Open Platform for the De-identification of Burned-in Texts in Medical Images using Deep Learning
(2024) 11th International Conference on Bioimaging (BIOIMAGING 2024) — Location: Rome, Italy (21.February.2024)
Files
2024-01-07-BIOSTEC-BIOIMAGING-ImageDeidentification-Final.pdf
Author preprint
Open Access - Adobe PDF
- 4.35 MB
Details
- Authors
- Abstract
- While the de-identification of DICOM tags is a standardized, well-established practice, the removal of protected health information burned into the pixels of medical images is a more complex challenge for which Deep Learning is especially well adapted. Unfortunately, there is currently a lack of accurate, effective, and freely available tools to this end. This motivates the release of a new benchmark dataset, together with free and open-source software leveraging dedicated Deep Learning algorithms, with the goal of improving patient confidentiality. The proposed methods consist of adapting scene-text detection models (SSD and TextBoxes) to the task of image de-identification. Results have shown that fine-tuning such generic text detection models on medical images significantly improves performance. The developed algorithms can be applied either from the command line or using a Web interface that is tightly integrated with a free and open-source PACS server.
- Affiliations
Citations
Langlois, Q., Szelagowski, N., Vanderdonckt, J., & Jodogne, S. (2024). Open Platform for the De-identification of Burned-in Texts in Medical Images using Deep Learning. Proc. of the 17th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2024), 1, 297-304. https://doi.org/10.5220/0012430300003657 (Original work published 2024)