{"id":"f2b8401c-c5c2-4226-93e7-c9c528355108","arxiv_id":"2606.20757","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"EMMS uses evidential fusion based on Dempster-Shafer theory to handle missing modalities in multimodal survival prediction without generative imputation, reporting SOTA results and calibrated uncertainty on four cancer datasets.","lead":"The paper proposes EMMS, a model that fuses multimodal data for cancer survival prediction by treating missing modalities as vacuous evidence via Dempster-Shafer theory and Gaussian Random Fuzzy Numbers. A smart generalist might read it because incomplete data is common in real clinical settings and this approach avoids imputation while reporting uncertainty.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"Reader correctly isolated the vacuous-evidence assumption as the key unproven step from the abstract alone. With full text unavailable in the query, no further internal inconsistency can be located; the standard DST neutrality property makes the assumption plausible pending the concrete definitions in the method section.","tokens_in":1695,"tokens_out":264,"duration_ms":22225,"concrete_test":"Extract the exact definition of the frame of discernment, the GRFN representation of a survival output, and the combination rule used in §3–4; verify that the vacuous mass assignment leaves the fused result identical to the observed-modality result and that the resulting uncertainty measure increases monotonically with the number of missing modalities.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on standard DST properties (vacuous belief is the neutral element under Dempster combination) together with GRFN modeling of per-modality survival evidence. The abstract states that missing modalities are assigned vacuous evidence, which by construction does not alter the combined mass from observed modalities and increases total uncertainty. No internal contradiction with this construction appears in the given description, and the claimed experiments on four datasets are presented as direct empirical support for calibration under missing-modality regimes.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript introduces the Evidential Missing Modality Survival Fusion (EMMS) model for multimodal survival prediction. It fuses per-modality evidence using Dempster-Shafer theory combined with Gaussian Random Fuzzy Numbers, explicitly models both aleatoric and epistemic uncertainty plus modality reliability, and assigns vacuous evidence to missing modalities so that they do not alter the combined mass from observed modalities while increasing total uncertainty. Experiments on four cancer datasets are reported to show state-of-the-art predictive performance together with calibrated uncertainty estimates and no extra computational cost relative to complete-modality baselines.","tokens_in":1765,"tokens_out":405,"duration_ms":22233,"significance":"If the empirical claims hold, the work supplies a direct, non-generative fusion rule that exploits the algebraic neutrality of vacuous belief under Dempster combination. This yields a computationally lightweight method for incomplete multimodal survival data that simultaneously reports calibrated uncertainty, which is a practical advantage in clinical settings where modality dropout is routine.","major_comments":[{"comment":"Abstract and experimental section: the central claim of state-of-the-art performance and calibrated uncertainty under missing-modality regimes is asserted without any description of the competing baselines, the statistical tests employed, the train/validation/test splits, or the precise handling of right-censoring; these omissions make the effectiveness claim unverifiable from the available text and are load-bearing for the paper's primary contribution.","section":"Abstract"}],"minor_comments":[{"comment":"Clarify the precise definition and parameterization of Gaussian Random Fuzzy Numbers when they are first introduced; the current description leaves the mapping from per-modality survival outputs to mass functions underspecified.","section":null},{"comment":"Add a short paragraph contrasting the proposed vacuous-evidence rule with standard imputation or modality-dropout baselines to highlight the claimed computational advantage.","section":null}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the detailed review and constructive suggestion to improve verifiability of the empirical claims. We address the comment below and will revise the manuscript accordingly.","responses":[{"response":"We agree that the abstract, as a concise summary, omits these specifics, and that the experimental section should make the evaluation protocol fully explicit to support the SOTA and calibration claims. In the revised manuscript we will (i) expand the abstract with a brief clause listing the main baselines (e.g., the complete-modality and missing-modality variants of the compared multimodal survival models), (ii) add a dedicated paragraph in the experimental section that enumerates the competing methods, reports the exact train/validation/test splits (including any cross-validation scheme), describes the statistical tests used for significance (e.g., paired Wilcoxon or log-rank tests with p-values), and details the right-censoring handling (standard negative log-likelihood under the Cox partial likelihood with censoring indicators). These additions will render the effectiveness claims directly verifiable while preserving the paper's length and focus.","revision_made":"yes","referee_comment":"[Abstract] Abstract and experimental section: the central claim of state-of-the-art performance and calibrated uncertainty under missing-modality regimes is asserted without any description of the competing baselines, the statistical tests employed, the train/validation/test splits, or the precise handling of right-censoring; these omissions make the effectiveness claim unverifiable from the available text and are load-bearing for the paper's primary contribution."}],"tokens_in":1250,"tokens_out":328,"duration_ms":12854,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The central move is to assign vacuous evidence to missing modalities so the combination rule leaves the mass from observed inputs unchanged while automatically raising total uncertainty. This is presented as a direct, non-generative alternative for clinical survival models.