arxiv: 2604.09280 · v1 · submitted 2026-04-10 · 📡 eess.IV · cs.CV

Recognition: unknown

AMO-ENE: Attention-based Multi-Omics Fusion Model for Outcome Prediction in Extra Nodal Extension and HPV-associated Oropharyngeal Cancer

Gautier H\'enique , William Le , Gabriel Dayan , Coralie Brodeur , Kristoff Nelson , Apostolos Christopoulos , Edith Filion , Phuc-Felix Nguyen-Tan

show 3 more authors

Laurent Letourneau-Guillon Houda Bahig Samuel Kadoury

Authors on Pith no claims yet

Pith reviewed 2026-05-10 17:13 UTC · model grok-4.3

classification 📡 eess.IV cs.CV

keywords extranodal extensionHPV-associated oropharyngeal canceroutcome predictionCT imagingattention-based fusionsemi-supervised segmentationradiomicsmultimodal model

0 comments

The pith

An attention-based fusion model combines CT-derived nodal features with clinical data to predict recurrence and survival in HPV-positive oropharyngeal cancer.

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

The paper builds a fully automated pipeline that starts with a semi-supervised 3D model to outline extranodal extension in lymph nodes on routine radiotherapy planning CT scans. From those outlines it pulls radiomics and deep features, trains a classifier for ENE status, and then feeds the results into a multimodal attention model that merges imaging signals with primary tumor and clinical variables. This setup is tested on hundreds of patients treated with radiation or chemoradiation to show improved forecasts of metastatic recurrence, overall survival, and disease-free survival compared with standard approaches. The work matters because extranodal extension carries strong prognostic weight yet remains hard to measure consistently by eye; automating its detection from existing scans could let clinicians refine staging and decisions without added manual effort or new tests.

Core claim

The authors describe an end-to-end system whose core is a hierarchical semi-supervised 3D segmentation stage that identifies and delineates intranodal and extranodal extension despite low contrast and annotation variability. Radiomics and learned features extracted from these segmentations feed an imaging-detected ENE grading classifier. These nodal descriptors are then combined with primary tumor characteristics inside an attention-based multi-omics fusion model that produces dynamic outcome predictions, with the entire pipeline evaluated for both prognostic value of the ENE label and superiority over baseline staging criteria on an internal patient cohort.

What carries the argument

The attention-based multi-omics fusion model, which uses learned attention weights to integrate radiomics and deep features from segmented nodal extension with clinical variables for outcome forecasting.

If this is right

The automated ENE grading adds measurable prognostic information beyond current staging systems for HPV-positive cases.
The pipeline removes the need for time-consuming manual contouring of nodal extension on planning CTs.
Fusing nodal imaging features with primary tumor and clinical data yields stronger forecasts of metastatic recurrence and survival than either source alone.
The approach lowers barriers to incorporating imaging-detected ENE into routine clinical decision-making for radiation planning.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same fusion structure could be retrained on other head-and-neck sites or imaging modalities to test whether attention-based integration generalizes.
If the model output is used to adjust radiation fields or chemotherapy intensity, follow-up studies could measure whether those adjustments change actual patient survival.
Extending the segmentation stage to include primary tumor margins might further tighten the link between automated imaging and treatment personalization.

Load-bearing premise

The semi-supervised segmentation model can reliably trace low-contrast extranodal extension boundaries even when manual labels are inconsistent, and the results from this single internal group of patients will hold for new cases without external or prospective checks.

What would settle it

An external validation set or prospective cohort in which the model's predicted ENE status shows poor agreement with expert review or its outcome forecasts no longer outperform standard clinical staging variables.

Figures

Figures reproduced from arXiv: 2604.09280 by Apostolos Christopoulos, Coralie Brodeur, Edith Filion, Gabriel Dayan, Gautier H\'enique, Houda Bahig, Kristoff Nelson, Laurent Letourneau-Guillon, Phuc-Felix Nguyen-Tan, Samuel Kadoury, William Le.

**Figure 2.** Figure 2: Overall schematic diagram of the proposed framework for CT-based [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

**Figure 3.** Figure 3: Graphical representation of the proposed attention-based modality [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

**Figure 4.** Figure 4: Qualitative results for the iENE segmentation task on radiotherapy [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗

**Figure 5.** Figure 5: ROC curves obtained for the binary classification of dichotomized [PITH_FULL_IMAGE:figures/full_fig_p020_5.png] view at source ↗

**Figure 6.** Figure 6: SHAP explanation analysis plots for the prediction of iENE- vs iENE [PITH_FULL_IMAGE:figures/full_fig_p022_6.png] view at source ↗

**Figure 7.** Figure 7: Kaplan Meier curves obtained by stratifying all 397 patients with the [PITH_FULL_IMAGE:figures/full_fig_p023_7.png] view at source ↗

**Figure 8.** Figure 8: Modality ablation experiment results for the 2-year therapeutic out [PITH_FULL_IMAGE:figures/full_fig_p025_8.png] view at source ↗

read the original abstract

Extranodal extension (ENE) is an emerging prognostic factor in human papillomavirus (HPV)-associated oropharyngeal cancer (OPC), although it is currently omitted as a clinical staging criteria. Recent works have advocated for the inclusion of iENE as a prognostic marker in HPV-positive OPC staging. However, several practical limitations continue to hinder its clinical integration, including inconsistencies in segmentation, low contrast in the periphery of metastatic lymph nodes on CT imaging, and laborious manual annotations. To address these limitations, we propose a fully automated end-to-end pipeline that uses computed tomography (CT) images with clinical data to assess the status of nodal ENE and predict treatment outcomes. Our approach includes a hierarchical 3D semi-supervised segmentation model designed to detect and delineate relevant iENE from radiotherapy planning CT scans. From these segmentations, a set of radiomics and deep features are extracted to train an imaging-detected ENE grading classifier. The predicted ENE status is then evaluated for its prognostic value and compared with existing staging criteria. Furthermore, we integrate these nodal features with primary tumor characteristics in a multimodal, attention-based outcome prediction model, providing a dynamic framework for outcome prediction. Our method is validated in an internal cohort of 397 HPV-positive OPC patients treated with radiation therapy or chemoradiotherapy between 2009 and 2020. For outcome prediction at the 2-year mark, our pipeline surpassed baseline models with 88.2% (4.8) in AUC for metastatic recurrence, 79.2% (7.4) for overall survival, and 78.1% (8.6) for disease-free survival. We also obtain a concordance index of 83.3% (6.5) for metastatic recurrence, 71.3% (8.9) for overall survival, and 70.0% (8.1) for disease-free survival, making it feasible for clinical decision making.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a technically integrated pipeline for semi-supervised ENE segmentation plus attention-based multimodal outcome prediction in HPV+ OPC, but everything rests on a single internal 397-patient cohort with no external validation shown.

read the letter

The main takeaway is a practical end-to-end setup that starts with hierarchical 3D semi-supervised segmentation on radiotherapy CT to handle low-contrast ENE, pulls radiomics and deep features for grading, then fuses those nodal signals with primary tumor data via attention to predict 2-year metastatic recurrence, overall survival, and disease-free survival. On their 397 HPV+ OPC patients treated 2009-2020, the model reports AUCs of 88.2% for metastatic recurrence, 79.2% for OS, and 78.1% for DFS, beating the baselines they tested, along with decent concordance indices. That combination for this specific clinical gap is not previously described in the cited work, and the attention fusion step is a sensible way to let the model weigh imaging versus clinical inputs dynamically. The cohort size is decent and the motivation around current staging limitations is clear. The soft spots are straightforward. All numbers come from one internal set with no external cohort, multi-center test, or prospective arm, so claims about clinical feasibility and superiority rest on untested generalization. The abstract flags annotation inconsistencies and low peripheral contrast on CT, yet the reliability of the semi-supervised segmenter on those exact issues is not demonstrated beyond internal performance. Cross-validation strategy, imbalance handling, and any risk of leakage between segmentation and prediction stages are not detailed enough in the provided text to judge. The reported standard deviations are useful but without statistical tests against baselines it is hard to know how meaningful the gains are. This is for people building multimodal models in head-and-neck oncology imaging who want an example of tying segmentation directly to survival endpoints. A reader already working on radiomics or attention fusion could borrow the architecture ideas, but anyone needing deployable results will want more validation first. It deserves a serious referee to check the methods section, request external data or at least detailed ablations, and clarify the statistical comparisons. I would send it to review rather than desk reject.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes AMO-ENE, a fully automated end-to-end pipeline for HPV-associated oropharyngeal cancer that employs a hierarchical 3D semi-supervised segmentation model on radiotherapy planning CT scans to delineate intranodal extranodal extension (iENE), extracts radiomics and deep features for ENE grading, and integrates these with primary tumor characteristics and clinical data via an attention-based multi-omics fusion model to predict 2-year outcomes (metastatic recurrence, overall survival, disease-free survival). The approach is validated on a single internal cohort of 397 patients (2009-2020), reporting AUCs of 88.2% (4.8), 79.2% (7.4), and 78.1% (8.6) respectively that surpass baselines, along with concordance indices, and concludes that the pipeline is feasible for clinical decision making.

