TOWARD AUTOMATIC IDENTIFICATION OF ANATOMIC ANOMALIES: CHARACTERIZING THE HYOID REGION OF CLEFT PALATE PATIENTS USING SHAPE STATE DESCRIPTORS

Abstract

BACKGROUNDThe identification of anatomic variants and anomalies is achieved objectively through genetic and/or morphometric approaches.OBJECTIVE(S)The purpose of this study is to determine whether the hyoid region quantitatively differs between subadult and adult patients with or without cleft lip and palate.STUDY DESIGNSixty scans derived from cone beam computed tomography of 3 groups were analyzed. The sample included control patients, patients with nongrowing cleft lip/palate, and patients with growing cleft lip/palate, with each group divided into 2 subgroups of patients with either uni- or bilateral cleft lip/palate. A quantitative analysis of the hyoid bone with contiguous vertebra was performed using least squares quantification and thin plate spline descriptors.RESULTSHyoidal regions of control patients were generally higher and wider compared with the regions of patients with cleft lip and palate, which were shorter and narrower based on the Procrustes and thin plate spline analyses. Size dominated the system, with patients with cleft lip and palate having quantitatively smaller regions in comparison to controls. Differences were related primarily to position of the hyoidal apparatus and secondarily to shape changes. Alterations in shape contributed minimally to the system in differentiating the sample with respect to height and width.DISCUSSION/CONCLUSIONSCleft lip and palate affected the size of the hyoidal apparatus but shape was minimally altered. The Procrustes superimposition analysis provides a potential approach for comparing individual patients to large samples of normative information. The identification of anatomic variants and anomalies is achieved objectively through genetic and/or morphometric approaches. The purpose of this study is to determine whether the hyoid region quantitatively differs between subadult and adult patients with or without cleft lip and palate. Sixty scans derived from cone beam computed tomography of 3 groups were analyzed. The sample included control patients, patients with nongrowing cleft lip/palate, and patients with growing cleft lip/palate, with each group divided into 2 subgroups of patients with either uni- or bilateral cleft lip/palate. A quantitative analysis of the hyoid bone with contiguous vertebra was performed using least squares quantification and thin plate spline descriptors. Hyoidal regions of control patients were generally higher and wider compared with the regions of patients with cleft lip and palate, which were shorter and narrower based on the Procrustes and thin plate spline analyses. Size dominated the system, with patients with cleft lip and palate having quantitatively smaller regions in comparison to controls. Differences were related primarily to position of the hyoidal apparatus and secondarily to shape changes. Alterations in shape contributed minimally to the system in differentiating the sample with respect to height and width. Cleft lip and palate affected the size of the hyoidal apparatus but shape was minimally altered. The Procrustes superimposition analysis provides a potential approach for comparing individual patients to large samples of normative information.