- Author
- Artuur Couckuyt (UGent) , Ruth Seurinck (UGent) , Annelies Emmaneel (UGent) , Katrien Quintelier (UGent) , David Novák (UGent) , Sofie Van Gassen (UGent) and Yvan Saeys (UGent)
- Organization
- Project
-
- Research Programme Artificial Intelligence - 2022
- ProFILE - Standardized multi-parameter flow cytometry as tool for early diagnosis, risk assessment and appropriate treatment in primary immunodefiency
- Improving the Diagnosis and Prognosis of Primary Immunodeficiencies using Computational Flow Cytometry
- How T cell autoreactivity develops in early rheumatoid arthritis: the smoking gun hypothesis
- Abstract
- Machine learning (ML) algorithms are increasingly being used to help implement clinical decision support systems. In this new field, we define as “translational machine learning”, joint efforts and strong communication between data scientists and clinicians help to span the gap between ML and its adoption in the clinic. These collaborations also improve interpretability and trust in translational ML methods and ultimately aim to result in generalizable and reproducible models. To help clinicians and bioinformaticians refine their translational ML pipelines, we review the steps from model building to the use of ML in the clinic. We discuss experimental setup, computational analysis, interpretability and reproducibility, and emphasize the challenges involved. We highly advise collaboration and data sharing between consortia and institutes to build multi-centric cohorts that facilitate ML methodologies that generalize across centers. In the end, we hope that this review provides a way to streamline translational ML and helps to tackle the challenges that come with it.
- Keywords
- Genetics (clinical), Genetics, ARTIFICIAL-INTELLIGENCE, DIABETIC-RETINOPATHY, EXPERIMENTAL-DESIGN, PREDICTION MODEL, CANCER, VALIDATION, SELECTION, HEALTH, FLOW, EXPLANATION
Downloads
-
Couckuyt2022 Article ChallengesInTranslationalMachi.pdf
- full text (Published version)
- |
- open access
- |
- |
- 6.98 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8744107
- MLA
- Couckuyt, Artuur, et al. “Challenges in Translational Machine Learning.” HUMAN GENETICS, vol. 141, 2022, pp. 1451–66, doi:10.1007/s00439-022-02439-8.
- APA
- Couckuyt, A., Seurinck, R., Emmaneel, A., Quintelier, K., Novák, D., Van Gassen, S., & Saeys, Y. (2022). Challenges in translational machine learning. HUMAN GENETICS, 141, 1451–1466. https://doi.org/10.1007/s00439-022-02439-8
- Chicago author-date
- Couckuyt, Artuur, Ruth Seurinck, Annelies Emmaneel, Katrien Quintelier, David Novák, Sofie Van Gassen, and Yvan Saeys. 2022. “Challenges in Translational Machine Learning.” HUMAN GENETICS 141: 1451–66. https://doi.org/10.1007/s00439-022-02439-8.
- Chicago author-date (all authors)
- Couckuyt, Artuur, Ruth Seurinck, Annelies Emmaneel, Katrien Quintelier, David Novák, Sofie Van Gassen, and Yvan Saeys. 2022. “Challenges in Translational Machine Learning.” HUMAN GENETICS 141: 1451–1466. doi:10.1007/s00439-022-02439-8.
- Vancouver
- 1.Couckuyt A, Seurinck R, Emmaneel A, Quintelier K, Novák D, Van Gassen S, et al. Challenges in translational machine learning. HUMAN GENETICS. 2022;141:1451–66.
- IEEE
- [1]A. Couckuyt et al., “Challenges in translational machine learning,” HUMAN GENETICS, vol. 141, pp. 1451–1466, 2022.
@article{8744107, abstract = {{Machine learning (ML) algorithms are increasingly being used to help implement clinical decision support systems. In this new field, we define as “translational machine learning”, joint efforts and strong communication between data scientists and clinicians help to span the gap between ML and its adoption in the clinic. These collaborations also improve interpretability and trust in translational ML methods and ultimately aim to result in generalizable and reproducible models. To help clinicians and bioinformaticians refine their translational ML pipelines, we review the steps from model building to the use of ML in the clinic. We discuss experimental setup, computational analysis, interpretability and reproducibility, and emphasize the challenges involved. We highly advise collaboration and data sharing between consortia and institutes to build multi-centric cohorts that facilitate ML methodologies that generalize across centers. In the end, we hope that this review provides a way to streamline translational ML and helps to tackle the challenges that come with it.}}, author = {{Couckuyt, Artuur and Seurinck, Ruth and Emmaneel, Annelies and Quintelier, Katrien and Novák, David and Van Gassen, Sofie and Saeys, Yvan}}, issn = {{0340-6717}}, journal = {{HUMAN GENETICS}}, keywords = {{Genetics (clinical),Genetics,ARTIFICIAL-INTELLIGENCE,DIABETIC-RETINOPATHY,EXPERIMENTAL-DESIGN,PREDICTION MODEL,CANCER,VALIDATION,SELECTION,HEALTH,FLOW,EXPLANATION}}, language = {{eng}}, pages = {{1451--1466}}, title = {{Challenges in translational machine learning}}, url = {{http://doi.org/10.1007/s00439-022-02439-8}}, volume = {{141}}, year = {{2022}}, }
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: