The Medical Segmentation Decathlon
Nature Communications, vol. 13, pp. 4128
Abstract
International challenges have become the de facto standard for comparative assessment of image analysis algorithms. Although segmentation is the most widely investigated medical image processing task, the various challenges have been organized to focus only on specific clinical tasks. We organized the Medical Segmentation Decathlon (MSD)-a biomedical image analysis challenge, in which algorithms compete in a multitude of both tasks and modalities to investigate the hypothesis that a method capable of performing well on multiple tasks will generalize well to a previously unseen task and potentially outperform a custom-designed solution. MSD results confirmed this hypothesis, moreover, MSD winner continued generalizing well to a wide range of other clinical problems for the next two years. Three main conclusions can be drawn from this study: (1) state-of-the-art image segmentation algorithms generalize well when retrained on unseen tasks; (2) consistent algorithmic performance across multiple tasks is a strong surrogate of algorithmic generalizability; (3) the training of accurate AI segmentation models is now commoditized to scientists that are not versed in AI model training.
Authors 58
-
Michela Antonelli corresponding
Affiliation as printed
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK. michela.antonelli@kcl.ac.uk
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK
-
German Cancer Research Center · Heidelberg University
Affiliation as printed
Div. Computer Assisted Medical Interventions, German Cancer Research Center (DKFZ), Heidelberg, Germany
Faculty of Mathematics and Computer Science, University of Heidelberg, Heidelberg, Germany
HI Helmholtz Imaging, German Cancer Research Center (DKFZ), Heidelberg, Germany
-
Affiliation as printed
Center for Biomedical Image Computing and Analytics (CBICA), University of Pennsylvania, Philadelphia, PA, USA
Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
-
National Institutes of Health · National Cancer Institute
Affiliation as printed
Center for Biomedical Informatics and Information Technology, National Cancer Institute (NIH), Bethesda, MD, USA
-
Affiliation as printed
Div. Biostatistics, German Cancer Research Center (DKFZ), Heidelberg, Germany
-
Affiliation as printed
Electrical Engineering and Computer Science, Vanderbilt University, Nashville, TN, USA
-
Radboud University Medical Center · Radboud Institute for Health Sciences
Affiliation as printed
Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, The Netherlands
-
Affiliation as printed
Quantitative Biomedicine, University of Zurich, Zurich, Switzerland
-
Google DeepMind (United Kingdom)
Affiliation as printed
DeepMind, London, UK
-
National Institutes of Health · National Institutes of Health Clinical Center
Affiliation as printed
Imaging Biomarkers and Computer-Aided Diagnosis Laboratory, Department of Radiology and Imaging Sciences, National Institutes of Health Clinical Center (NIH), Bethesda, MD, USA
-
Radboud University Medical Center · Radboud Institute for Health Sciences
Affiliation as printed
Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, The Netherlands
-
Affiliation as printed
Center for Biomedical Image Computing and Analytics (CBICA), University of Pennsylvania, Philadelphia, PA, USA
-
Technical University of Munich
Affiliation as printed
Department of Informatics, Technische Universität München, München, Germany
-
Technical University of Munich
Affiliation as printed
Department of Informatics, Technische Universität München, München, Germany
-
Memorial Sloan Kettering Cancer Center
Affiliation as printed
Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
-
Memorial Sloan Kettering Cancer Center
Affiliation as printed
Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
-
Vanderbilt University Medical Center
Affiliation as printed
Department of Psychiatry & Behavioral Sciences, Vanderbilt University Medical Center, Nashville, TN, USA
-
Radboud University Medical Center · Radboud Institute for Health Sciences
Affiliation as printed
Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, The Netherlands
-
Memorial Sloan Kettering Cancer Center
Affiliation as printed
Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, NY, USA
-
Vanderbilt University Medical Center
Affiliation as printed
Department of Psychiatry & Behavioral Sciences, Vanderbilt University Medical Center, Nashville, TN, USA
-
Stanford Medicine · Stanford University
Affiliation as printed
Department of Radiology, Stanford University, Stanford, CA, USA
-
Memorial Sloan Kettering Cancer Center
Affiliation as printed
Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
-
