Machine learning in orthodontics: Introducing a 3D auto-segmentation and auto-landmark finder of CBCT images to assess maxillary constriction in unilateral impacted canine patients

Abstract

Objectives: To (1) introduce a novel machine learning method and (2) assess maxillary structure variation in unilateral canine impaction for advancing clinically viable information.

Materials and methods: A machine learning algorithm utilizing Learning-based multi-source IntegratioN frameworK for Segmentation (LINKS) was used with cone-beam computed tomography (CBCT) images to quantify volumetric skeletal maxilla discrepancies of 30 study group (SG) patients with unilaterally impacted maxillary canines and 30 healthy control group (CG) subjects. Fully automatic segmentation was implemented for maxilla isolation, and maxillary volumetric and linear measurements were performed. Analysis of variance was used for statistical evaluation.

Results: Maxillary structure was successfully auto-segmented, with an average dice ratio of 0.80 for three-dimensional image segmentations and a minimal mean difference of two voxels on the midsagittal plane for digitized landmarks between the manually identified and the machine learning-based (LINKS) methods. No significant difference in bone volume was found between impaction ([2.37 ± 0.34] [Formula: see text] 104 mm3) and nonimpaction ([2.36 ± 0.35] [Formula: see text] 104 mm3) sides of SG. The SG maxillae had significantly smaller volumes, widths, heights, and depths (P < .05) than CG.

Conclusions: The data suggest that palatal expansion could be beneficial for those with unilateral canine impaction, as underdevelopment of the maxilla often accompanies that condition in the early teen years. Fast and efficient CBCT image segmentation will allow large clinical data sets to be analyzed effectively.

Keywords: CBCT; Canine impaction; Image segmentation; Machine learning; Orthodontics.

Figures

Figure 1.

Flowchart of the automatic segmentation method. In the training stage (left), a series of sequential random forest classifiers were obtained through iterative training using the appearance features from the original CBCT, the context features from the updated segmentation probability maps, and the training labels. In the application stage (right), these classifiers were sequentially applied to the new target CBCT to iteratively generate the final segmentation.

Figure 2.

The proposed location of the three landmarks (Ba, Na, and ANS) used to define the midsagittal plane are shown by the cross-points of two black lines.

Figure 3.

An example of automatic segmentation results. Per the study design, the craniofacial area was segmented into three regions of interest (ROIs): maxilla (yellow), mandible (red), and the rest of the craniofacial skeleton (blue).

Figure 4.

(A) Segmentation results for the maxilla. (B) The superimposition results of three different types of impaction (buccal, mid-alveolus, and palatal) are shown, allowing for geometric difference determination.