Significance. If the performance generalizes beyond the single internal cohort, the work could support clinical integration of iENE as a prognostic marker by automating segmentation on low-contrast CT images with inconsistent annotations. The attention-based multimodal fusion of nodal and tumor features with clinical data offers a dynamic framework for outcome prediction that addresses practical limitations in current staging. The semi-supervised segmentation component is a targeted response to annotation challenges, though its reliability remains unquantified in the provided description.

major comments (3)

[Abstract] Abstract: The reported performance metrics (e.g., AUC 88.2% (4.8) for metastatic recurrence) and claim of clinical feasibility rest on an internal 397-patient cohort without any description of cross-validation strategy, train/test split details, handling of class imbalance, or statistical comparison methods against baselines. This is load-bearing for the central superiority and feasibility claims.
[Abstract] Abstract and Methods (implied): The pipeline depends on the hierarchical 3D semi-supervised segmentation model to reliably delineate low-contrast iENE despite acknowledged annotation inconsistencies, yet no quantitative segmentation metrics (e.g., Dice scores, sensitivity/specificity) or ablation on segmentation accuracy are referenced, which directly affects the validity of downstream radiomics/deep feature extraction and ENE grading.
[Abstract] Abstract: The validation is limited to a single internal cohort (2009-2020) with no external validation, multi-center testing, or prospective evaluation described, undermining the generalization of the reported AUCs and concordance indices to support claims of clinical decision-making feasibility.

minor comments (2)

[Title] The title uses 'Extra Nodal Extension' which should be standardized to 'Extranodal Extension' for consistency with the abstract and field terminology.
[Abstract] The abstract reports standard deviations in parentheses (e.g., 4.8) but does not clarify whether these represent standard deviation across folds, bootstraps, or another variability measure.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for their constructive feedback on our manuscript. We provide point-by-point responses to the major comments and indicate the revisions made to address them.

read point-by-point responses

Referee: [Abstract] Abstract: The reported performance metrics (e.g., AUC 88.2% (4.8) for metastatic recurrence) and claim of clinical feasibility rest on an internal 397-patient cohort without any description of cross-validation strategy, train/test split details, handling of class imbalance, or statistical comparison methods against baselines. This is load-bearing for the central superiority and feasibility claims.

Authors: We agree that the abstract would benefit from more methodological detail to support the reported metrics. The full manuscript describes these aspects in the Methods section. We have revised the abstract to include a concise summary of the cross-validation strategy, train/test split details, class imbalance handling, and statistical comparison methods against baselines. revision: yes
Referee: [Abstract] Abstract and Methods (implied): The pipeline depends on the hierarchical 3D semi-supervised segmentation model to reliably delineate low-contrast iENE despite acknowledged annotation inconsistencies, yet no quantitative segmentation metrics (e.g., Dice scores, sensitivity/specificity) or ablation on segmentation accuracy are referenced, which directly affects the validity of downstream radiomics/deep feature extraction and ENE grading.

Authors: We concur that quantitative evaluation of the segmentation model is essential. In the revised manuscript, we have incorporated segmentation performance metrics such as Dice scores, sensitivity, and specificity, as well as an ablation analysis on the effect of segmentation accuracy on the subsequent feature extraction and outcome prediction tasks. revision: yes
Referee: [Abstract] Abstract: The validation is limited to a single internal cohort (2009-2020) with no external validation, multi-center testing, or prospective evaluation described, undermining the generalization of the reported AUCs and concordance indices to support claims of clinical decision-making feasibility.

Authors: We recognize the importance of external validation for broader applicability. We have updated the abstract, results, and discussion sections to explicitly note the single-institution nature of the cohort and to moderate the language regarding clinical feasibility, emphasizing the need for further validation. However, we do not have external datasets available for inclusion in this revision. revision: partial

standing simulated objections not resolved

External validation on multi-center or prospective data, which is not available in the current study.

