Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

The 2024 Brain Tumor Segmentation (BraTS) Challenge: Glioma Segmentation on Post-treatment MRI (2405.18368v1)

Published 28 May 2024 in cs.CV

Abstract: Gliomas are the most common malignant primary brain tumors in adults and one of the deadliest types of cancer. There are many challenges in treatment and monitoring due to the genetic diversity and high intrinsic heterogeneity in appearance, shape, histology, and treatment response. Treatments include surgery, radiation, and systemic therapies, with magnetic resonance imaging (MRI) playing a key role in treatment planning and post-treatment longitudinal assessment. The 2024 Brain Tumor Segmentation (BraTS) challenge on post-treatment glioma MRI will provide a community standard and benchmark for state-of-the-art automated segmentation models based on the largest expert-annotated post-treatment glioma MRI dataset. Challenge competitors will develop automated segmentation models to predict four distinct tumor sub-regions consisting of enhancing tissue (ET), surrounding non-enhancing T2/fluid-attenuated inversion recovery (FLAIR) hyperintensity (SNFH), non-enhancing tumor core (NETC), and resection cavity (RC). Models will be evaluated on separate validation and test datasets using standardized performance metrics utilized across the BraTS 2024 cluster of challenges, including lesion-wise Dice Similarity Coefficient and Hausdorff Distance. Models developed during this challenge will advance the field of automated MRI segmentation and contribute to their integration into clinical practice, ultimately enhancing patient care.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (35)
  1. Q. T. Ostrom, M. Price, C. Neff, G. Cioffi, K. A. Waite, C. Kruchko, and J. S. Barnholtz-Sloan, “Cbtrus statistical report: primary brain and other central nervous system tumors diagnosed in the united states in 2016—2020,” Neuro-oncology, vol. 25, no. Supplement_4, pp. iv1–iv99, 2023.
  2. Lyon: International Agency for Research on Cancer, 5 ed., 2021.
  3. R. Stupp, W. P. Mason, M. J. Van Den Bent, M. Weller, B. Fisher, M. J. Taphoorn, K. Belanger, A. A. Brandes, C. Marosi, U. Bogdahn, et al., “Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma,” New England journal of medicine, vol. 352, no. 10, pp. 987–996, 2005.
  4. K. V. Hoebel, C. P. Bridge, S. Ahmed, O. Akintola, C. Chung, R. Y. Huang, J. M. Johnson, A. Kim, K. I. Ly, K. Chang, et al., “Expert-centered evaluation of deep learning algorithms for brain tumor segmentation,” Radiology: Artificial Intelligence, vol. 6, no. 1, p. e220231, 2023.
  5. K. Clark, B. Vendt, K. Smith, J. Freymann, J. Kirby, P. Koppel, S. Moore, S. Phillips, D. Maffitt, M. Pringle, et al., “The cancer imaging archive (tcia): maintaining and operating a public information repository,” Journal of digital imaging, vol. 26, pp. 1045–1057, 2013.
  6. B. H. Menze, A. Jakab, S. Bauer, J. Kalpathy-Cramer, K. Farahani, J. Kirby, Y. Burren, N. Porz, J. Slotboom, R. Wiest, et al., “The multimodal brain tumor image segmentation benchmark (brats),” IEEE transactions on medical imaging, vol. 34, no. 10, pp. 1993–2024, 2014.
  7. S. Bakas, M. Reyes, A. Jakab, S. Bauer, M. Rempfler, A. Crimi, R. T. Shinohara, C. Berger, S. M. Ha, M. Rozycki, et al., “Identifying the best machine learning algorithms for brain tumor segmentation, progression assessment, and overall survival prediction in the brats challenge,” arXiv preprint arXiv:1811.02629, 2018.
  8. U. Baid, S. Ghodasara, S. Mohan, M. Bilello, E. Calabrese, E. Colak, K. Farahani, J. Kalpathy-Cramer, F. C. Kitamura, S. Pati, et al., “The rsna-asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification,” arXiv preprint arXiv:2107.02314, 2021.
  9. P. Kickingereder, F. Isensee, I. Tursunova, J. Petersen, U. Neuberger, D. Bonekamp, G. Brugnara, M. Schell, T. Kessler, M. Foltyn, et al., “Automated quantitative tumour response assessment of mri in neuro-oncology with artificial neural networks: a multicentre, retrospective study,” The Lancet Oncology, vol. 20, no. 5, pp. 728–740, 2019.
  10. K. Chang, A. L. Beers, H. X. Bai, J. M. Brown, K. I. Ly, X. Li, J. T. Senders, V. K. Kavouridis, A. Boaro, C. Su, et al., “Automatic assessment of glioma burden: a deep learning algorithm for fully automated volumetric and bidimensional measurement,” Neuro-oncology, vol. 21, no. 11, pp. 1412–1422, 2019.
  11. J. D. Rudie, E. Calabrese, R. Saluja, D. Weiss, J. B. Colby, S. Cha, C. P. Hess, A. M. Rauschecker, L. P. Sugrue, and J. E. Villanueva-Meyer, “Longitudinal assessment of posttreatment diffuse glioma tissue volumes with three-dimensional convolutional neural networks,” Radiology: Artificial Intelligence, vol. 4, no. 5, p. e210243, 2022.
  12. E. Lotan, B. Zhang, S. Dogra, W. Wang, D. Carbone, G. Fatterpekar, E. Oermann, and Y. Lui, “Development and practical implementation of a deep learning–based pipeline for automated pre-and postoperative glioma segmentation,” American Journal of Neuroradiology, vol. 43, no. 1, pp. 24–32, 2022.
  13. M. Ghaffari, G. Samarasinghe, M. Jameson, F. Aly, L. Holloway, P. Chlap, E.-S. Koh, A. Sowmya, and R. Oliver, “Automated post-operative brain tumour segmentation: A deep learning model based on transfer learning from pre-operative images,” Magnetic resonance imaging, vol. 86, pp. 28–36, 2022.
  14. P. J. Sørensen, J. F. Carlsen, V. A. Larsen, F. L. Andersen, C. N. Ladefoged, M. B. Nielsen, H. S. Poulsen, and A. E. Hansen, “Evaluation of the hd-glio deep learning algorithm for brain tumour segmentation on postoperative mri,” Diagnostics, vol. 13, no. 3, p. 363, 2023.
  15. P. Vollmuth, M. Foltyn, R. Y. Huang, N. Galldiks, J. Petersen, F. Isensee, M. J. van den Bent, F. Barkhof, J. E. Park, Y. W. Park, et al., “Artificial intelligence (ai)-based decision support improves reproducibility of tumor response assessment in neuro-oncology: an international multi-reader study,” Neuro-oncology, vol. 25, no. 3, pp. 533–543, 2023.
  16. S. Baig, I. Vidic, G. M. Mastorakos, R. X. Smith, N. White, S. Bash, A. M. Dale, C. R. McDonald, T. Beaumont, T. M. Seibert, J. Hattangadi-Gluth, S. Kesari, N. Farid, and J. D. Rudie, “Segmentation of Pre- and Post-Treatment Diffuse Glioma Tissue Subregions Including Resection Cavities,” In Revision at Neuro-Oncology Advances, 2024.
  17. B. K. Fields, E. Calabrese, J. Mongan, S. Cha, C. P. Hess, L. P. Sugrue, S. M. Chang, T. L. Luks, J. E. Villanueva-Meyer, A. M. Rauschecker, and J. D. Rudie, “The University of California, San Francisco Longitudinal Post-Treatment Diffuse Glioma (UCSF-ALPTDG) MRI Dataset.,” In Revision at Radiology: Artificial Intelligence, 2024.
  18. S. Cepeda, S. García-García, I. Arrese, F. Herrero, T. Escudero, T. Zamora, and R. Sarabia, “The río hortega university hospital glioblastoma dataset: A comprehensive collection of preoperative, early postoperative and recurrence mri scans (rhuh-gbm),” Data in Brief, vol. 50, p. 109617, 2023.
  19. S. Bakas, H. Akbari, A. Sotiras, M. Bilello, M. Rozycki, J. S. Kirby, J. B. Freymann, K. Farahani, and C. Davatzikos, “Advancing the cancer genome atlas glioma mri collections with expert segmentation labels and radiomic features,” Scientific data, vol. 4, no. 1, pp. 1–13, 2017.
  20. R. W. Cox, J. Ashburner, H. Breman, K. Fissell, C. Haselgrove, C. J. Holmes, J. L. Lancaster, D. E. Rex, S. M. Smith, J. B. Woodward, et al., “A (sort of) new image data format standard: Nifti-1,” in 10th Annual Meeting of the Organization for Human Brain Mapping, vol. 22, p. 01, 2004.
  21. F. Isensee, M. Schell, I. Pflueger, G. Brugnara, D. Bonekamp, U. Neuberger, A. Wick, H.-P. Schlemmer, S. Heiland, W. Wick, et al., “Automated brain extraction of multisequence mri using artificial neural networks,” Human brain mapping, vol. 40, no. 17, pp. 4952–4964, 2019.
  22. S. Pati, A. Singh, S. Rathore, A. Gastounioti, M. Bergman, P. Ngo, S. M. Ha, D. Bounias, J. Minock, G. Murphy, et al., “The cancer imaging phenomics toolkit (captk): technical overview,” in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 5th International Workshop, BrainLes 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Revised Selected Papers, Part II 5, pp. 380–394, Springer, 2020.
