Evaluation of the factors effective in the decision making for mandibular single incisior extraction using artificial intelligence method

Aim: The aim of this study was to evaluate the factors that were effective in deciding on mandibular single incisor extraction by using artificial intelligence. Materials and Methods: In our study, the plaster models taken before treatment of 105 patients with lower single incisor extraction with an average age of 15,82, 100 patients with premolar extraction with an average age of 15,21, and 100 patients with fixed treatment without extraction with an average age of 15,10 were scanned using a desktop model scanning device (3Shape R700TM) and digital models were obtained. Space analysis, Bolton analysis, irregularity index, ICON index, intercanine and intermolar distance, overjet and overbite, molar relationship, arch form, IMPA°, U1- SPP°, SN-GoGn°, ANB° parameters were calculated. Measurements and analyzes on digital models were applied using OrthoAnalyzer software. The relationship between the measurements made in the study groups and the results of the cephalomeric analysis was analyzed using ANOVA analysis, Tukey-Kramer multiple comparison test, Kruskal-Wallis and Chi-Square tests. Results: Irregularity index scores were found to be significantly higher in the lower single incisor extraction group. In the lower single extraction group, the SN-GoGn° values were significantly lower than the premolar extraction group. Triangular arch form was found to be significantly higher in the lower single incisor extraction group. No statistically significant difference was found in lower anterior Bolton discrepancy between the premolar extraction and lower single incisor extraction groups. The decision nodes were determined by the decision tree algorithm as ICON scores of 36 and 58, irregularity index scores of 5,5 and 9,3, lower crowding values of 4,2 and 6,2 mm, upper crowding value of 5,7 mm, anterior Bolton discrepancy of 0,095, and SNGoGn° value of 30°. Conclusion: Fixed treatment with lower single incisor extraction was applied to the patients with lower anterior Bolton discrepancy (mean=0,85), mandibular crowding greater than 6,2 mm, and irregularity index greater than 5,5. On the other hand, premolar extraction was decided for the patients with ICON value greater than 58, upper crowding value greater than 5.7 mm, and SN-GoGN° value greater than 30.