Equipping computational pathology systems with artifact processing pipelines: a showcase for computation and performance trade-offs
The proposed artifact detection pipeline will not only ensure reliable CPATH predictions but may also provide quality control. In this work, the best-performing pipeline for artifact detection is MoE with DCNNs. Our detailed experiments show that there is always a trade-off between performance and computational complexity, and no straightforward DL solution equally suits all types of data and applications. The code and HistoArtifacts dataset can be found online at Github and Zenodo , respectively.
Authors
- Umay Kiraz (ORCID: https://orcid.org/0000-0002-6721-4877)
- Tahlita C.M. Zuiverloon
- Carlos Couceiro Monteagudo (ORCID: https://orcid.org/0000-0002-9381-9976)
- Farbod Khoraminia (ORCID: https://orcid.org/0000-0002-2817-3996)
- Kjersti Engan (ORCID: https://orcid.org/0000-0002-8970-0067)
- Neel Kanwal (ORCID: https://orcid.org/0000-0002-8115-0558)
- Andrés Mosquera‐Zamudio (ORCID: https://orcid.org/0000-0002-6158-9619)
- Chunming Rong (ORCID: https://orcid.org/0000-0002-8347-0539)
- Emiel A. M. Janssen
Institutions
- Universitat de València (ES)
- Stavanger University Hospital (NO)
- Erasmus MC Cancer Institute (NL)
- INCLIVA Health Research Institute (ES)
- University of Stavanger (NO)
Publication Details
- Journal
- BMC Medical Informatics and Decision Making
- Published
- 2024-10-07
- DOI
- https://doi.org/10.1186/s12911-024-02676-z
- Citations
- 11
- Primary Topic
- AI in cancer detection
- Type
- article
- Field-Weighted Citation Impact
- 3.82
Funders
- European Commission
- Horizon 2020 Framework Programme