  23. F. Isensee, P. F. Jaeger, S. A. Kohl, J. Petersen, and K. H. Maier-Hein, “nnu-net: a self-configuring method for deep learning-based biomedical image segmentation,” Nature methods, vol. 18, no. 2, pp. 203–211, 2021.
  24. A. Myronenko, M. M. R. Siddiquee, D. Yang, Y. He, and D. Xu, “Automated head and neck tumor segmentation from 3d pet/ct,” arXiv preprint arXiv:2209.10809, 2022.
  25. T. Rohlfing, D. B. Russakoff, and C. R. Maurer, “Performance-based classifier combination in atlas-based image segmentation using expectation-maximization parameter estimation,” IEEE transactions on medical imaging, vol. 23, no. 8, pp. 983–994, 2004.
  26. A. F. Kazerooni, N. Khalili, X. Liu, D. Haldar, Z. Jiang, S. M. Anwar, J. Albrecht, M. Adewole, U. Anazodo, H. Anderson, et al., “The brain tumor segmentation (brats) challenge 2023: Focus on pediatrics (cbtn-connect-dipgr-asnr-miccai brats-peds),” ArXiv, 2023.
  27. A. W. Moawad, A. Janas, U. Baid, D. Ramakrishnan, L. Jekel, K. Krantchev, H. Moy, R. Saluja, K. Osenberg, K. Wilms, et al., “The brain tumor segmentation (brats-mets) challenge 2023: Brain metastasis segmentation on pre-treatment mri,” ArXiv, 2023.
  28. D. LaBella, M. Adewole, M. Alonso-Basanta, T. Altes, S. M. Anwar, U. Baid, T. Bergquist, R. Bhalerao, S. Chen, V. Chung, et al., “The asnr-miccai brain tumor segmentation (brats) challenge 2023: Intracranial meningioma,” arXiv preprint arXiv:2305.07642, 2023.
  29. M. Adewole, J. D. Rudie, A. Gbdamosi, O. Toyobo, C. Raymond, D. Zhang, O. Omidiji, R. Akinola, M. A. Suwaid, A. Emegoakor, et al., “The brain tumor segmentation (brats) challenge 2023: Glioma segmentation in sub-saharan africa patient population (brats-africa),” ArXiv, 2023.
  30. D. LaBella, U. Baid, O. Khanna, S. McBurney-Lin, R. McLean, P. Nedelec, A. Rashid, N. H. Tahon, T. Altes, R. Bhalerao, et al., “Analysis of the brats 2023 intracranial meningioma segmentation challenge,” arXiv preprint arXiv:2405.09787, 2024.
  31. D. LaBella, O. Khanna, S. McBurney-Lin, R. Mclean, P. Nedelec, A. S. Rashid, N. h. Tahon, T. Altes, U. Baid, R. Bhalerao, et al., “A multi-institutional meningioma mri dataset for automated multi-sequence image segmentation,” Scientific Data, vol. 11, no. 1, p. 496, 2024.
  32. R. Duan, J. Tong, L. Lin, L. Levine, M. Sammel, J. Stoddard, T. Li, C. H. Schmid, H. Chu, and Y. Chen, “Palm: Patient-centered treatment ranking via large-scale multivariate network meta-analysis,” The Annals of Applied Statistics, vol. 17, no. 1, pp. 815–837, 2023.
  33. M. Niyazi, N. Andratschke, M. Bendszus, A. J. Chalmers, S. C. Erridge, N. Galldiks, F. J. Lagerwaard, P. Navarria, P. M. af Rosenschöld, U. Ricardi, et al., “Estro-eano guideline on target delineation and radiotherapy details for glioblastoma,” Radiotherapy and Oncology, vol. 184, p. 109663, 2023.
  34. E. Ermiş, A. Jungo, R. Poel, M. Blatti-Moreno, R. Meier, U. Knecht, D. M. Aebersold, M. K. Fix, P. Manser, M. Reyes, et al., “Fully automated brain resection cavity delineation for radiation target volume definition in glioblastoma patients using deep learning,” Radiation oncology, vol. 15, pp. 1–10, 2020.
  35. J. D. Rudie, A. M. Rauschecker, R. N. Bryan, C. Davatzikos, and S. Mohan, “Emerging applications of artificial intelligence in neuro-oncology,” Radiology, vol. 290, no. 3, pp. 607–618, 2019.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (85)
  1. Maria Correia de Verdier (6 papers)
  2. Rachit Saluja (12 papers)
  3. Louis Gagnon (2 papers)
  4. Dominic LaBella (13 papers)
  5. Ujjwall Baid (1 paper)
  6. Nourel Hoda Tahon (6 papers)
  7. Martha Foltyn-Dumitru (4 papers)
  8. Jikai Zhang (9 papers)
  9. Maram Alafif (1 paper)
  10. Saif Baig (1 paper)
  11. Ken Chang (28 papers)
  12. Gennaro D'Anna (2 papers)
  13. Lisa Deptula (4 papers)
  14. Diviya Gupta (1 paper)
  15. Muhammad Ammar Haider (5 papers)
  16. Ali Hussain (4 papers)
  17. Michael Iv (11 papers)
  18. Marinos Kontzialis (2 papers)
  19. Paul Manning (1 paper)
  20. Farzan Moodi (2 papers)
Citations (11)
View on Semantic Scholar

