Pith. sign in

REVIEW 3 major objections 5 minor 37 references

CoC: Chain-of-Cancer based on Cross-Modal Autoregressive Traction for Survival Prediction

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that combining four data modalities—gene expression, DNA methylation, global and local pathology images, and handcrafted language prompts—through a chain-like autoregressive fusion mechanism improves cancer survival…

desk verdict Novel four-modality survival model with a solid ablation study, but the headline SOTA claim hinges on an undescribed feature-reduction step that must be clarified before the result is believable. read the letter →

arxiv 2506.15696 v1 pith:6K57EHFH submitted 2025-05-28 cs.LG

Haipeng Zhou , Sicheng Yang , Sihan Yang , Jing Qin , Lei Chen , Lei Zhu This is my paper
classification cs.LG
keywords survivalpredictionmultimodallearningcross-modalautoregressivemodelDNAmethylationlanguageguidancewholeslideimagesC-indexcancerriskstratification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that cancer survival prediction improves when four data modalities are fused through a language-guided, autoregressive mechanism: gene expression, DNA methylation, global and local pathology images, and handcrafted text prompts. The authors claim to be the first to bring methylation and textual descriptions into survival prediction. Their Chain-of-Cancer (CoC) framework uses a CoC-Adapter to inject text guidance into each modality and an Autoregressive Mutual Traction (AMT) module to make the modalities predict one another in a causal sequence. Across five public TCGA cancer datasets, the method reaches an overall C-index of 0.655, against 0.638 for the runner-up, and separates high- and low-risk patients with log-rank p-values below 0.01. If the claim holds, epigenetic state and linguistic priors are useful, currently underused signals for patient risk stratification.

What carries the argument

The load-bearing mechanism is the Chain-of-Cancer framework's two-stage inter-learning branch. First, the CoC-Adapter fuses each modality's raw feature with a handcrafted clinical description—for example, 'An H&E stained image of {cancer} with abnormal methylation'—by concatenating feature and text-embedding and passing them through a learnable MLP. Second, the Autoregressive Mutual Traction module interleaves these text-embedded tokens into a single sequence, prepends a start token, and decodes the sequence with a two-layer causal Transformer, so each modality's representation is trained to predict the next modality's representation. A reconstruction loss and a mutual-information loss on cross-modal pairs keep the model from over-reconstructing while preserving shared information. The raw features are also kept through a separate projector as 'intra-learning' tokens, and both intra and inter tokens are concatenated and fed to the survival classifier.

What would settle it

Run CoC with the exact gene and methylation token-reduction step removed and replaced by a trivial projection (for example, randomly selected features averaged into 6 and 8 tokens); if the overall C-index stays near 0.655, the reported improvement is not attributable to the adapter and traction modules. Alternatively, if the released code shows the reduction is a standard feature-selection step, one can compare that step alone against the full model to bound its contribution.

Watch

Extended reading notes

Core claim

The central claim is that combining three clinical modalities—gene expression, methylation, and whole-slide pathology at both global and patch levels—with language guidance produces better survival risk stratification than any existing single- or dual-modal method. The authors report that CoC outperforms twelve baselines on all five datasets, with per-dataset C-index gains of 0.6%, 1.4%, 1.5%, 2.0%, and 1.3% on CESC, LIHC, BRCA, COAD, and KIRC, and an overall 0.655 versus 0.638 for the best prior method. They attribute the gain to two design choices: the CoC-Adapter, which concatenates each modality feature with a text embedding from a medical vision-language tokenizer and projects through an MLP, and the Autoregressive Mutual Traction module, which interleaves the text-embedded features and reconstructs them with a causal Transformer. The ablation study shows that removing the text prompt, using a vanilla prompt instead of tailored ones, or dropping the methylation modality each lowers the overall C-index, which the authors take as evidence that each component contributes.

Load-bearing premise

The method assumes that 60,660 gene features and 80,000 methylation features can be collapsed into just 6 and 8 learned tokens, yet the paper never says how this reduction is performed; the entire comparison with prior work may hinge on this unstated projection step.

