IADR Abstract Archives
Automated Characterization of Maxillary Asymmetry in Unilateral Impacted Canine Patients
Objectives: Maxillary canines are the second most frequently impacted teeth. Palatal expansion, surgical exposure, and space opening have been used clinically to remove anatomic obstacles to allow canine eruption. However, less is known about the skeletal variation in patients with impacted maxillary canines, especially 3D volumetric characterization. We hypothesize that alveolar structure of maxillae in patients may be genetically constricted, causing the impaction.
Methods: To test the hypothesis, we introduced a new algorithm of CBCT images to quantify the volumetric discrepancy of the skeletal maxilla, defined as the maxillary asymmetry differential (MAD), via 30 Chinese individuals with unilaterally impacted maxillary canines (UIMC). Another 30 healthy age- and gender- matched subjects served as the control. The mean ages were 16.5 and 18.8 years at the time of the CBCT scans for the study and the control group, respectively. A fully automatic segmentation method based on advanced machine learning technologies was utilized to isolate the maxilla. To assess the asymmetry of the maxilla, a midsagittal plane was constructed using an automatic landmark finder. ANOVA was used for statistical analysis.
Results: The auto-segmentation successfully identified the maxilla structure and split the midsagittal plane for asymmetric measures; the computational time for completion of each CBCT analysis took an hour. The reliability of the new algorism was validated with a minimum mean difference of 1 voxel (=0.3mm) between the manually identified landmarks and the automatic digitized landmarks. No significant difference in MAD bone volume between the impacted side (2.10±0.33) x104 mm3 and non-impacted side (2.11±0.31) x104 mm3 in the patient with UIMC was found when compared to control patients. The maxilla of the study group had a significantly smaller volume than that of the control (p<0.05).
Conclusions: Our data provides evidence of the maxillary constriction and support arch expansion for the patients having UIMC.
Methods: To test the hypothesis, we introduced a new algorithm of CBCT images to quantify the volumetric discrepancy of the skeletal maxilla, defined as the maxillary asymmetry differential (MAD), via 30 Chinese individuals with unilaterally impacted maxillary canines (UIMC). Another 30 healthy age- and gender- matched subjects served as the control. The mean ages were 16.5 and 18.8 years at the time of the CBCT scans for the study and the control group, respectively. A fully automatic segmentation method based on advanced machine learning technologies was utilized to isolate the maxilla. To assess the asymmetry of the maxilla, a midsagittal plane was constructed using an automatic landmark finder. ANOVA was used for statistical analysis.
Results: The auto-segmentation successfully identified the maxilla structure and split the midsagittal plane for asymmetric measures; the computational time for completion of each CBCT analysis took an hour. The reliability of the new algorism was validated with a minimum mean difference of 1 voxel (=0.3mm) between the manually identified landmarks and the automatic digitized landmarks. No significant difference in MAD bone volume between the impacted side (2.10±0.33) x104 mm3 and non-impacted side (2.11±0.31) x104 mm3 in the patient with UIMC was found when compared to control patients. The maxilla of the study group had a significantly smaller volume than that of the control (p<0.05).
Conclusions: Our data provides evidence of the maxillary constriction and support arch expansion for the patients having UIMC.