Continual learning framework for a multicenter study with an application to electrocardiogram
Abstract Deep learning has been increasingly utilized in the medical field and achieved many goals. Since the size of data dominates the performance of deep learning, several medical institutions are conducting joint research to obtain as much data as possible. However, sharing data is usually prohi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2024-03-01
|
| Series: | BMC Medical Informatics and Decision Making |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12911-024-02464-9 |
| _version_ | 1827139370988077056 |
|---|---|
| author | Junmo Kim Min Hyuk Lim Kwangsoo Kim Hyung-Jin Yoon |
| author_facet | Junmo Kim Min Hyuk Lim Kwangsoo Kim Hyung-Jin Yoon |
| author_sort | Junmo Kim |
| collection | DOAJ |
| description | Abstract Deep learning has been increasingly utilized in the medical field and achieved many goals. Since the size of data dominates the performance of deep learning, several medical institutions are conducting joint research to obtain as much data as possible. However, sharing data is usually prohibited owing to the risk of privacy invasion. Federated learning is a reasonable idea to train distributed multicenter data without direct access; however, a central server to merge and distribute models is needed, which is expensive and hardly approved due to various legal regulations. This paper proposes a continual learning framework for a multicenter study, which does not require a central server and can prevent catastrophic forgetting of previously trained knowledge. The proposed framework contains the continual learning method selection process, assuming that a single method is not omnipotent for all involved datasets in a real-world setting and that there could be a proper method to be selected for specific data. We utilized the fake data based on a generative adversarial network to evaluate methods prospectively, not ex post facto. We used four independent electrocardiogram datasets for a multicenter study and trained the arrhythmia detection model. Our proposed framework was evaluated against supervised and federated learning methods, as well as finetuning approaches that do not include any regulation to preserve previous knowledge. Even without a central server and access to the past data, our framework achieved stable performance (AUROC 0.897) across all involved datasets, achieving comparable performance to federated learning (AUROC 0.901). |
| first_indexed | 2024-04-25T01:06:31Z |
| format | Article |
| id | doaj.art-e0476afe6ca4426faeea518a761b9924 |
| institution | Directory Open Access Journal |
| issn | 1472-6947 |
| language | English |
| last_indexed | 2025-03-20T18:38:10Z |
| publishDate | 2024-03-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Informatics and Decision Making |
| spelling | doaj.art-e0476afe6ca4426faeea518a761b99242024-08-25T11:23:17ZengBMCBMC Medical Informatics and Decision Making1472-69472024-03-0124111310.1186/s12911-024-02464-9Continual learning framework for a multicenter study with an application to electrocardiogramJunmo Kim0Min Hyuk Lim1Kwangsoo Kim2Hyung-Jin Yoon3Interdisciplinary Program in Bioengineering, Seoul National UniversityTransdisciplinary Department of Medicine and Advanced Technology, Seoul National University HospitalTransdisciplinary Department of Medicine and Advanced Technology, Seoul National University HospitalInterdisciplinary Program in Bioengineering, Seoul National UniversityAbstract Deep learning has been increasingly utilized in the medical field and achieved many goals. Since the size of data dominates the performance of deep learning, several medical institutions are conducting joint research to obtain as much data as possible. However, sharing data is usually prohibited owing to the risk of privacy invasion. Federated learning is a reasonable idea to train distributed multicenter data without direct access; however, a central server to merge and distribute models is needed, which is expensive and hardly approved due to various legal regulations. This paper proposes a continual learning framework for a multicenter study, which does not require a central server and can prevent catastrophic forgetting of previously trained knowledge. The proposed framework contains the continual learning method selection process, assuming that a single method is not omnipotent for all involved datasets in a real-world setting and that there could be a proper method to be selected for specific data. We utilized the fake data based on a generative adversarial network to evaluate methods prospectively, not ex post facto. We used four independent electrocardiogram datasets for a multicenter study and trained the arrhythmia detection model. Our proposed framework was evaluated against supervised and federated learning methods, as well as finetuning approaches that do not include any regulation to preserve previous knowledge. Even without a central server and access to the past data, our framework achieved stable performance (AUROC 0.897) across all involved datasets, achieving comparable performance to federated learning (AUROC 0.901).https://doi.org/10.1186/s12911-024-02464-9Multicenter studyDeep learningContinual learningElectrocardiogram |
| spellingShingle | Junmo Kim Min Hyuk Lim Kwangsoo Kim Hyung-Jin Yoon Continual learning framework for a multicenter study with an application to electrocardiogram BMC Medical Informatics and Decision Making Multicenter study Deep learning Continual learning Electrocardiogram |
| title | Continual learning framework for a multicenter study with an application to electrocardiogram |
| title_full | Continual learning framework for a multicenter study with an application to electrocardiogram |
| title_fullStr | Continual learning framework for a multicenter study with an application to electrocardiogram |
| title_full_unstemmed | Continual learning framework for a multicenter study with an application to electrocardiogram |
| title_short | Continual learning framework for a multicenter study with an application to electrocardiogram |
| title_sort | continual learning framework for a multicenter study with an application to electrocardiogram |
| topic | Multicenter study Deep learning Continual learning Electrocardiogram |
| url | https://doi.org/10.1186/s12911-024-02464-9 |
| work_keys_str_mv | AT junmokim continuallearningframeworkforamulticenterstudywithanapplicationtoelectrocardiogram AT minhyuklim continuallearningframeworkforamulticenterstudywithanapplicationtoelectrocardiogram AT kwangsookim continuallearningframeworkforamulticenterstudywithanapplicationtoelectrocardiogram AT hyungjinyoon continuallearningframeworkforamulticenterstudywithanapplicationtoelectrocardiogram |