Circularity Check

0 steps flagged

No circularity in derivation chain; pipeline is standard ML training and internal validation

full rationale

The paper presents a multi-stage pipeline (hierarchical 3D semi-supervised segmentation for iENE, radiomics/deep feature extraction, ENE grading classifier, then attention-based multimodal fusion for outcome prediction) trained and evaluated on one internal 397-patient cohort. No equations, derivations, or self-citations are shown that reduce the reported AUC/C-index values to quantities defined by construction from the same fitted parameters. The abstract and description frame the results as empirical validation of a trained model rather than tautological renaming or self-referential fitting. Absence of full methods text prevents ruling out subtle data leakage, but no load-bearing circular step is exhibited by the provided text.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review is limited to the abstract; the model relies on standard deep-learning assumptions (semi-supervised learning works on low-contrast CT, radiomic features capture ENE biology, attention fusion improves multimodal prediction) that are not independently verified here.

axioms (1)

domain assumption Semi-supervised learning can produce accurate segmentations of low-contrast metastatic lymph nodes on CT despite annotation inconsistencies
Invoked to justify the hierarchical 3D segmentation stage that feeds all downstream features.

pith-pipeline@v0.9.0 · 5715 in / 1539 out tokens · 79612 ms · 2026-05-10T17:13:40.030765+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

45 extracted references · 13 canonical work pages · 4 internal anchors

[1]

Abboud, H

Z. Abboud, H. Lombaert, and S. Kadoury. Sparse bayesian networks: Effi- cient uncertainty quantification in medical image analysis. InInternational Conference on Medical Image Computing and Computer-Assisted Interven- tion, pages 675–684. Springer, 2024

2024
[2]

personalized

M. B. Amin, F. L. Greene, S. B. Edge, C. C. Compton, J. E. Gershenwald, R. K. Brookland, L. Meyer, D. M. Gress, D. R. Byrd, and D. P. Winchester. The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more "personalized" approach to cancer staging.CA: a cancer journal for clinicians, 67(2):93–99, Mar. 2017

2017
[3]

Andrearczyk, V

V. Andrearczyk, V. Oreiller, M. Abobakr, A. Akhavanallaf, P. Balermpas, S. Boughdad, L. Capriotti, J. Castelli, C. C. Le Rest, P. Decazes, R. Cor- reia, D. El-Habashy, H. Elhalawani, C. D. Fuller, M. Jreige, Y. Khamis, A. La Greca, A. Mohamed, M. Naser, J. O. Prior, S. Ruan, S. Tanadini- Lang, O. Tankyevych, Y. Salimi, M. Vallières, P. Vera, D. Visvikis, ...

2022
[4]

K. M. Boehm, P. Khosravi, R. Vanguri, J. Gao, and S. P. Shah. Harnessing multimodaldataintegrationtoadvanceprecisiononcology.Nature Reviews. Cancer, 22(2):114–126, Feb. 2022

2022
[5]

M. J. Cardoso, W. Li, R. Brown, N. Ma, E. Kerfoot, Y. Wang, B. Murrey, A. Myronenko, C. Zhao, D. Yang, V. Nath, Y. He, Z. Xu, A. Hatamizadeh, A. Myronenko, W. Zhu, Y. Liu, M. Zheng, Y. Tang, I. Yang, M. Zephyr, B. Hashemian, S. Alle, M. Z. Darestani, C. Budd, M. Modat, T. Ver- cauteren, G. Wang, Y. Li, Y. Hu, Y. Fu, B. Gorman, H. Johnson, 30 B. Genereaux,...

work page internal anchor Pith review arXiv 2022
[6]

N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer. SMOTE: Synthetic Minority Over-sampling Technique.Journal of Artificial Intelli- gence Research, 16:321–357, June 2002. arXiv:1106.1813 [cs]

work page arXiv 2002
[7]

C. Chen, J. Miao, D. Wu, A. Zhong, Z. Yan, S. Kim, J. Hu, Z. Liu, L. Sun, X. Li, T. Liu, P.-A. Heng, and Q. Li. MA-SAM: Modality-agnostic SAM adaptation for 3D medical image segmentation.Medical Image Analysis, 98:103310, Dec. 2024

2024
[8]

M. Chen, K. Wang, and J. Wang. Advancing Head and Neck Cancer Sur- vival Prediction via Multi-Label Learning and Deep Model Interpretation. ArXiv, page arXiv:2405.05488v1, May 2024

work page arXiv 2024
[9]