Editorial extensions

If this is right

  • If the central claim is right, adding DNA methylation as a third clinical modality is a reliable source of survival signal that prior gene-plus-pathology models miss.
  • Handcrafted, modality-specific text prompts improve prediction over vanilla prompts, so the quality of the linguistic prior matters, not just its presence.
  • Forcing modalities to reconstruct one another autoregressively yields better risk stratification than co-attention or optimal-transport fusion used by the dual-modal baselines.
  • The gains replicate across five cancer types (CESC, LIHC, BRCA, COAD, KIRC), suggesting the mechanism generalizes rather than fitting one tumor's biology.
  • Because the encoders (ResNet-50 and SNN) are standard, the same framework could be applied to other TCGA or clinical cohorts without new feature extractors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper never describes how 60,660 gene features and 80,000 methylation features are reduced to 6 and 8 tokens; if that reduction step—rather than the traction mechanism—carries the performance, the reported gains might survive even without the autoregressive module. (Our inference from the unstated projection in Section 3.2.)
  • A natural test the authors did not run is to permute the modality order in the autoregressive sequence; if order does not matter, the 'chain' is a modeling convenience rather than a true causal structure.
  • The text prompts are handcrafted for each modality; a testable extension is to compare them with LLM-generated or learned prompts to see whether further gains come from richer linguistic priors.
  • The same adapter-plus-traction recipe could be transferred to other multimodal clinical tasks where one modality is a high-dimensional molecular profile and another is imaging or free text.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes CoC (Chain-of-Cancer), a multimodal survival prediction framework that combines gene expression, DNA methylation, whole-slide pathology at global and local scales, and handcrafted text prompts. The method has two branches: an intra-learning branch that uses raw features directly, and an inter-learning branch that injects text guidance via CoC-Adapters and then applies an Autoregressive Mutual Traction (AMT) module with reconstruction and mutual-information losses. Evaluation on five TCGA datasets (CESC, LIHC, BRCA, COAD, KIRC) reports state-of-the-art C-indexes, with an overall 0.655 versus 0.638 for the runner-up, and ablation studies support the contribution of each proposed component.

Significance. If the reported results hold, the paper would be the first to integrate methylation data and language guidance into survival prediction, and the claimed gains are clinically meaningful in magnitude. The ablation study is a genuine strength: it isolates the contribution of text prompts, the CoC-Adapter, and the AMT module, and the Kaplan-Meier analysis with log-rank p-values below 0.01 is a useful falsifiable check. The promise of code release is also a positive. However, the central quantitative claim currently rests on an undocumented dimensionality reduction of high-dimensional molecular profiles, and the headline improvements are not accompanied by statistical significance tests. These two issues make the significance of the empirical contribution conditional on revisions.

major comments (3)
  1. [Section 3.2 Implementation Details and Fig. 1] The paper does not describe how the raw 60,660-dimensional gene profile and 80,000-dimensional methylation profile are mapped to f_gene in R^{6x512} and f_meth in R^{8x512}. Section 3.2 only states that SNN is used to encode 1-D data and gives Ng=6, Nm=8. Because every variant in Table 2 shares this unstated reduction, the reported improvements over prior methods in Table 1 cannot be attributed to the proposed CoC/AMT design rather than to an undocumented, potentially survival-informative feature selection or projection. Please specify the SNN architecture, the exact feature-selection or projection procedure, normalization, and whether all preprocessing is performed per training fold only; also provide code or pseudo-code for this step.
  2. [Section 3.3, Table 1] The headline claim of outperforming all baselines (overall C-index 0.655 vs. 0.638) is not supported by any statistical significance test. The per-dataset standard deviations overlap substantially with the best baselines, e.g., CESC 0.643±0.028 vs. MOTCAT 0.637±0.035, LIHC 0.630±0.047 vs. SurvPath 0.616±0.047, and KIRC 0.709±0.048 vs. MOTCAT 0.696±0.054. Please report paired per-fold significance tests (e.g., Wilcoxon signed-rank or permutation tests) or confidence intervals for all pairwise comparisons, and define how the 'Overall' column is computed (mean of dataset means, pooled concordance, etc.).
  3. [Section 2.3, Eq. (4)] The mutual-information loss is central to the AMT module, but its notation is ambiguous and possibly incorrect. The outer sum is over pairs (m1,m2), yet the expectation uses variables x_i^{m1} and x_i^{m2} without defining how positive and negative samples are constructed for each pair. The text also says there are 'four traction chains' but then states 'a total of 6 (3×2) pairs', which is inconsistent because four modalities yield six unordered pairs. Please clarify the exact indexing, the positive/negative sampling procedure including the shuffling step, and correct the pair count.
minor comments (5)
  1. [Section 3.2 Implementation Details] Please report the exact hyperparameters used for all baseline methods; 'reproduced by their official codes' is insufficient for reproducibility.
  2. [Section 3.3 and Fig. 3] The Kaplan-Meier figure is not accompanied by the log-rank p-values or hazard ratios; please report these numbers in the text or captions.
  3. [Section 2.2 and Fig. 1] The text says the 'tokenizer from CONCH' is used, but Fig. 1 marks the text encoder as frozen; please clarify whether E_text uses the full CONCH text encoder or only a tokenizer plus a separate encoder, and whether it is frozen.
  4. [Table 2] The labels 'w/o CoC-1' and 'w/ CoC-1' are confusing; please clarify that 'w/o' means without the CoC-Adapter for that modality and 'w/' means with it.
  5. [Section 3.1 Datasets] Please specify the preprocessing and normalization used for methylation data (e.g., beta values, M values, filtering criteria) and how the 80,000 methylation features are obtained.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is trained end-to-end with a supervised survival loss plus regularizers and evaluated on held-out folds; reported gains are not forced by construction.

