Machine learning-based risk prediction for major adverse cardiovascular events in a Brazilian hospital: Development, external validation, and interpretability
Among the machine learning algorithms evaluated, Random Forest showed the best generalization ability, both internally and externally. Shapley values for local interpretability were more informative than LIME ones, which is in line with our exploratory analysis and global interpretation of the final model. Machine learning algorithms with good generalization and accompanied by interpretability analyses are recommended for assessments of individual risks of cardiovascular diseases and development of personalized preventive actions.
Authors
- Gilson Shimizu (ORCID: https://orcid.org/0000-0003-3711-5592)
- Hilton Vicente César (ORCID: https://orcid.org/0000-0002-2701-8898)
- Peter P. Rainer (ORCID: https://orcid.org/0000-0002-2840-6454)
- Stefanie Jauk (ORCID: https://orcid.org/0000-0003-1287-594X)
- Kátia Mitiko Firmino Suzuki
- João Mazzoncini de Azevedo Marques (ORCID: https://orcid.org/0000-0002-3100-3883)
- Paulo Mazzoncini de Azevedo‐Marques (ORCID: https://orcid.org/0000-0002-7271-2774)
- Diether Kramer
- Michael Schrempf
- José Abrão Cardeal da Costa (ORCID: https://orcid.org/0000-0002-7400-5598)
- Elen Almeida Romão (ORCID: https://orcid.org/0000-0002-9302-3089)
- Sandro Scarpelini
Institutions
- Universidade de São Paulo (BR)
- Medical University of Graz (AT)
Publication Details
- Journal
- PLoS ONE
- Published
- 2024-10-11
- DOI
- https://doi.org/10.1371/journal.pone.0311719
- Citations
- 4
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 1.39
Funders
- Fundação de Amparo à Pesquisa do Estado de São Paulo
- Conselho Nacional de Desenvolvimento Científico e Tecnológico