Acute deep neck infection MRI: deep learning segmentation and clinical relevance of retropharyngeal edema volume
RPE volume correlated with markers of severe illness and outperformed binary RPE classification. We developed a DL-based algorithm for slice-wise classification and automatic segmentation of RPE. The classification model achieved excellent performance, while segmentation yielded modest Dice similarity coefficients consistent with prior imaging-based tumor segmentation algorithms.
Authors
- Aapo Sirén (ORCID: https://orcid.org/0000-0002-6822-1171)
- Oona Rainio (ORCID: https://orcid.org/0000-0002-7775-7656)
- Mikko J. Nyman (ORCID: https://orcid.org/0000-0002-2104-6302)
- Heidi Huhtanen (ORCID: https://orcid.org/0000-0002-1581-1709)
- Riku Klén
- J. Hirvonen (ORCID: https://orcid.org/0000-0002-0368-3746)
- Ville Sakari Viertonen (ORCID: https://orcid.org/0009-0009-3575-6979)
Institutions
- University of Turku (FI)
- Turku University Hospital (FI)
- Turku PET Centre (FI)
- Tampere University Hospital (FI)
- Tampere University (FI)
Publication Details
- Journal
- European Radiology Experimental
- Published
- 2026-02-23
- DOI
- https://doi.org/10.1186/s41747-026-00686-2
- Primary Topic
- Otolaryngology and Infectious Diseases
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Sigrid Juséliuksen Säätiö