|| Checking for direct PDF access through Ovid
The purpose of this study was to quantify temporomandibular joint disk-slice information produced by magnetic resonance imaging by means of a stepwise discriminant analysis. One hundred ninety-four adolescents consented to magnetic resonance imaging evaluation of their temporomandibular joints. Sagittal magnetic resonance imaging slices of each joint were assigned to one of six subjective categories of disk position by an experienced maxillofacial radiologist. Standardized reference planes transferred to each magnetic resonance image from corresponding lateral cephalometric radiographics facilitated the measurement of disk length and disk displacement and the computation of ratio values of these measurements. Discriminant analysis revealed that all three quantitative variables were descriptive and discriminant for grouping slice data into pre-established subjective categories. Cross-validation and misclassification error calculations showed a 69.3% agreement between subjective and discriminant classification. Therefore quantification of disk displacement can be used in place of subjective evaluation. In addition, discriminant analysis disclosed a reduction in disk length associated with increased severity of disk displacement.