full rationale

The paper's derivation chain is self-contained and non-circular. The model takes raw clinical features (gene, methylation, global and local pathology) extracted by standard encoders (SNN, ResNet-50), applies a learnable CoC-Adapter and an Autoregressive Mutual Traction module, and is trained with L = L_surv + L_rec + lambda * L_MI, where L_surv is a standard negative log-likelihood survival loss. The central claim, the C-index comparison in Table 1, is obtained by 5-fold cross-validation on held-out folds, not by fitting or renaming an input quantity. The hyperparameter lambda=0.3 is an empirical constant, not derived from the test results, and the compared methods are reproduced with their official codes. The self-citations present (refs. 26, 27, 29, 34, 37) are background references to prior work on segmentation, shadow detection, and chain-of-thought prompting; none of them is invoked as a uniqueness theorem or as the justification for the survival prediction result. The paper's main weakness is an underspecified reduction from 60,660 gene and 80,000 methylation features to Ng=6 and Nm=8 tokens, which is a reproducibility and correctness-risk concern, but not a circularity: the reduction is not defined in terms of the reported C-index, and the reported prediction is not statistically forced by it. No step in the paper equates an output with an input by definition, so the appropriate circularity score is 0.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The model depends on a modest set of handpicked hyperparameters and on unstated feature-tokenization choices. No new entities are introduced.

free parameters (9)
  • lambda (weight of MI loss) = 0.3
    Empirically set; no sensitivity analysis.
  • Number of time bins for survival discretization = 4
    Chosen to convert survival to classification (Section 2.1).
  • Number of tokens for gene features (N_g) = 6
    Empirically chosen; no justification for the value.
  • Number of tokens for methylation features (N_m) = 8
    Empirically chosen; no justification.
  • Number of inter tokens (N) = 4
    Possibly the number of modalities; unclear in the paper.
  • Latent dimension d = 512
    Model capacity choice.
  • Number of transformer layers in AMT = 2
    Model capacity choice.
  • Learning rate = 1e-4
    Standard choice; not tuned.
  • Number of training epochs = 20
    Empirically set.
assumptions (5)
  • domain assumption Discretizing survival time into 4 bins and training with negative log-likelihood approximates the continuous hazard function.
    Used in Section 2.1; a standard simplification, but it loses fine-grained time resolution.
  • domain assumption The handcrafted text prompts (e.g., 'An H&E stained image of {cancer} with abnormal gene expression') provide meaningful guidance to the CONCH text encoder.
    There is no evaluation that these prompts capture the intended biological concepts; the encoder was trained on different data.
  • domain assumption The autoregressive reconstruction plus contrastive regularizer yields representations that help the downstream survival classifier.
    This is the core training hypothesis of the AMT module, validated only by internal ablations.
  • ad hoc to paper Gene and methylation profiles can be summarized into 6 and 8 tokens without losing information relevant to survival, using the SNN encoder.
    Crucial unstated assumption; the feature reduction is not described in the paper.
  • standard math Frozen foundation models (ResNet-50, CONCH, SNN) provide suitable feature spaces for all modalities.
    These are standard pretrained models treated as inputs from prior literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CoC: Chain-of-Cancer based on Cross-Modal Autoregressive Traction for Survival Prediction." pith.science (2026). https://pith.science/paper/6K57EHFH