XGBoost: A Scalable Tree Boosting System

T. Chen and C. Guestrin. XGBoost: A Scalable Tree Boosting Sys- tem. InProceedings of the 22nd ACM SIGKDD International Confer- ence on Knowledge Discovery and Data Mining, pages 785–794, Aug. 2016. arXiv:1603.02754 [cs]

work page Pith review arXiv 2016
[10]

Davidson-Pilon

C. Davidson-Pilon. lifelines: survival analysis in python.Journal of Open Source Software, 4(40):1317, 2019

2019
[11]

G. S. Dayan, G. Hénique, H. Bahig, K. Nelson, C. Brodeur, A. Christopou- los, E. Filion, P.-F. Nguyen-Tan, B. O’Sullivan, T. Ayad, et al. Artificial intelligence model for imaging-based extranodal extension detection and outcome prediction in human papillomavirus- positive oropharyngeal can- cer.JAMA Otolaryngology–Head & Neck Surgery, 2025

2025
[12]

An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Un- terthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale, June 2021. arXiv:2010.11929 [cs]

work page internal anchor Pith review Pith/arXiv arXiv 2021
[13]

Falcon and The PyTorch Lightning team

W. Falcon and The PyTorch Lightning team. PyTorch Lightning, Mar. 2019

2019
[14]

arXiv preprint arXiv:2201.01266 , year=

A. Hatamizadeh, V. Nath, Y. Tang, D. Yang, H. Roth, and D. Xu. Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images, Jan. 2022. arXiv:2201.01266 [eess]

work page arXiv 2022
[15]

He and et al

K. He and et al. Masked Autoencoders Are Scalable Vision Learners, Dec
[16]

arXiv:2111.06377 [cs]. 31

work page arXiv
[17]

Y. He, V. Nath, D. Yang, Y. Tang, A. Myronenko, and D. Xu. SwinUNETR-V2: Stronger Swin Transformers with Stagewise Convolu- tions for 3D Medical Image Segmentation. In H. Greenspan, A. Mad- abhushi, P. Mousavi, S. Salcudean, J. Duncan, T. Syeda-Mahmood, and R. Taylor, editors,Medical Image Computing and Computer Assisted In- tervention – MICCAI 2023, page...

2023
[18]

Hiyama, H

T. Hiyama, H. Kuno, T. Nagaki, K. Sekiya, S. Oda, S. Fujii, R. Hayashi, and T. Kobayashi. Extra-nodal extension in head and neck cancer: how radiologists can help staging and treatment planning.Japanese Journal of Radiology, 38(6):489–506, June 2020

2020
[20]

Hoebers, E

F. Hoebers, E. Yu, B. O’Sullivan, A. A. Postma, W. M. Palm, E. Bartlett, J. Lee, S. Stock, S. Koyfman, J. Su, W. Xu, and S. H. Huang. Augmenting inter-rater concordance of radiologic ex- tranodal extension in HPV-positive oropharyngeal carcinoma: A mul- ticenter study.Head & Neck, 44(11):2361–2369, 2022. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10...

work page doi:10.1002/hed.27130 2022
[21]

arXiv preprint arXiv:2404.09556 , year=

F. Isensee, T. Wald, C. Ulrich, M. Baumgartner, S. Roy, K. Maier-Hein, and P. F. Jaeger. nnU-Net Revisited: A Call for Rigorous Validation in 3D Medical Image Segmentation, July 2024. arXiv:2404.09556 [cs]

work page arXiv 2024
[22]

B. H. Kann, S. Aneja, G. V. Loganadane, J. R. Kelly, S. M. Smith, R. H. Decker, J. B. Yu, H. S. Park, W. G. Yarbrough, A. Malhotra, B. A. Burt- ness, and Z. A. Husain. Pretreatment Identification of Head and Neck Cancer Nodal Metastasis and Extranodal Extension Using Deep Learning Neural Networks.Scientific Reports, 8(1):14036, Sept. 2018. Publisher: Natu...

2018
[23]

Kazmierski, M

M. Kazmierski, M. Welch, S. Kim, C. McIntosh, K. Rey-McIntyre, S. H. Huang, T. Patel, T. Tadic, M. Milosevic, F.-F. Liu, A. Ryczkowski, J. Kazmierska, Z. Ye, D. Plana, H. J. Aerts, B. H. Kann, S. V. Bratman, A. J. Hope, and B. Haibe-Kains. Multi-institutional Prognostic Modeling in Head and Neck Cancer: Evaluating Impact and Generalizability of Deep Learn...

