Craniofacial Dysmorphology in Unilateral Coronal Synostosis Using Three-Dimensional Landmark-Based Analysis With Generalized Procrustes Superimposition

Abstract

BACKGROUND: Geometric morphometric analysis with Procrustes superimposition is an advanced computational tool that can be used to quantify dynamic changes in complex three-dimensional structures. The present study couples high resolution CT imaging with a Geometric Morphometric approach in order to further understand the complex dysmorphology that occurs in unilateral coronal synostosis (UCS). METHODS: Forty-one UCS patients and 41 age- and sex-matched controls received high-resolution CT imaging. Thirty-one anatomical landmarks were identified on each imaging set. A geometric morphometric workflow was used to perform a Procrustes superimposition to register landmarks into a common space. Procrustes-aligned landmarks were used to derive angle calculations, lengths, and other anatomical measurements. Three-dimensional coordinates were also used to perform a principal components analysis (PCA). RESULTS: Unilateral coronal synostosis patients exhibited significant angular deviation at the levels of the inferior skull base, mid-posterior fossa, and vertex. Both left- and right-sided UCS patients showed increased lengthening in the transverse (left-right) dimension, exhibiting increased length between the left and right EAC (P = 0.047). Conversely, UCS patients revealed shortening in the midline AP dimension as evidenced by the decreased Nasal root-Lambda (P < 0.0001) and Nasal root-superior dorsum sellae (P = 0.01) distances compared with controls. PCA revealed that 25.26% of variation in shape among the patients sampled to be driven by flattening of the skull and that18.93% of variation was driven by right-sided deformity and mediolateral expansion. CONCLUSIONS: Significant dysmorphology between UCS patients and controls was quantified using Geometric Morphometric approach, which may be useful in further characterizing the dynamic craniofacial changes in UCS.