Algorithmic analysis of the structure of mixed odontogenic tumors
Document Type
Article
Publication Date
12-1-2026
Abstract
This study aims to characterize the histomorphology of mixed odontogenic tumors, using mathematical morphology algorithms applied to digital images. Five cases of primordial odontogenic tumor (POT), 5 cases of ameloblastic fibroma, 5 cases of developing odontoma (DO), and 5 cases of tooth germs (TG) were analyzed. Histological sections stained with Haematoxylin & Eosin were digitized and the epithelial compartments were segmented into ‘virtual cells’ to further characterize the tissue compartment architecture. A comparison of the mean area of virtual epithelial cells in the entities investigated showed that, despite data distribution between the entities being similar, statistically significant differences (p < 0.001) were found, being larger for DO and smaller for AF. Additionally, DO exhibits a broader data distribution of the area compared to the other entities. Significant differences were not found between TG and POT without subepithelial condensation. Quantitative tissue analysis showed that, in focal areas, POT more closely resembles TG than other mixed odontogenic tumors. These findings suggest that virtual cell–based morphometric analysis may provide complementary quantitative information in diagnostically challenging cases, although validation in larger datasets is required.
Keywords
Algorithmic analysis, Ameloblastic fibroma, Mixed odontogenic tumors, Primordial odontogenic tumor, Tooth germ
Publication Title
Scientific Reports
ISSN
2045-2322
DOI
10.1038/s41598-026-38399-6
Recommended Citation
Pereira-Prado, Vanesa; Sicco, Estefanía; Silveira, Felipe Martins; Schuch, Lauren Frenzel; Hunter, Keith; Niklander, Sven Eric; Tilakaratne, Wanninayake Mudiyanselage; Gomez, Ricardo Santiago; Landini, Gabriel; and Bologna-Molina, Ronell, "Algorithmic analysis of the structure of mixed odontogenic tumors" (2026). Research Publications (2026 to 2030). 467.
https://knova.um.edu.my/research_publications_2026_2030/467
Volume
16
Issue
1