A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm
Distributed learning is a promising alternative to central learning for machine learning (ML) model training, overcoming data-sharing problems in healthcare. Previous studies exploring federated learning (FL) or the traveling model (TM) setup for medical image-based disease classification often reli...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2024-02-01
|
| Series: | Frontiers in Artificial Intelligence |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/frai.2024.1301997/full |
| _version_ | 1827356643703128064 |
|---|---|
| author | Raissa Souza Raissa Souza Raissa Souza Raissa Souza Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Milton Camacho Milton Camacho Milton Camacho Milton Camacho Richard Camicioli Oury Monchi Oury Monchi Oury Monchi Oury Monchi Oury Monchi Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Matthias Wilms Matthias Wilms Matthias Wilms Matthias Wilms Nils D. Forkert Nils D. Forkert Nils D. Forkert Nils D. Forkert |
| author_facet | Raissa Souza Raissa Souza Raissa Souza Raissa Souza Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Milton Camacho Milton Camacho Milton Camacho Milton Camacho Richard Camicioli Oury Monchi Oury Monchi Oury Monchi Oury Monchi Oury Monchi Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Matthias Wilms Matthias Wilms Matthias Wilms Matthias Wilms Nils D. Forkert Nils D. Forkert Nils D. Forkert Nils D. Forkert |
| author_sort | Raissa Souza |
| collection | DOAJ |
| description | Distributed learning is a promising alternative to central learning for machine learning (ML) model training, overcoming data-sharing problems in healthcare. Previous studies exploring federated learning (FL) or the traveling model (TM) setup for medical image-based disease classification often relied on large databases with a limited number of centers or simulated artificial centers, raising doubts about real-world applicability. This study develops and evaluates a convolution neural network (CNN) for Parkinson's disease classification using data acquired by 83 diverse real centers around the world, mostly contributing small training samples. Our approach specifically makes use of the TM setup, which has proven effective in scenarios with limited data availability but has never been used for image-based disease classification. Our findings reveal that TM is effective for training CNN models, even in complex real-world scenarios with variable data distributions. After sufficient training cycles, the TM-trained CNN matches or slightly surpasses the performance of the centrally trained counterpart (AUROC of 83% vs. 80%). Our study highlights, for the first time, the effectiveness of TM in 3D medical image classification, especially in scenarios with limited training samples and heterogeneous distributed data. These insights are relevant for situations where ML models are supposed to be trained using data from small or remote medical centers, and rare diseases with sparse cases. The simplicity of this approach enables a broad application to many deep learning tasks, enhancing its clinical utility across various contexts and medical facilities. |
| first_indexed | 2024-03-08T05:12:52Z |
| format | Article |
| id | doaj.art-29d7e62d83354025a6e62fa38cfd298f |
| institution | Directory Open Access Journal |
| issn | 2624-8212 |
| language | English |
| last_indexed | 2024-03-08T05:12:52Z |
| publishDate | 2024-02-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Artificial Intelligence |
| spelling | doaj.art-29d7e62d83354025a6e62fa38cfd298f2024-02-07T05:24:31ZengFrontiers Media S.A.Frontiers in Artificial Intelligence2624-82122024-02-01710.3389/frai.2024.13019971301997A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigmRaissa Souza0Raissa Souza1Raissa Souza2Raissa Souza3Emma A. M. Stanley4Emma A. M. Stanley5Emma A. M. Stanley6Emma A. M. Stanley7Milton Camacho8Milton Camacho9Milton Camacho10Milton Camacho11Richard Camicioli12Oury Monchi13Oury Monchi14Oury Monchi15Oury Monchi16Oury Monchi17Zahinoor Ismail18Zahinoor Ismail19Zahinoor Ismail20Zahinoor Ismail21Matthias Wilms22Matthias Wilms23Matthias Wilms24Matthias Wilms25Nils D. Forkert26Nils D. Forkert27Nils D. Forkert28Nils D. Forkert29Department of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaBiomedical Engineering Graduate Program, University of Calgary, Calgary, AB, CanadaAlberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, CanadaDepartment of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaBiomedical Engineering Graduate Program, University of Calgary, Calgary, AB, CanadaAlberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, CanadaDepartment of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaBiomedical Engineering Graduate Program, University of Calgary, Calgary, AB, CanadaAlberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, CanadaDepartment of Medicine (Neurology), Neuroscience and Mental Health Institute, University of Alberta, Edmonton, AB, CanadaDepartment of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaDepartment of Radiology, Radio-oncology