\n\nWhat is actually new is the specific pairing of Dempster-Shafer fusion with Gaussian Random Fuzzy Numbers for per-modality survival evidence, plus the explicit use of vacuous belief to handle incomplete multimodal inputs in a survival setting. The approach builds on standard properties of the theory rather than inventing new combination rules.\n\nThe paper does well on the practical side: it keeps computation light, sidesteps the need to impute or generate missing scans or records, and ties uncertainty calibration to the evidential framework. The stress-test note correctly notes that the vacuous-evidence construction introduces no internal contradiction.\n\nThe soft spots are in the experimental claims. The abstract asserts state-of-the-art results on four cancer datasets and calibrated uncertainty, yet supplies no information on baseline methods, concordance-index handling of censoring, dataset splits, or statistical tests. Without those details the performance numbers cannot be evaluated, so the effectiveness argument rests on the theoretical construction alone until the full results are checked.\n\nThis is for researchers who build multimodal survival predictors and need a lightweight way to cope with missing modalities. It deserves a serious referee because the problem is common, the proposed fix is simple and grounded, and the only way to settle the empirical side is through review.","headline":"EMMS fuses multimodal survival predictions by treating missing modalities as vacuous evidence under Dempster-Shafer theory with Gaussian Random Fuzzy Numbers, avoiding any generative step for absent data.","tokens_in":2239,"tokens_out":377,"would_cite":false,"duration_ms":18171,"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":"EMMS fuses multimodal survival data by treating missing modalities as vacuous evidence within a Dempster-Shafer framework.","keywords":["multimodal survival prediction","missing modalities","evidential fusion","Dempster-Shafer theory","uncertainty estimation","cancer datasets","Gaussian Random Fuzzy Numbers"],"falsifier":"Run the model on the same cancer datasets but with increasing fractions of modalities removed at random and check whether the reported uncertainty rises in proportion to the actual rise in prediction error; persistent under- or over-estimation of uncertainty would falsify the central claim.","tokens_in":2590,"feed_emoji":"🧬","tokens_out":616,"duration_ms":19520,"temperature":0.7,"pith_summary":"The paper presents a model that performs survival prediction from multiple clinical data sources even when some sources are absent. It combines predictions from available modalities through an evidential fusion process that accounts for different kinds of uncertainty and the trustworthiness of each source. Missing inputs receive no weight in the combination step, which raises overall uncertainty in a controlled manner. This matters because incomplete records are routine in medical practice, and the method avoids the separate step of creating substitute values for absent data.","feed_headline":"Evidential fusion handles missing data in cancer survival prediction","feed_subtitle":"Treating absent modalities as vacuous evidence yields calibrated uncertainty without any imputation step.","key_machinery":"The evidential fusion mechanism based on Dempster-Shafer theory and Gaussian Random Fuzzy Numbers that integrates modality predictions while assigning vacuous evidence to absent modalities.","core_discovery":"The EMMS model performs multimodal survival prediction under missing modalities by employing Dempster-Shafer theory and Gaussian Random Fuzzy Numbers for decision fusion. It considers aleatoric and epistemic uncertainty along with modality reliability. Missing modalities are treated as vacuous evidence, which prevents them from interfering with available inputs and leads to increased yet calibrated uncertainty. Experiments on four cancer datasets show state-of-the-art performance with calibrated uncertainty estimates and no extra computational cost.","pith_inferences":["The same fusion rule could be applied to other multimodal clinical tasks such as diagnosis or treatment recommendation.","Clinical data collection protocols might tolerate more missing entries if this style of fusion is used downstream.","The uncertainty signal could be used to trigger additional tests only when missing data meaningfully degrades reliability."],"forward_implications":["The approach yields calibrated uncertainty estimates that increase when modalities are absent.","No separate generative model or imputation step is required for missing data.","Performance reaches state-of-the-art levels on four cancer survival datasets.","Computational cost remains comparable to models trained on complete data.","Uncertainty estimates are produced alongside the survival predictions."],"fun_headline_variants":["Evidential model for multimodal survival prediction under missing modalities","Treating absent modalities as vacuous evidence in survival analysis","Multimodal fusion using Dempster-Shafer theory and fuzzy numbers","Cancer survival prediction with calibrated uncertainty and missing data","EMMS avoids imputation by handling missing modalities as vacuous evidence"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"Treating missing modalities as vacuous evidence will prevent interference with available inputs and will automatically produce increased but calibrated uncertainty.","fun_headline_variants_meta":{"raw":{"variants":["Evidential model for multimodal survival prediction under missing modalities","Treating absent modalities as vacuous evidence in survival analysis","Multimodal fusion using Dempster-Shafer theory and fuzzy numbers","Cancer survival prediction with calibrated uncertainty and missing data","EMMS avoids imputation by handling missing modalities as vacuous evidence"]},"model":"grok-4.3","cost_usd":0.005815,"raw_usage":{"total_tokens":2736,"prompt_tokens":605,"num_sources_used":0,"completion_tokens":80,"cost_in_usd_ticks":58149500,"prompt_tokens_details":{"text_tokens":605,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2051,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":605,"tokens_out":80,"duration_ms":15509,"temperature":1.0,"reasoning_tokens":2051,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-26T18:13:17.194092+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Run the model on the same cancer datasets but with increasing fractions of modalities removed at random and check whether the reported uncertainty rises in proportion to the actual rise in prediction error; persistent under- or over-estimation of uncertainty would falsify the central claim.","supporting_citations":[],"review_version":1}