Diagnostic Performance of a Deep Learning-Based Tool for the Detection and Staging of Rectal Cancers on Endoscopic Ultrasound: Prospective Study
: The DL tool shows promise for aiding operators in identifying rectal lesions suitable for endoscopic resection. The semi-supervised training approach reduces manual segmentation burden and achieves performance comparable to expert physicians.
Authors
- Lucio Molini
- Andrea Ghezzi (ORCID: https://orcid.org/0000-0002-6073-0107)
- Marco Valvano (ORCID: https://orcid.org/0000-0002-2201-6279)
- Nicola Panvini (ORCID: https://orcid.org/0000-0002-0967-7101)
- Mariangela Rutigliani (ORCID: https://orcid.org/0000-0002-5773-0309)
- Amedeo Montale (ORCID: https://orcid.org/0000-0003-4213-1917)
- M. Oppezzi
- Francesco Paparo (ORCID: https://orcid.org/0000-0002-2050-8966)
- Marco Magliocco
Institutions
- Ente Ospedaliero Ospedali Galliera (IT)
- University of Genoa (IT)
Publication Details
- Journal
- Diagnostics
- Published
- 2026-04-14
- DOI
- https://doi.org/10.3390/diagnostics16081161
- Primary Topic
- Colorectal Cancer Surgical Treatments
- Type
- article
- Field-Weighted Citation Impact
- 0.00