@misc{pith2026250615696,
  author       = {Pith},
  title        = {Pith review of: CoC: Chain-of-Cancer based on Cross-Modal Autoregressive Traction for Survival Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6K57EHFH}},
  note         = {Machine review of arXiv:2506.15696}
}
read the original abstract

Survival prediction aims to evaluate the risk level of cancer patients. Existing methods primarily rely on pathology and genomics data, either individually or in combination. From the perspective of cancer pathogenesis, epigenetic changes, such as methylation data, could also be crucial for this task. Furthermore, no previous endeavors have utilized textual descriptions to guide the prediction. To this end, we are the first to explore the use of four modalities, including three clinical modalities and language, for conducting survival prediction. In detail, we are motivated by the Chain-of-Thought (CoT) to propose the Chain-of-Cancer (CoC) framework, focusing on intra-learning and inter-learning. We encode the clinical data as the raw features, which remain domain-specific knowledge for intra-learning. In terms of inter-learning, we use language to prompt the raw features and introduce an Autoregressive Mutual Traction module for synergistic representation. This tailored framework facilitates joint learning among multiple modalities. Our approach is evaluated across five public cancer datasets, and extensive experiments validate the effectiveness of our methods and proposed designs, leading to producing \sota results. Codes will be released.

Figures

Figures reproduced from arXiv: 2506.15696 by the authors.

Figure 1
Figure 1. An overview of our CoC framework. It consists of the basic feature extraction, the (a) inter-learning, and the (b) intra-learning. In (a) we use CoC-Adapter to embed language guidance to produce homogeneity features autoregressively. In (b) we use a simple projector to yield heterogeneity features. Finally, by concatenating these features we can utilize a classifier to conduct survival prediction. t via a discrete-t… view at source ↗
Figure 2
Figure 2. The Inter-Learning branch and it consists of two steps. (1) We first use CoC￾Adapter to embed the raw feature with the text guidance. (2) We deploy Autoregressive Mutual Traction (AMT) module to conduct synergistic representation. 2.2 Chain-of-Cancer Adapter Prompt methods like Chain-of-Thought (CoT) [28,13,34] demonstrate that ex￾plicitly using language prompts can enhance the model’s reasoning ability for multimod… view at source ↗
Figure 3
Figure 3. Kaplan-Meier survival curves. The prognostic separation be￾tween high-risk and low-risk cohorts are stratified by median prognostic scores. The shaded areas denote the confidence intervals. Please zoom in for the best view. Kaplan-Meier Analysis. We further conduct statistical analysis based on Kaplan-Meier analysis, and the visualization can be found in [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

37 extracted references · 22 canonical work pages

  1. [1]

    Artificial intelligence201, 81–105 (2013)

    Amores, J.: Multiple instance classification: Review, taxonomy and comparative study. Artificial intelligence201, 81–105 (2013)

    work page 2013

  2. [2]

    NIPS 13 (2000)

    Bengio, Y., Ducharme, R., Vincent, P.: A neural probabilistic language model. NIPS 13 (2000)

    work page 2000

  3. [3]

    In: ICCV

    Chen, R.J., Lu, M.Y., Weng, W.H., Chen, T.Y., Williamson, D.F., Manz, T., Shady, M., Mahmood, F.: Multimodal co-attention transformer for survival pre- diction in gigapixel whole slide images. In: ICCV. pp. 4015–4025 (2021)

    2021

  4. [4]

    Biometrika 62(2), 269–276 (08 1975)

    COX, D.R.: Partial likelihood. Biometrika 62(2), 269–276 (08 1975). https://doi.org/10.1093/biomet/62.2.269

    doi:10.1093/biomet/62.2.269 1975

  5. [5]

    Journal of the Royal Statistical So- ciety: Series B (Methodological)34(2), 187–202 (1972)

    Cox, D.R.: Regression models and life-tables. Journal of the Royal Statistical So- ciety: Series B (Methodological)34(2), 187–202 (1972)

    1972

  6. [6]

    Journal of Machine Learning Research21(85), 1–63 (2020)

    Haider, H., Hoehn, B., Davis, S., Greiner, R.: Effective ways to build and evaluate individual survival distributions. Journal of Machine Learning Research21(85), 1–63 (2020)

    work page 2020

  7. [7]

