Explainability and uncertainty: Two sides of the same coin for enhancing the interpretability of deep learning models in healthcare
Our analysis indicates that integrating UQ with XAI could significantly enhance the reliability of DL models in practice. This approach has the potential to reduce interpretation biases and over-reliance, leading to more cautious and conscious use of AI in healthcare.
Authors
- Arkadiusz Gertych (ORCID: https://orcid.org/0000-0002-3107-602X)
- Federico Cabitza (ORCID: https://orcid.org/0000-0002-4065-3415)
- Andrea Campagner (ORCID: https://orcid.org/0000-0002-0027-5157)
- Massimo Salvi (ORCID: https://orcid.org/0000-0001-7225-7401)
- Filippo Molinari (ORCID: https://orcid.org/0000-0003-1150-2244)
- Silvia Seoni (ORCID: https://orcid.org/0000-0001-5372-2654)
- U. Rajendra Acharya (ORCID: https://orcid.org/0000-0003-2689-8552)
Institutions
- Cedars-Sinai Medical Center (US)
- Politecnico di Torino (IT)
- University of Southern Queensland (AU)
- Istituto Clinico Sant'Ambrogio (IT)
- University of Milano-Bicocca (IT)
Publication Details
- Journal
- International Journal of Medical Informatics
- Published
- 2025-02-21
- DOI
- https://doi.org/10.1016/j.ijmedinf.2025.105846
- Citations
- 36
- Primary Topic
- Explainable Artificial Intelligence (XAI)
- Type
- article
- Field-Weighted Citation Impact
- 89.31