2023
[24]

Kharytaniuk, P

N. Kharytaniuk, P. Molony, S. Boyle, G. O’Leary, R. Werner, C. Hef- fron, L. Feeley, and P. Sheahan. Association of Extracapsular Spread With 32 Survival According to Human Papillomavirus Status in Oropharynx Squa- mous Cell Carcinoma and Carcinoma of Unknown Primary Site.JAMA otolaryngology– head & neck surgery, 142(7):683–690, July 2016

2016
[25]

Segment Anything

A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y. Lo, P. Dollár, and R. Girshick. Segment Anything, Apr. 2023. arXiv:2304.02643 [cs]

work page internal anchor Pith review arXiv 2023
[26]

Krizhevsky, I

A. Krizhevsky, I. Sutskever, and G. E. Hinton. ImageNet Classification with Deep Convolutional Neural Networks. In F. Pereira, C. J. Burges, L. Bottou, and K. Q. Weinberger, editors,Advances in Neural Information Processing Systems, volume 25. Curran Associates, Inc., 2012

2012
[27]

W. T. Le, E. Vorontsov, F. P. Romero, L. Seddik, M. M. Elsharief, P. F. Nguyen-Tan, D. Roberge, H. Bahig, and S. Kadoury. Cross-institutional outcome prediction for head and neck cancer patients using self-attention neural networks.Scientific Reports, 12(1):3183, Feb. 2022

2022
[28]

J. Ma, Y. He, F. Li, L. Han, C. You, and B. Wang. Segment anything in medical images.Nature Communications, 15(1):654, Jan. 2024. Publisher: Nature Publishing Group

2024
[29]

M. Meng, L. Bi, D. Feng, and J. Kim. Radiomics-enhanced Deep Multi-task Learning for Outcome Prediction in Head and Neck Cancer, Nov. 2022

2022
[30]

Myronenko, M

A. Myronenko, M. M. R. Siddiquee, D. Yang, Y. He, and D. Xu. Auto- mated head and neck tumor segmentation from 3D PET/CT, Sept. 2022. arXiv:2209.10809 [eess]

work page arXiv 2022
[31]

M. M. Oken, R. H. Creech, D. C. Tormey, J. Horton, T. E. Davis, E. T. McFadden, and P. P. Carbone. Toxicity and response criteria of the East- ern Cooperative Oncology Group.American Journal of Clinical Oncology, 5(6):649–655, Dec. 1982

1982
[32]

S. Pai, D. Bontempi, I. Hadzic, V. Prudente, M. Sokač, T. L. Chaunzwa, S. Bernatz, A. Hosny, R. H. Mak, N. J. Birkbak, and H. J. W. L. Aerts. Foundation model for cancer imaging biomarkers.Nature Machine Intelli- gence, 6(3):354–367, Mar. 2024. Publisher: Nature Publishing Group

2024
[33]

J. Peng, Y. Lu, L. Chen, K. Qiu, F. Chen, J. Liu, W. Xu, W. Zhang, Y. Zhao, Z. Yu, and J. Ren. The prognostic value of machine learning techniques versus cox regression model for head and neck cancer.Methods, 205:123–132, Sept. 2022

2022
[34]

K. Peng, D. Zhou, and S. Gong. OAR-UNet: Enhancing Long-Distance Dependencies for Head and Neck OAR Segmentation.Electronics, 13(18):3771, Jan. 2024. Number: 18 Publisher: Multidisciplinary Digital Publishing Institute. 33

2024
[35]

Rebaud, T

L. Rebaud, T. Escobar, F. Khalid, K. Girum, and I. Buvat. Simplicity Is All You Need: Out-of-the-Box nnUNet Followed by Binary-Weighted Ra- diomic Model for Segmentation and Outcome Prediction in Head and Neck PET/CT. InHead and Neck Tumor Segmentation and Outcome Predic- tion: Third Challenge, HECKTOR 2022, Held in Conjunction with MIC- CAI 2022, Singapo...