Affiliation as printed
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK
-
Affiliation as printed
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK
-
Affiliation as printed
Department of Computer Science and Software Engineering, École Polytechnique de Montréal, Montréal, QC, Canada
-
Radboud University Medical Center · Radboud Institute for Health Sciences
Affiliation as printed
Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, The Netherlands
-
Affiliation as printed
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK
-
Affiliation as printed
Div. Biostatistics, German Cancer Research Center (DKFZ), Heidelberg, Germany
-
Affiliation as printed
Universidad de los Andes, Bogota, Colombia
-
Affiliation as printed
VUNO Inc., Seoul, Korea
-
Affiliation as printed
Tencent Jarvis Lab, Shenzhen, China
-
Affiliation as printed
Universidad de los Andes, Bogota, Colombia
-
Affiliation as printed
Department of Automation, Tsinghua University, Beijing, China
-
Chinese Academy of Sciences · Shenzhen Institutes of Advanced Technology
Affiliation as printed
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
-
Affiliation as printed
HI Applied Computer Vision Lab, Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, Germany
-
Affiliation as printed
Department of Computer Science, Xiamen University, Xiamen, China
-
Chinese Academy of Sciences · Shenzhen Institutes of Advanced Technology
Affiliation as printed
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
-
Affiliation as printed
Kakao Brain, Seongnam-si, Republic of Korea
-
Heidelberg University · University Hospital Heidelberg
Affiliation as printed
Cerebriu A/S, Copenhagen, Denmark
Pattern Analysis and Learning Group, Department of Radiation Oncology, Heidelberg University Hospital, Heidelberg, Germany
-
Fraunhofer Institute for Digital Medicine · RWTH Aachen University
Affiliation as printed
Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany
Institute of Imaging & Computer Vision, RWTH Aachen University, Aachen, Germany
-
Affiliation as printed
Cerebriu A/S, Copenhagen, Denmark
Department of Computer Science, University of Copenhagen, Copenhagen, Denmark
-
Affiliation as printed
VUNO Inc., Seoul, Korea
-
Affiliation as printed
Department of Computer Science, University of Copenhagen, Copenhagen, Denmark
-
Affiliation as printed
MaaDoTaa.com, San Diego, CA, USA
-
Affiliation as printed
Institute of Imaging & Computer Vision, RWTH Aachen University, Aachen, Germany
-
LMU Klinikum · Ludwig-Maximilians-Universität München
Affiliation as printed
Lab for Artificial Intelligence in Medical Imaging (AI-Med), Department of Child and Adolescent Psychiatry, University Hospital, LMU München, Germany
-
Affiliation as printed
MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University, Shanghai, China
-
Affiliation as printed
VUNO Inc., Seoul, Korea
-
LMU Klinikum · Ludwig-Maximilians-Universität München
Affiliation as printed
Lab for Artificial Intelligence in Medical Imaging (AI-Med), Department of Child and Adolescent Psychiatry, University Hospital, LMU München, Germany
-
Affiliation as printed
Department of Computer Science, Xiamen University, Xiamen, China
-
Affiliation as printed
Shanghai Key Laboratory of Multidimensional Information Processing, East China Normal University, Shanghai, China
-
Affiliation as printed
Johns Hopkins University, Baltimore, MD, USA
-
Affiliation as printed
NVIDIA, Santa Clara, CA, USA
-
Affiliation as printed
Department of Automation, Tsinghua University, Beijing, China
-
Affiliation as printed
Tencent Jarvis Lab, Shenzhen, China
-
Affiliation as printed
School of Computing/Department of Biomedical and Molecular Sciences, Queen's University, Kingston, ON, Canada
-
German Cancer Research Center · Heidelberg University
Affiliation as printed
Div. Computer Assisted Medical Interventions, German Cancer Research Center (DKFZ), Heidelberg, Germany
Faculty of Mathematics and Computer Science, University of Heidelberg, Heidelberg, Germany
HI Helmholtz Imaging, German Cancer Research Center (DKFZ), Heidelberg, Germany
Medical Faculty, University of Heidelberg, Heidelberg, Germany
-
Affiliation as printed
School of Biomedical Engineering & Imaging Sciences, King's College London, London, UK
Cited by 1,329 stored of 1,340
Cited by patents worldwide 5 (Lens.org)
-
Learning from partially labeled datasets for medical imaging analysisUS12688928B2 2026-07-21 Active
-
Adaptive token sampling for efficient transformerUS12306906B2 2025-05-20 Active
-
SYSTÈME ET PROCÉDÉ COMBINÉ DE SÉLECTION, D’ANNOTATION ET D’ENTRAÎNEMENT VIA UN MODÈLE D’APPRENTISSAGE AUTOMATIQUE PARTAGÉFR3146529A1 2024-09-13 Pending
-
TRAINING A VOLUMETRIC MODEL ON SPARSELY ANNOTATED TRAINING DATAWO2024056793A1 2024-03-21 Pending