AI Models Show Promise for Improving Diagnosis and Prognosis of Recurrent Jaw Cysts
-
Odontogenic keratocyst (OKC) is a cystic lesion of the jaws with a high recurrence rate after surgery. The study developed AI models for diagnosis and prognosis of OKC.
-
The diagnostic model differentiated OKC from orthokeratinized odontogenic cyst (OOC) and Gorlin syndrome (GS) with high accuracy, but some confusion existed between OKC and GS.
-
The prognostic model predicted recurrence of sporadic OKC with good performance. Multiple slides fusion was better than single slide for recurrence prediction.
-
AI features correlated significantly with OKC pathological findings like daughter cysts, epithelial proliferation, inflammation, unilocular/multilocular, and basal cell proliferation.
-
The AI systems demonstrated excellent testing performance, evidencing their potential for assisting in OKC clinical management for diagnosis, recurrence prediction, and unraveling disease insights.