Cephalometric regional superimpositions -- digital vs. analog accuracy and precision: 2. the mandible

Introduction: Lateral cephalometric superimpositions (LCS) are used to measure dental and skeletal changes that occur in the craniofacial complex over time. Orthodontists use LCSs to assess treatment outcomes. The purpose of this study was to conduct an assessment of the measured displacement of defined dental landmarks across digital and analog methods of mandibular regional serial superimposition as compared to an implant-registered superimposition reference. The data used in this study was derived from the Mathew\u27s Acquisition Group implant sample; the first United States longitudinal study of growing children with maxillary and mandibular Björk type metallic implants. Methods: Sixty-six lateral cephalometric radiographs were selected from twenty-two children. Three cephalometric tracings were completed for each subject that were then superimposed pairwise (T1 vs. T2, T2 vs. T3) across four separate methods of superimposition, two analog: Implant, Structural; and two digital: Dolphin, Quick Ceph. Each superimposition was then imported into Adobe Photoshop where the images were scaled and the displacement of defined dental structures was measured. Defined dental structures included: (1) first molar mesial contact point, (2) first molar apical root bisection, (3) central incisor root apex, and (4) central incisor crown incisal edge. A random-effects, generalized linear model was used to contrast dental landmark displacement measurements. Results: There was no difference between the mean displacement of defined dental structures between different methods (p=0.145). There was no difference between the different methods by defined dental structure (p=0.150). Conclusions: Our study demonstrated that there are no statistically significant differences among three methods of mandibular regional superimposition in comparison to an implant-registered (reference) method (analog: Structural, Implant; digital: Dolphin, and Quick Ceph). The historical data set utilized in our study, limited by the small sample size, resulted in a relatively low power (0.15). A low power increases the likelihood of incorrectly failing to reject a null hypothesis that is actually false. which must be considered in our study