    Statistics in medicine15(4), 361–387 (1996)

    Harrell Jr, F.E., Lee, K.L., Mark, D.B.: Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Statistics in medicine15(4), 361–387 (1996)

    work page 1996

  8. [8]

    In: CVPR

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR. pp. 770–778 (2016)

    2016

Show all 37 references
  1. [9]

    In: ICML

    Ilse,M.,Tomczak,J.,Welling,M.:Attention-baseddeepmultipleinstancelearning. In: ICML. pp. 2127–2136. PMLR (2018)

    2018

  2. [10]

    In: CVPR

    Jaume, G., Vaidya, A., Chen, R.J., Williamson, D.F., Liang, P.P., Mahmood, F.: Modeling dense multimodal interactions between biological pathways and histology for survival prediction. In: CVPR. pp. 11579–11590 (2024)

    2024

  3. [11]

    Journal of the American statistical association53(282), 457–481 (1958)

    Kaplan, E.L., Meier, P.: Nonparametric estimation from incomplete observations. Journal of the American statistical association53(282), 457–481 (1958)

    1958

  4. [12]

    NIPS 30 (2017)

    Klambauer, G., Unterthiner, T., Mayr, A., Hochreiter, S.: Self-normalizing neural networks. NIPS 30 (2017)

    2017

  5. [13]

    NIPS35, 22199–22213 (2022)

    Kojima, T., Gu, S.S., Reid, M., Matsuo, Y., Iwasawa, Y.: Large language models are zero-shot reasoners. NIPS35, 22199–22213 (2022)

    2022

  6. [14]

    arXiv preprint arXiv:2406.11838 (2024) 10 H.Zhou et al

    Li, T., Tian, Y., Li, H., Deng, M., He, K.: Autoregressive image generation without vector quantization. arXiv preprint arXiv:2406.11838 (2024) 10 H.Zhou et al

    2024 arXiv

  7. [15]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Li, W., Ju, L., Tang, F., Xia, P., Xiong, X., Hu, M., Zhu, L., Ge, Z.: Towards realistic semi-supervised medical image classification. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 39, pp. 4968–4976 (2025)

    2025

  8. [16]

    Nature Medicine30, 863–874 (2024)

    Lu, M.Y., Chen, B., Williamson, D.F., Chen, R.J., Liang, I., Ding, T., Jaume, G., Odintsov, I., Le, L.P., Gerber, G., et al.: A visual-language foundation model for computational pathology. Nature Medicine30, 863–874 (2024)

    2024

  9. [17]

    Nature biomedical engineering5(6), 555–570 (2021)

    Lu, M.Y., Williamson, D.F., Chen, T.Y., Chen, R.J., Barbieri, M., Mahmood, F.: Data-efficient and weakly supervised computational pathology on whole-slide images. Nature biomedical engineering5(6), 555–570 (2021)

    2021

  10. [18]

    Nature communications11(1), 6350 (2020)

    Qiu, Y.L., Zheng, H., Devos, A., Selby, H., Gevaert, O.: A meta-learning approach for genomic survival analysis. Nature communications11(1), 6350 (2020)

    2020

  11. [19]

    In: ICML

    Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al.: Learning transferable visual models from natural language supervision. In: ICML. pp. 8748–8763. PMLR (2021)

    2021

  12. [20]

    NIPS35, 36479–36494 (2022)

    Saharia, C., Chan, W., Saxena, S., Li, L., Whang, J., Denton, E.L., Ghasemipour, K.,GontijoLopes,R.,KaragolAyan,B.,Salimans,T.,etal.:Photorealistictext-to- image diffusion models with deep language understanding. NIPS35, 36479–36494 (2022)

    2022

  13. [21]

    NIPS 34, 2136–2147 (2021)

    Shao, Z., Bian, H., Chen, Y., Wang, Y., Zhang, J., Ji, X., et al.: Transmil: Trans- former based correlated multiple instance learning for whole slide image classifica- tion. NIPS 34, 2136–2147 (2021)

    2021

  14. [22]

    British journal of cancer127(10), 1808–1815 (2022)

    Syriopoulou, E., Wästerlid, T., Lambert, P.C., Andersson, T.M.L.: Standardised survival probabilities: a useful and informative tool for reporting regression models for survival data. British journal of cancer127(10), 1808–1815 (2022)

