GeneralAI / InformaticsResearch
Added perilesional radiomic features improve odontogenic cyst classification on panoramic radiographs
Journal of clinical medicine2w ago
Radiomics on panoramic radiographs moderately differentiated odontogenic cysts; adding perilesional/contextual features raised four-class balanced accuracy from 0.416 to 0.487 (p=0.002). Best pairwise accuracy was 0.838 for dentigerous versus unspecified odontogenic cyst.
- Study included 63 patients with histopathologically confirmed cysts: odontogenic keratocyst (33), radicular (12), dentigerous (10), and unspecified odontogenic cyst (8).
- Texture features, especially neighborhood gray-tone difference matrix metrics, were most informative; pairwise accuracy was low for keratocyst vs radicular cyst (0.602) and keratocyst vs dentigerous (0.644).
- Substantial overlap between cyst types persists; authors call for larger validation cohorts and multimodal approaches.
Automated summary
RadPigeon summaries are original and for information only. They are not clinical advice.