2022
[36]

U-Net: Convolutional Networks for Biomedical Image Segmentation

O. Ronneberger, P. Fischer, and T. Brox. U-Net: Convolutional Networks for Biomedical Image Segmentation, May 2015. arXiv:1505.04597 [cs]

work page internal anchor Pith review Pith/arXiv arXiv 2015
[37]

B. S. Rose, J.-H. Jeong, S. K. Nath, S. M. Lu, and L. K. Mell. Population- Based Study of Competing Mortality in Head and Neck Cancer.Journal of Clinical Oncology, 29(26):3503–3509, Sept. 2011. Publisher: Wolters Kluwer

2011
[38]

Saber, M

R. Saber, M. Tonneau, S. Bahig, J. Malo, W. Belkaid, M. Messaoudene, N. Bouchard, F. Coulombe, P. Joubert, H. Bahig, S. Turcotte, B. Routy, and S. Kadoury. Feature Tokenizer-Transformers with Self-Training for The Prediction of PD-L1 Expression of Non-Small Cell Lung Cancer from CT. In2024 IEEE International Symposium on Biomedical Imaging (ISBI), pages 1...

2024
[39]

Sinha, D

P. Sinha, D. Kallogjeri, H. Gay, W. L. Thorstad, J. S. Lewis, R. Chernock, B. Nussenbaum, and B. H. Haughey. High metastatic node number, not extracapsular spread or N-classification is a node-related prognosticator in transorally-resected, neck-dissected p16-positive oropharynx cancer.Oral Oncology, 51(5):514–520, May 2015

2015
[40]

Vallières, E

M. Vallières, E. Kay-Rivest, L. J. Perrin, X. Liem, C. Furstoss, H. J. W. L. Aerts, N. Khaouam, P. F. Nguyen-Tan, C.-S. Wang, K. Sultanem, J. Seun- tjens, and I. El Naqa. Radiomics strategies for risk assessment of tumour failure in head-and-neck cancer.Scientific Reports, 7(1):10117, Aug. 2017. Publisher: Nature Publishing Group

2017
[41]

J. J. M. van Griethuysen, A. Fedorov, C. Parmar, A. Hosny, N. Aucoin, V. Narayan, R. G. H. Beets-Tan, J.-C. Fillion-Robin, S. Pieper, and H. J. W.L.Aerts. ComputationalRadiomicsSystemtoDecodetheRadiographic Phenotype.Cancer Research, 77(21):e104–e107, Nov. 2017

2017
[42]

R. S. Vanguri, J. Luo, A. T. Aukerman, J. V. Egger, C. J. Fong, N. Horvat, A. Pagano, J. d. A. B. Araujo-Filho, L. Geneslaw, H. Rizvi, et al. Mul- timodal integration of radiology, pathology and genomics for prediction of response to pd-(l) 1 blockade in patients with non-small cell lung cancer. Nature cancer, 3(10):1151–1164, 2022

2022
[43]

H. Wang, S. Guo, J. Ye, Z. Deng, J. Cheng, T. Li, J. Chen, Y. Su, Z. Huang, Y. Shen, B. Fu, S. Zhang, J. He, and Y. Qiao. SAM-Med3D: 34 Towards General-purpose Segmentation Models for Volumetric Medical Im- ages, Sept. 2024. arXiv:2310.15161 [cs]

work page arXiv 2024
[44]

Wodzinski

M. Wodzinski. Benchmark of Deep Encoder-Decoder Architectures for Head and Neck Tumor Segmentation in Magnetic Resonance Images: Contribution to the HNTSMRG Challenge. In K. A. Wahid, C. Dede, M. A. Naser, and C. D. Fuller, editors,Head and Neck Tumor Segmen- tation for MR-Guided Applications, pages 204–213, Cham, 2025. Springer Nature Switzerland

2025
[45]

Xiong, B

X. Xiong, B. J. Smith, S. A. Graves, M. M. Graham, J. M. Buatti, and R. R. Beichel. Head and Neck Cancer Segmentation in FDG PET Im- ages: Performance Comparison of Convolutional Neural Networks and Vi- sion Transformers.Tomography, 9(5):1933–1948, Oct. 2023. Number: 5 Publisher: Multidisciplinary Digital Publishing Institute

1933
[46]

C.-N. Yu, R. Greiner, H.-C. Lin, and V. Baracos. Learning Patient-Specific Cancer Survival Distributions as a Sequence of Dependent Regressors. In Advances in Neural Information Processing Systems, volume 24. Curran Associates, Inc., 2011. 35

2011