Digital setup accuracy for moderate crowding correction with fixed orthodontic appliances: a prospective study.

ABSTRACT

OBJECTIVES: To evaluate the accuracy of a semi-automatic 3D digital setup process in predicting the orthodontic treatment outcome achieved by labial fixed appliances. SUBJECTS AND METHODS: Twenty-five adult patients (18 to 24 years old) with class I malocclusion and moderate crowding were prospectively enrolled and received treatment on both jaws through the straight-wire technique. Prior to treatment commencement, a semi-automatic digital setup simulating the predicted treatment outcome was performed for each patient through Orthoanalyzer software (3Shape®, Copenhagen, Denmark) to obtain the prediction model. This was compared to the final outcome model through 3D superimposition methods. Metric variables and inspection of color-coded distance maps were used to detect how accurately the digital setup predicts the actual treatment outcome. RESULTS: The mean absolute distances (MAD) between the superimposed dental arches of the predicted and the final models were 0.77 ± 0.13 mm following superimposition on the palate, 0.52 ± 0.06 mm following superimposition on the maxillary dental arch, and 0.55 ± 0.15 mm following superimposition on the mandibular dental arch. The MAD at the palatal reference area was 0.09 ± 0.04 mm. Visualization of color-coded distance maps indicated that the digital setup accurately predicted the final teeth position in a few cases. Almost half of the cases had posteriorly wider upper and lower dental arches and palatally/lingually positioned or inclined anterior teeth, whereas the rest still showed errors within 2-3 mm, distributed over the entire dental arches with no distinct pattern. CONCLUSIONS: The accuracy of semi-automatic prediction of the labial fixed appliance treatment outcome in Class I cases with moderate crowding is not yet sufficient. While average measures showed deviations less than 1 mm, examination of individual color-coded distance maps revealed significant disparities between the simulated and the actual results.