Presentation Blocks: 03-23-2018 - Friday - 03:45 PM - 05:00 PM

Title: Predicting Facial Type From Mandibular Landmark Data at Young Ages


HeeSoo Oh (Presenter)
University of the PacificUniversity of the Pacific Arthur A. Dugoni School of Dentistry

Richard Sherwood, University of Missouri
Dana Duren, University of Missouri
Manish Valiathan, Case Western Reserve University,School of Dental Medicine
Emily Leary, University of Missouri
Kieran McNulty, University of Minnesota


Objectives: To assess the potential of predicting adult facial type at different stages of mandibular ontogeny.

Methods: 956 participants of the Denver, Iowa, Michigan, and Oregon Growth Studies with longitudinal lateral cephalograms between ages 3-18 years were placed into one of three facial types based on mandibular plane angle (MPA) at 15 years of age or older: normodivergent (MPA 28-39°), hypodivergent (MPA < 28°), and hyperdivergent (MPA > 39°). 31 2D anatomical landmarks, representing soft and hard tissues of the mandible, were digitized and superimposed using generalized Procrustes analysis placing them within a common shape space. Data were analyzed within three-year age categories using stepwise discriminant function analysis with jackknife crossvalidation to test how well young individuals can be re-classified to their adult facial types.

Results: Individuals were correctly classified significantly better than chance even among even the youngest age category (3-6 yrs.). Crossvalidation rates increased with age, and only in the 12-15 year category were rates similar (71% correctly reclassified) to results achieved among the oldest category (> 15 years). Although each category has a few unique landmarks best discriminating among types, several landmarks were common across all age categories: menton, gonion, articulare, and intersection of the posterior extension of the palatal plane with the posterior ramus. One additional point (intersection of basisphenoid synchondrosis and the neck of the condyle) becomes important for recognizing facial type in late adolescence and adults. Finally, when mandibular landmark data are first superimposed among a broader suite of facial landmarks, thereby preserving jaw position relative to the face, identification of facial type from only the mandible improves markedly.

Conclusions: The discovery of important indicators of facial type in adolescent mandibles helps improve our capacity to identify different facial types at a younger age and, potentially, options for timing of treatment.