Classification of Maxillofacial Morphology by Artificial Intelligence Using Cephalometric Analysis Measurements

Abstract

The characteristics of maxillofacial morphology play a major role in orthodontic diagnosis and treatment planning. While Sassouni's classification scheme outlines different categories of maxillofacial morphology, there is no standardized approach to assigning these classifications to patients. This study aimed to create an artificial intelligence (AI) model that uses cephalometric analysis measurements to accurately classify maxillofacial morphology, allowing for the standardization of maxillofacial morphology classification. This study used the initial cephalograms of 220 patients aged 18 years or older. Three orthodontists classified the maxillofacial morphologies of 220 patients using eight measurements as the accurate classification. Using these eight cephalometric measurement points and the subject's gender as input features, a random forest classifier from the Python sci-kit learning package was trained and tested with a k-fold split of five to determine orthodontic classification; distinct models were created for horizontal-only, vertical-only, and combined maxillofacial morphology classification. The accuracy of the combined facial classification was 0.823 ± 0.060; for anteroposterior-only classification, the accuracy was 0.986 ± 0.011; and for the vertical-only classification, the accuracy was 0.850 ± 0.037. ANB angle had the greatest feature importance at 0.3519. The AI model created in this study accurately classified maxillofacial morphology, but it can be further improved with more learning data input.

Keywords: artificial intelligence (AI); cephalograms; k-fold; machine learning; orthodontics; random forest classifier (RF).

Figure 1

Illustration of cephalometric measurements considered for classification. Each cephalometric measurement is shown. The blue line (1) is an illustration of the measurement used in anteroposterior analysis, and the red lines (2–8) are illustrations of the measurements used in vertical analysis.

Figure 2

Cephalometric inputs for the determination of anteroposterior, vertical, and combined facial classifications. The input data and output classifications are listed in each category.

Figure 3

Confusion matrix for 5 runs of the combined facial classification RF. The horizontal axis represents the AI classification result (Predicted label), and the vertical axis represents the accurate classification (True label). The numbers I, II, and III represent Class I, Class II, and Class III, respectively, and S, M, and L represent Short, Medium, and Long, respectively. Blue represents patients who were classified correctly, and gray represents the misclassified patients.

Figure 4

RF confusion matrix anteroposterior classifications using k-fold (n = 5). The horizontal axis represents the AI classification results (predicted label), and the vertical axis represents the accurate classifications (true label). Blue represents patients who were classified correctly, and gray represents the misclassified patients.

Figure 5

RF confusion matrix vertical classifications using k-fold (n = 5). The horizontal axis represents the AI classification results (Predicted label), and the vertical axis represents the accurate classification (True label). Blue represents patients who were correctly classified, and gray represents the misclassified patients.

Figure 6

Feature importance for the combined facial classification RF. A total of 9 features were shown for gender, and 8 cephalometric items were used for classification.