics from a more stabilized convergence' yet reports no numerical values (contrast-recovery coefficients, standard deviation of background, iteration numbers at stopping, or statistical comparisons against conventional ADMM). Without these quantities or error analysis, the magnitude and reproducibility of the claimed benefit cannot be assessed.","section":"Abstract / Results"}],"minor_comments":[{"comment":"The manuscript should clarify whether the constraint in the ADMM formulation is enforced exactly or approximately and how the penalty parameter is chosen.","section":"Methods"}],"recommendation":"major_revision","confidential_remarks":"The work is presented as a simulation-only proof-of-concept; the journal may wish to consider whether a methods paper whose key practical step remains unaddressed fits the scope without additional real-data or robustness results."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the detailed review and constructive comments. Our responses to the major comments are provided below. The work is presented under the explicit assumption of available event labels, as stated throughout the manuscript, and focuses on the reconstruction algorithm rather than the upstream labeling process.","responses":[{"response":"The manuscript explicitly frames the TOF-decomp ADMM under the assumption that per-event kernel labels are available, as stated in the abstract and methods. The contribution is the constrained decomposition that balances convergence rates of the fast- and slow-CTR terms once labels are given. We do not provide or claim a labeling algorithm, which would be a separate signal-processing task. The oracle-label simulations demonstrate the potential benefit of the split formulation; we agree that label-error sensitivity is an important practical consideration and will expand the discussion section to address this limitation and outline possible labeling strategies based on timing-signal features.","revision_made":"partial","referee_comment":"[Abstract] Abstract: The central claim that the split enables 'early stopping at iterations that yield improved contrast-noise trade-offs' rests entirely on the assumption that 'events are labeled with the appropriate kernels.' The manuscript supplies neither an algorithm for obtaining these labels from real detector signals nor any sensitivity analysis showing how label errors degrade the balancing property. Because the simulations use oracle labels, the reported stabilization is not shown to survive the labeling step that would be required in practice."},{"response":"We acknowledge that the current validation relies primarily on visual inspection of convergence behavior and image quality in the presented figures. To improve quantitative assessment, the revised manuscript will include tables reporting contrast-recovery coefficients, background standard deviations, and the specific iteration numbers selected for early stopping, together with direct numerical comparisons against conventional ADMM on the same simulated datasets.","revision_made":"yes","referee_comment":"[Abstract / Results] Validation description (Abstract and Results): The abstract asserts 'improved contrast-noise characteristics from a more stabilized convergence' yet reports no numerical values (contrast-recovery coefficients, standard deviation of background, iteration numbers at stopping, or statistical comparisons against conventional ADMM). Without these quantities or error analysis, the magnitude and reproducibility of the claimed benefit cannot be assessed."}],"tokens_in":1421,"tokens_out":510,"duration_ms":20228,"standing_objections":["Development and validation of a reliable per-event kernel labeling algorithm from measured detector signals","Experimental validation using real (non-simulated) measured TOF-PET data"]},"desk_editor":{"model":"grok-4.3","letter":"The core idea is an ADMM variant that splits the log-likelihood into separate fast-CTR and slow-CTR terms and optimizes them under a constraint. This is presented as a way to fix the convergence imbalance that comes from mixing kernels in detectors that combine Cherenkov and scintillation light.\n\nWhat the work actually does is show, in brain and phantom simulations, that the split lets you stop at an earlier iteration and get a better contrast-noise curve than a standard joint optimization. The math on the splitting itself looks straightforward and the simulations line up with the stated goal.\n\nThe limitation is the assumption that every event already carries the correct kernel label. The abstract is explicit about this: the method is proposed “assuming that the events are labeled with the appropriate kernels,” and the reported results use oracle labels generated in simulation. No part of the work shows how those labels would be obtained from real detector signals or measures what happens when the labels contain even modest error. Because the claimed gain comes directly from the split, any labeling noise would remove the balancing effect.