Abstract
The objective of this study was to visually classify soft tissue profile patterns of patients with skeletal mandibular prognathism using self-organizing maps (SOM).
The subjects of this retrospective study were 90 Japanese adult females (mean age: 22.6±4.2 years) who underwent surgical orthodontic treatment for skeletal mandibular prognathism. The control subjects were 27 Japanese adult females (mean age: 23.8±4.4 years) who underwent non-surgical orthodontic treatment. Pre-treatment cephalograms for the 90 mandibular prognathism and post-treatment cephalograms for the control subjects were traced, and 48 points of measurement were digitalized. Among the 48 points, 27 soft tissue points were used as input data for the SOM with multiple calculation units. The unit groups performed automatic learning of the input data by means of a self-organizing algorithm, and virtual profiles were obtained for each unit. As a result, four soft tissue profile patterns were visualized, and close relationships to the skeletal and dental characteristics were found in each pattern. The most typical soft tissue pattern in Japanese mandibular prognathism seems to be a pattern showing a moderately retrusive maxilla and a moderately protrusive mandible with the lower lip protrusion and the small nasolabial angle.