Summary

  • The paper introduces an automated segmentation framework for post-treatment gliomas using the nnU-Net approach and STAPLE fusion strategy.
  • It utilizes a comprehensive mpMRI dataset from 2,200 cases and evaluates segmentation performance with lesion-wise Dice and Hausdorff metrics across multiple tumor regions.
  • The study addresses challenges from treatment-induced changes and resection cavities, paving the way for enhanced clinical decision-making in glioma management.

Automated Segmentation in Post-Treatment Gliomas: A Review of the 2024 BraTS Challenge

The 2024 Brain Tumor Segmentation (BraTS) challenge seeks to address a critical gap in the field of medical imaging and oncology, focusing on the segmentation of post-treatment gliomas. Gliomas, comprising about 80% of malignant primary brain tumors, require a multi-faceted treatment strategy including surgery, radiation therapy, and systemic therapies, with MRI playing an indispensable role in treatment planning and monitoring. Despite extensive research efforts, significant challenges persist in the accurate and efficient evaluation of post-treatment gliomas due to the intrinsic heterogeneity and treatment-induced changes in their appearance.

Data and Methods

The dataset for the 2024 BraTS challenge consists of post-treatment glioma cases contributed from seven academic medical centers, amassing approximately 2,200 cases. The dataset includes multiparametric MRI (mpMRI) sequences such as pre-contrast T1-weighted, contrast-enhanced T1-weighted, T2-weighted, and FLAIR images. This comprehensive imaging data undergoes a predefined preprocessing workflow, involving brain extraction, registration, and conversion to a standardized file format.

Following preprocessing, the data is subjected to a series of automated segmentation approaches using the nnU-Net framework. Five primary segmentation models are deployed, including versions trained on data from the Duke and UCSF cohorts and those incorporating additional datasets. These results are merged using the STAPLE fusion algorithm, and digital subtraction images further facilitate the annotation process for radiologists.

Evaluation and Performance Metrics

The challenge evaluates segmentation performance according to lesion-wise metrics to focus on individual lesion detection rather than overall image accuracy. The two primary evaluation metrics are:

  1. Lesion-wise Dice Similarity Coefficient (DSC) - Measures voxelwise segmentation overlap.
  2. Lesion-wise 95% Hausdorff Distance (HD95) - Measures the spatial distance between the predicted and ground truth segmentations.

Participants are evaluated across several tumor sub-regions:

  • Enhancing tissue (ET)
  • Non-enhancing tumor core (NETC)
  • Surrounding non-enhancing FLAIR hyperintensity (SNFH)
  • Resection cavity (RC)

The combination of these metrics aims to provide a comprehensive assessment of model performance in detecting various glioma sub-regions, crucial for clinical decision-making.

Discussion

This challenge underscores a shift from pre-treatment to post-treatment glioma segmentation, addressing the complexity introduced by treatment-related changes. The inclusion of resection cavities as a novel tissue sub-region marks a notable expansion in segmentation categories, essential for precise therapy planning and follow-up.

While prior studies have demonstrated the potential of automated segmentation to enhance workflow efficiency, they often fall short in accurately capturing smaller regions of enhancement and other subtle changes. This challenge aims to refine these models, making them more robust in the post-treatment context.

The process of creating these annotated datasets involves significant effort, ensuring high-quality labels through careful manual correction by expert reviewers. Despite this, variability in annotations remains an inherent challenge, suggesting that future efforts may benefit from methods to further mitigate inter-observer variability.

Implications and Future Directions

The 2024 BraTS challenge represents a critical step toward integrating advanced segmentation tools into clinical practice. By making high-quality, annotated datasets publicly available, it supports the development and validation of new algorithms aimed at improving glioma management. The outcomes of this challenge can significantly impact practical workflows, enabling quicker and more reliable tumor evaluation.

Future iterations of BraTS challenges might incorporate longitudinal and multimodal data to better distinguish between residual tumor and treatment-induced changes. Additionally, developing models that can predict treatment responses and recurrence could offer substantial benefits in tailoring personalized treatment strategies.

In conclusion, the 2024 BraTS post-treatment glioma challenge sets a new standard in the field, aiming to enhance the accuracy and efficiency of glioma segmentation. These advancements not only promise to improve patient outcomes but also pave the way for sophisticated future research in tumor characterization and treatment response prediction.