HEad and neCK TumOR (HECKTOR) Lesion Segmentation, Staging and Prognosis using Multimodal Data¶
The HECKTOR challenge returns for its 2026 edition at the 29th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2026) in Abu Dhabi, United Arab Emirates. Building on the success of previous editions (2020–2025), HECKTOR continues to advance clinically relevant machine learning solutions for head and neck cancer.
Challenge Task¶
Participants are invited to develop a multimodal pipeline leveraging FDG PET, CT, and clinical data to:
- Segment primary tumors and lymph nodes
- Infer radiological TN staging
- Predict recurrence-free survival
This unified task reflects a realistic clinical workflow, integrating diagnosis, staging, and prognosis into a single framework.
Clinical Background and Motivation¶
Head and neck cancers represent a significant global health burden, disproportionately affecting men and older populations (Barsouk et al. 2023). Despite advances such as radiotherapy combined with cetuximab, disease control remains challenging, with locoregional recurrence rates reaching up to 40% within two years (Bonner et al. 2010; Chajon et al. 2013). Medical imaging plays a central role in patient management, and recent radiomics studies using PET and CT have shown promise for non-invasive tumor characterization and outcome prediction (Vallières et al. 2017; Bogowicz et al. 2017; Castelli et al. 2017). These modalities provide complementary insights into tumor biology by capturing metabolic and anatomical information. However, most existing approaches are limited by small, single-center cohorts, highlighting the need for large-scale, multi-institutional datasets to develop robust, generalizable, and clinically translatable models.
Medical imaging plays a central role in patient management. PET and CT provide complementary information:
- PET captures metabolic activity
- CT provides detailed anatomical structure
Together, they enable improved tumor characterization and risk stratification.
Recent advances in radiomics and deep learning have shown promise in extracting prognostic information from these modalities. However, most existing studies rely on small, single-center cohorts, limiting generalizability. There is a critical need for large-scale, multi-institutional benchmarks to develop robust and clinically translatable models.
The HECKTOR challenge has progressively expanded in scope and complexity:
- Earlier editions focused on tumor segmentation and outcome prediction
- The 2025 edition introduced HPV prediction and expanded the dataset with new centers and modalities
- Evaluation protocols evolved to better reflect clinical realism
For 2026, HECKTOR takes a major step forward by unifying tasks into a single end-to-end clinical pipeline, enabling:
- Treatment planning support
- Improved risk stratification
- More accurate outcome prediction
By jointly modeling anatomy, disease stage, and prognosis, the challenge promotes the development of clinically meaningful AI systems ready for real-world integration.
What’s New in HECKTOR 2026¶
1. Expanded Multi-Center Dataset¶
- Total cohort increased to ~1423 patients
- Includes:
- ~270 newly annotated cases from CHU Brest (primary tumor + lymph nodes)
- Completion of radiological TN staging annotations
- Previously filtered public datasets (~96 cases)
- New data from UAE medical centers (~150 cases)
This results in one of the largest multi-institutional datasets for head and neck cancer imaging.
2. Unified End-to-End Task¶
For the first time, HECKTOR introduces a single integrated task covering:
Segmentation → TN Staging → Prognosis
Participants may:
- Submit modular pipelines, or
- Develop end-to-end models (strongly encouraged)
This design mirrors real clinical workflows and promotes:
- Knowledge sharing across subtasks
- Reduced redundancy
- Improved computational efficiency
3. Focus on Clinical Translation¶
Rather than optimizing isolated components, participants are encouraged to develop holistic models capable of supporting real-world decision-making, similar to how radiologists interpret imaging, assign staging, and assess prognosis. This unified approach reflects a more practical and scalable deployment strategy, where a single model can support multiple steps of patient management.
4. Introduction of TN Staging¶
- New task component: Radiological TN staging
- Clinically critical for:
- Treatment planning
- Risk stratification
- Prognostic assessment
5. Refined Segmentation Evaluation¶
While segmentation remains consistent with 2025:
- Evaluation metrics are refined to better assess lymph node detection and delineation
- Increased emphasis on multi-lesion performance
References¶
[Saeed et al. 2025] Numan Saeed, Salma Hassan, Shahad Hardan, et al. "A Multimodal and Multi-centric Head and Neck Cancer Dataset for Segmentation, Diagnosis and Outcome Prediction," arXiv:2509.00367 (2025)
[Andrearczyk et al. 2022] Andrearczyk V, et al. "Overview of the HECKTOR Challenge at MICCAI 2021: Automatic Head and Neck Tumor Segmentation and Outcome Prediction in PET/CT Images", in: Head and Neck Tumor Segmentation and Outcome Prediction, 1-37 (2022).
[Bonner et al. 2010] Bonner, James A., Paul M. Harari, Jordi Giralt, Roger B. Cohen, Christopher U. Jones, Ranjan K. Sur, David Raben, et al. 2010. “Radiotherapy plus Cetuximab for Locoregionally Advanced Head and Neck Cancer: 5-Year Survival Data from a Phase 3 Randomised Trial, and Relation between Cetuximab-Induced Rash and Survival.” The Lancet Oncology 11 (1): 21–28.
[Bogowicz et al. 2017] Bogowicz M, et al. "Comparison of PET and CT radiomics for prediction of local tumor control in head and neck squamous cell carcinoma." Acta Oncologica 56.11 (2017): 1531-1536.
[Castelli et al. 2017] Castelli J, et al. "A PET-based nomogram for oropharyngeal cancers." European Journal of Cancer 75 (2017): 222-230.
[Chajon et al. 2013] Chajon E, et al. "Salivary gland-sparing other than parotid-sparing in definitive head-and-neck intensity-modulated radiotherapy does not seem to jeopardize local control." Radiation Oncology 8.1 (2013): 1-9.
[Parkin et al. 2005] Parkin DM, et al. "Global cancer statistics, 2002." CA: a cancer journal for clinicians 55.2 (2005): 74-108.
[Vallières et al. 2017] Vallières M, et al. “Radiomics strategies for risk assessment of tumour failure in head-and-neck cancer.” Nature Scientific Reports, 7(1):10117 (2017).