and Nuclear Medicine, Université de Montréal, Montréal, QC, CanadaCentre de Recherche, Institut Universitaire de Gériatrie de Montréal, Montréal, QC, CanadaDepartment of Clinical Neurosciences, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaDepartment of Clinical Neurosciences, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaDepartment of Psychiatry, University of Calgary, Calgary, AB, Canada0Clinical and Biomedical Sciences, Faculty of Health and Life Sciences, University of Exeter, Exeter, United KingdomHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaAlberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, Canada1Department of Pediatrics, University of Calgary, Calgary, AB, Canada2Department of Community Health Sciences, University of Calgary, Calgary, AB, CanadaDepartment of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaHotchkiss Brain Institute, University of Calgary, Calgary, AB, CanadaAlberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, CanadaDepartment of Clinical Neurosciences, Cumming School of Medicine, University of Calgary, Calgary, AB, CanadaDistributed learning is a promising alternative to central learning for machine learning (ML) model training, overcoming data-sharing problems in healthcare. Previous studies exploring federated learning (FL) or the traveling model (TM) setup for medical image-based disease classification often relied on large databases with a limited number of centers or simulated artificial centers, raising doubts about real-world applicability. This study develops and evaluates a convolution neural network (CNN) for Parkinson's disease classification using data acquired by 83 diverse real centers around the world, mostly contributing small training samples. Our approach specifically makes use of the TM setup, which has proven effective in scenarios with limited data availability but has never been used for image-based disease classification. Our findings reveal that TM is effective for training CNN models, even in complex real-world scenarios with variable data distributions. After sufficient training cycles, the TM-trained CNN matches or slightly surpasses the performance of the centrally trained counterpart (AUROC of 83% vs. 80%). Our study highlights, for the first time, the effectiveness of TM in 3D medical image classification, especially in scenarios with limited training samples and heterogeneous distributed data. These insights are relevant for situations where ML models are supposed to be trained using data from small or remote medical centers, and rare diseases with sparse cases. The simplicity of this approach enables a broad application to many deep learning tasks, enhancing its clinical utility across various contexts and medical facilities.https://www.frontiersin.org/articles/10.3389/frai.2024.1301997/fulltraveling modelfederated learningdistributed learningParkinson's diseasemulti-center |
| spellingShingle | Raissa Souza Raissa Souza Raissa Souza Raissa Souza Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Emma A. M. Stanley Milton Camacho Milton Camacho Milton Camacho Milton Camacho Richard Camicioli Oury Monchi Oury Monchi Oury Monchi Oury Monchi Oury Monchi Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Zahinoor Ismail Matthias Wilms Matthias Wilms Matthias Wilms Matthias Wilms Nils D. Forkert Nils D. Forkert Nils D. Forkert Nils D. Forkert A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm Frontiers in Artificial Intelligence traveling model federated learning distributed learning Parkinson's disease multi-center |
| title | A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm |
| title_full | A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm |
| title_fullStr | A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm |
| title_full_unstemmed | A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm |
| title_short | A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm |
| title_sort | multi center distributed learning approach for parkinson s disease classification using the traveling model paradigm |
| topic | traveling model federated learning distributed learning Parkinson's disease multi-center |
| url | https://www.frontiersin.org/articles/10.3389/frai.2024.1301997/full |
| work_keys_str_mv | AT raissasouza amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT richardcamicioli amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert amulticenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT raissasouza multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT emmaamstanley multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT miltoncamacho multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT richardcamicioli multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT ourymonchi multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT zahinoorismail multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT matthiaswilms multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm AT nilsdforkert multicenterdistributedlearningapproachforparkinsonsdiseaseclassificationusingthetravelingmodelparadigm |