\n\nThis paper is aimed at people who already work on iterative reconstruction for next-generation TOF-PET hardware. A referee could usefully check the ADMM derivation and push for either a labeling procedure or a sensitivity study on label error. It is narrow enough that it does not need to be rejected outright, but the practical scope is limited by the simulation-only validation.","headline":"The paper gives a clean ADMM split for balancing fast and slow CTR terms in multi-kernel TOF-PET, but the whole approach only works under oracle event labeling that exists only in simulation.","tokens_in":2325,"tokens_out":369,"would_cite":false,"duration_ms":14144,"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":"TOF-decomp ADMM splits fast- and slow-CTR log-likelihood terms under a constraint to balance their contributions in multi-kernel TOF-PET reconstruction.","keywords":["TOF-PET","image reconstruction","ADMM","multi-kernel","coincidence time resolution","iterative reconstruction","positron emission tomography"],"falsifier":"A head-to-head run on the same phantom data where standard ADMM and TOF-decomp ADMM are stopped at the same early iteration number and the contrast-noise curve of the proposed method is compared directly to the conventional curve.","tokens_in":2603,"feed_emoji":"","tokens_out":649,"duration_ms":19696,"temperature":0.7,"pith_summary":"The paper proposes an alternating direction method of multipliers called TOF-decomp ADMM for image reconstruction in time-of-flight positron emission tomography when detectors produce events from multiple coincidence time resolution components. It assumes events carry correct kernel labels and splits the fast-CTR and slow-CTR log-likelihood terms so each can be optimized separately while linked by a constraint. This explicit balancing counters the fact that faster timing components converge more quickly than slower ones in standard joint optimization. A reader would care because the split allows stopping the iteration early at points where contrast-to-noise ratio is higher than what conventional joint methods achieve at the same step.","feed_headline":"ADMM splits fast and slow timing terms for better multi-kernel TOF-PET","feed_subtitle":"Separate optimization of CTR components under a constraint stabilizes convergence and raises contrast-noise ratio at early stopping points.","key_machinery":"TOF-decomp ADMM, which splits the fast- and slow-CTR log-likelihood terms and optimizes them separately under a constraint to balance their contributions.","core_discovery":"The TOF-decomp ADMM explicitly balances the contributions of fast- and slow-CTR components by splitting their log-likelihood terms and optimizing them separately under a constraint. This strategy addresses the convergence imbalance inherent to multi-kernel TOF-PET and enables early stopping at iterations that yield improved contrast-noise trade-offs compared with conventional methods, as shown in brain and image quality phantom simulations that demonstrate more stabilized convergence.","pith_inferences":["The same splitting idea could be tested on data sets containing three or more distinct CTR kernels to check whether the balancing effect scales.","If accurate kernel labels are available from hardware, the method may shorten total reconstruction time in settings where full convergence is computationally expensive.","Phantom results suggest the technique could be applied to other iterative PET algorithms that suffer from heterogeneous timing statistics."],"forward_implications":["Improved contrast-noise characteristics result from more stabilized convergence.","Early stopping becomes viable at iterations that already deliver better trade-offs than full runs of conventional methods.","The approach supplies a framework for using timing information from detectors that mix Cherenkov and scintillation photons.","The imbalance between fast and slow components is removed by separate optimization under the linking constraint."],"fun_headline_variants":["ADMM splits fast and slow CTR terms in TOF-PET","Balancing CTR components with ADMM in multi-kernel TOF-PET","ADMM separates CTR log-likelihood optimization in TOF-PET","TOF-decomp ADMM splits timing kernels for PET reconstruction"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The method requires that each detected event is correctly labeled with its timing kernel.","fun_headline_variants_meta":{"raw":{"variants":["ADMM splits fast and slow CTR terms in TOF-PET","Balancing CTR components with ADMM in multi-kernel TOF-PET","ADMM separates CTR log-likelihood optimization in TOF-PET","TOF-decomp ADMM splits timing kernels for PET reconstruction"]},"model":"grok-4.3","cost_usd":0.005656,"raw_usage":{"total_tokens":2698,"prompt_tokens":656,"num_sources_used":0,"completion_tokens":68,"cost_in_usd_ticks":56562000,"prompt_tokens_details":{"text_tokens":656,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1974,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":656,"tokens_out":68,"duration_ms":13710,"temperature":1.0,"reasoning_tokens":1974,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-29T00:05:31.112270+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A head-to-head run on the same phantom data where standard ADMM and TOF-decomp ADMM are stopped at the same early iteration number and the contrast-noise curve of the proposed method is compared directly to the conventional curve.","supporting_citations":[],"review_version":1}