A device to predict lower third molar eruption.

This study was designed to create a transparent device to be superimposed on a panoramic tomogram taken of a patient at age 20 years to make predictions of future eruption or impaction of lower third molars. The device was developed from data on 40 lower third molars initially retained at age 20 years; one half of these remained impacted, and the other half of them erupted by age 26 years. Tracings were made from panoramic tomograms taken at age 20 years. The critical point for prediction in the overlay was the intersection of a horizontal reference line and the anterior border of the ascending ramus. To estimate this critical separation line, Bayes' Decision Theory was used. The sum of false negatives and false positives was least at a distance of 14.5 mm from the distal surface of the second molar. Mesial from this point, the probability of impaction was 76%; distal from this point, the probability of eruption was 72%. Tested against 35 initially unerupted lower third molars, the predictions made by the device and the actual clinical findings were in conformity in relation to 97% of the test teeth. It could thus be concluded that the method was simple to use and may prove a good addition for predicting lower third molar development.