How accurate is the soft tissue prediction of Dolphin Imaging for orthognathic surgery?

The purpose of this retrospective study is to evaluate the accuracy of soft tissue prediction models with Dolphin Imaging.

Preoperative and 1-year postoperative Lateral cephalograms (LCG) of patients undergoing Le Fort I, bilateral sagittal split osteotomy (BSSO) or bimaxillary osteotomy (Bimax) between 2006 and 2009 were retrospectively collected. A hard tissue virtual surgery was performed on the preoperative LCG to match the hard tissue of the 1-year postoperative LCG. The soft tissue changes were then modelled by Dolphin Imaging (version 11.5b) and the differences in x- and y-coordinates of 11 hard tissue and 14 soft tissue landmarks between the predicted model and 1-year postoperative LCG were determined. Statistical analyses were performed using a one-tailed, one-sample t-test for both soft- and hard tissue differences, and a Bland-Altman plot for interobserver bias.

A total of 108 patients were included, the mean age was 30 years and 23% were male. The virtual treatment was considered accurate; all hard tissue landmark differences were less than 1mm. The soft tissue model by Dolphin Imaging showed significant differences of more than 2mm for several landmarks, including the vertical positioning of Stomion Inferius (P=0.007), Lower lip (P=0.005) and Pogonion (P=0.03) in the Bimax group and horizontal positioning of Stomion Inferius (P<0.001) in the BSSO group.

Dolphin Imaging gives reasonable predictions of postoperative outcome. There is, however, room for improvement, especially regarding the vertical prediction in the lower lip region.