    2022

  15. [23]

    arXiv preprint arXiv:2404.02905 (2024)

    Tian, K., Jiang, Y., Yuan, Z., Peng, B., Wang, L.: Visual autoregressive modeling: Scalable image generation via next-scale prediction. arXiv preprint arXiv:2404.02905 (2024)

    2024 arXiv

  16. [24]

    The lancet oncology3(12), 755–763 (2002)

    Verma, M., Srivastava, S.: Epigenetics in cancer: implications for early detection and prevention. The lancet oncology3(12), 755–763 (2002)

    2002

  17. [25]

    IEEE Transactions on Medical Imaging (2024)

    Wang, H., Chen, J., Zhang, S., He, Y., Xu, J., Wu, M., He, J., Liao, W., Luo, X.: Dual-reference source-free active domain adaptation for nasopharyngeal carci- noma tumor segmentation across multiple hospitals. IEEE Transactions on Medical Imaging (2024)

    2024

  18. [26]

    In: 2023 IEEE International Con- ference on Robotics and Automation (ICRA)

    Wang, H., Jin, Y., Zhu, L.: Dynamic interactive relation capturing via scene graph learning for robotic surgical report generation. In: 2023 IEEE International Con- ference on Robotics and Automation (ICRA). pp. 2702–2709. IEEE (2023)

    2023

  19. [27]

    In: Proceedings of the 32nd ACM International Conference on Multimedia

    Wang, H., Wang, W., Zhou, H., Xu, H., Wu, S., Zhu, L.: Language-driven interac- tive shadow detection. In: Proceedings of the 32nd ACM International Conference on Multimedia. pp. 5527–5536 (2024)

    2024

  20. [28]

    Advances in neural information processing systems35, 24824–24837 (2022)

    Wei, J., Wang, X., Schuurmans, D., Bosma, M., Xia, F., Chi, E., Le, Q.V., Zhou, D., et al.: Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems35, 24824–24837 (2022)

    2022

  21. [29]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Xing, Z., Ye, T., Yang, Y., Liu, G., Zhu, L.: Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 578–588. Springer (2024)

    2024

  22. [30]

    In: MICCAI

    Xiong, C., Chen, H., Zheng, H., Wei, D., Zheng, Y., Sung, J.J., King, I.: Mome: Mixture of multimodal experts for cancer survival prediction. In: MICCAI. pp. 318–328. Springer (2024) CoC: Chain-of-Cancer for Survival Prediction 11

    2024

  23. [31]

    In: ICCV

    Xu, Y., Chen, H.: Multimodal optimal transport-based co-attention transformer with global structure consistency for survival prediction. In: ICCV. pp. 21241– 21251 (2023)

    2023

  24. [32]

    IEEE TPAMI43(9), 3126–3137 (2020)

    Zadeh, S.G., Schmid, M.: Bias in cross-entropy-based training of deep survival networks. IEEE TPAMI43(9), 3126–3137 (2020)

    2020

  25. [33]

    NIPS30 (2017)

    Zaheer, M., Kottur, S., Ravanbakhsh, S., Poczos, B., Salakhutdinov, R.R., Smola, A.J.: Deep sets. NIPS30 (2017)

    2017

  26. [34]

    NIPS36, 5168– 5191 (2023)

    Zheng, G., Yang, B., Tang, J., Zhou, H.Y., Yang, S.: Ddcot: Duty-distinct chain-of- thought prompting for multimodal reasoning in language models. NIPS36, 5168– 5191 (2023)

    2023

  27. [35]

    NPJ genomic medicine5(1), 11 (2020)

    Zheng, H., Momeni, A., Cedoz, P.L., Vogel, H., Gevaert, O.: Whole slide images reflect dna methylation patterns of human tumors. NPJ genomic medicine5(1), 11 (2020)

    2020

  28. [36]

    In: ICCV

    Zhou, F., Chen, H.: Cross-modal translation and alignment for survival analysis. In: ICCV. pp. 21485–21494 (2023)

    2023

  29. [37]

    In: Proceedings of the 32nd ACM International Conference on Multimedia

    Zhou, H., Wang, H., Ye, T., Xing, Z., Ma, J., Li, P., Wang, Q., Zhu, L.: Timeline and boundary guided diffusion network for video shadow detection. In: Proceedings of the 32nd ACM International Conference on Multimedia. pp. 166–175 (2024)

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.