Fig. 6
From: Development and validation of an AI-enabled oral score using large-scale dental data
(A) Probability of various treatment types based on the bone level of teeth in the AL & PD dataset. The error bars in this panel illustrate the 95% confidence intervals of the computed probabilities. Advanced treatments may include procedures such as gingival flap surgery, osseous surgery, bone replacement grafts, and distal or proximal wedge procedures. (B) Relationship between treatment cost and bone level for the same set of teeth. The blue curve represents the estimated treatment costs, while the red line shows a first-degree polynomial approximation of cost. (C) Tooth Score vs. Bone Level. Tooth score decreases linearly as bone level exceeds 2.0 mm, with deductions scaled to treatment cost (B). At 6.71 mm, the average extraction threshold, 63.5 points are deducted. This scoring approach aligns with caries-based indices for consistent assessment of tooth health.
