Other Open Access
Vincent Andrearczyk; Valentin Oreiller; Martin Vallières; Joel Castelli; Hesham Elhalawani; Mario Jreige; Sarah Boughdad; John O. Prior; Adrien Depeursinge
This is the challenge design document for the "3D Head and Neck Tumor Segmentation in PET/CT", accepted for MICCAI 2020.
Head and Neck (H&N) cancers are among the most common cancers worldwide (5th leading cancer by incidence) (Parkin et al. 2005). Radiotherapy combined with cetuximab has been established as standard treatment (Bonner et al. 2010). However, locoregional failures remain a major challenge and occur in up to 40% of patients in the first two years after the treatment (Chajon et al. 2013). Recently, several radiomics studies based on Positron Emission Tomography (PET) and Computed Tomography (CT) imaging were proposed to better identify patients with a worse prognosis in a non-invasive fashion and by reusing images acquired for diagnosis and treatment planning (Vallières et al. 2017),(Bogowicz et al. 2017),(Castelli et al. 2017). Although highly promising, these methods were validated on 100-400 patients. Further validation on larger cohorts (e.g. 300-3000 patients) is required to respect an adequate ratio between the number of variables and observations in order to avoid an overestimation of the generalization performance. Achieving such a validation requires the manual delineation of primary tumors and nodal metastases for every patient and in three dimensions, which is intractable and error-prone.
Methods for automated lesion segmentation in medical images were proposed in various contexts, often achieving expert-level performance (Heimann and Meinzer 2009), (Menze et al. 2015). Surprisingly few studies evaluated the performance of computerized automated segmentation of tumor lesions in PET and CT images (Song et al. 2013),(Blanc-Durand et al. 2018), (Moe et al. 2019).
Therefore, it is timely to propose a MICCAI challenge to advance the methodological aspects and their validation for automated tumor and metastatic lymph nodes segmentation in PET/CT images. We also expect these progress and knowledge to be transferable for the segmentation of other types of cancer in the aforementioned imaging modalities. By focusing on metabolic and morphological tissue properties respectively, PET and CT modalities include complementary and synergistic information for cancerous lesion segmentation, that only modern image analysis methods can fully leverage.
This challenge will offer an opportunity for participants working on 3D segmentation algorithms to develop automatic bi-modal approaches for the segmentation of H&N tumors in PET/CT scans, focusing on oropharyngeal cancers. Various approaches must be explored and compared to extract and merge information from the two modalities, including early or late fusion, full volume or patch based approaches, 2-, 2.5- or 3-D approach. The data used in this challenge will be multi-centric, including four centers in Canada (Vallières et al. 2017) and one center in Switzerland (Castelli et al. 2017) for a total of 249 patients with both tumor and metastatic lymph nodes contoured.
In addition to these 249 cases for which we already have all the agreements and information, we will likely include approximately 330 additional cases, including 215 public cases from (Grossberg et al. 2017), 88 public cases from (Wee et al. 2019) both for the training set, as well as approximately 30 non-public cases from McGill, for which we still need to obtain the data agreement. We do not include these cases in the data description as we do not have all data description or data agreement yet.
Blanc-Durand, Paul, Axel Van Der Gucht, Niklaus Schaefer, Emmanuel Itti, and John O. Prior. 2018. “Automatic
Lesion Detection and Segmentation of 18F-FET PET in Gliomas: A Full 3D U-Net Convolutional Neural Network
Study.” PloS One 13 (4): e0195798.
Bogowicz, Marta, Oliver Riesterer, Luisa Sabrina Stark, Gabriela Studer, Jan Unkelbach, Matthias Guckenberger, and
Stephanie Tanadini-Lang. 2017. “Comparison of PET and CT Radiomics for Prediction of Local Tumor Control in
Head and Neck Squamous Cell Carcinoma.” Acta Oncologica 56 (11): 1531–36.
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.
Castelli, J., A. Depeursinge, V. Ndoh, J. O. Prior, M. Ozsahin, A. Devillers, H. Bouchaab, et al. 2017. “A PET-Based
Nomogram for Oropharyngeal Cancers.” European Journal of Cancer 75 (April): 222–30.
Chajon, Enrique, Caroline Lafond, Guillaume Louvel, Joël Castelli, Danièle Williaume, Olivier Henry, Franck Jégoux,
et al. 2013. “Salivary Gland-Sparing Other than Parotid-Sparing in Definitive Head-and-Neck Intensity-Modulated
Radiotherapy Does Not Seem to Jeopardize Local Control.” Radiation Oncology. https://doi.org/10.1186/1748-
Grossberg A, Mohamed A, Elhalawani H, Bennett W, Smith K, Nolan T, Chamchod S, Kantor M, Browne T,
Hutcheson K, Gunn G, Garden A, Frank S, Rosenthal D, Freymann J, Fuller C.(2017). Data from Head and Neck
Cancer CT Atlas. The Cancer Imaging Archive. DOI: 10.7937/K9/TCIA.2017.umz8dv6s
Heimann, Tobias, and Hans-Peter Meinzer. 2009. “Statistical Shape Models for 3D Medical Image Segmentation: A
Review.” Medical Image Analysis. https://doi.org/10.1016/j.media.2009.05.004.
Menze, Bjoern H., Andras Jakab, Stefan Bauer, Jayashree Kalpathy-Cramer, Keyvan Farahani, Justin Kirby, Yuliya
Burren, et al. 2015. “The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS).” IEEE Transactions on
Medical Imaging 34 (10): 1993–2024.
Moe, Yngve Mardal, Aurora Rosvoll Groendahl, Martine Mulstad, Oliver Tomic, Ulf Indahl, Einar Dale, Eirik
Malinen, and Cecilia Marie Futsaether. "Deep learning for automatic tumour segmentation in PET/CT images of
patients with head and neck cancers." (2019).
Parkin, D. M., F. Bray, J. Ferlay, and P. Pisani. 2005. “Global Cancer Statistics, 2002.” CA: A Cancer Journal for
Song, Qi, Junjie Bai, Dongfeng Han, Sudershan Bhatia, Wenqing Sun, William Rockey, John E. Bayouth, John M.
Buatti, and Xiaodong Wu. 2013. “Optimal Co-Segmentation of Tumor in PET-CT Images with Context Information.”
IEEE Transactions on Medical Imaging 32 (9): 1685–97.
Vallières, Martin, Emily Kay-Rivest, Léo Jean Perrin, Xavier Liem, Christophe Furstoss, Hugo J. W. L. Aerts, Nader
Khaouam, et al. 2017. “Radiomics Strategies for Risk Assessment of Tumour Failure in Head-and-Neck Cancer.”
Scientific Reports 7 (1): 10117.
Wee, L., & Dekker, A. (2019). Data from Head-Neck-Radiomics-HN1 [Data set]. The Cancer Imaging Archive.