الفهرس | Only 14 pages are availabe for public view |
Abstract This study aimed to assess accuracy and compare four different third party software programs regarding 3D superimposition of CBCT models in follow up of orthodontic cases. It was based on pre-treatment (T) and post-treatment (T1) CBCT DICOM images of 21 patients treated with RME. The DICOM files were transferred to the data base of the following third party software programs; Dolphin 3D (Dolphin Imaging and Management, Chatsworth, California), InVivo (Anatomage, San Jose, California, USA), OnDemand 3D (Ondemand 3D; Cybermed Co., Seoul, Korea) and Mimics (Materialise, Louvain, Belgium) programs. Six superimposed models were obtained from the programs: Dolphin PBR and SBR, InVivo PBR and VBR, OnDemand VBR and Mimics PBR superimposed models. Accuracy of these software programs was tested by comparing results of the superimposed models with the gold standard. The gold standard was calculated from 3D tracing of each CBCT image (T and T1) using InVivoDental (Anatomage, San Jose, CA, USA) after setting the coordinate system for analysis formed by; Frankfort horizontal plane and mid sagittal plane perpendicular to it. The points of evaluation were the mid buccal points of maxillary; canine, second premolar and first molar bilaterally. The distance from each evaluation point to the mid sagittal plane was measured automatically by the software at T CBCT image then the same procedure was repeated at T1 CBCT image. The gold standard was the mathematical difference between T and T1. Inter-observer reliability was investigated as well by comparing the results between the three observers. We also provided a qualitative analysisof software options provided by each program regarding the 3D superimposition techniques by 15-point evaluation list. After all, the study provided quantitative and qualitative ranking of the software programs as well as methods of the 3D superimposition based on the results. The current results revealed poor accuracy of all methods under investigation detected from the high error percent compared to the gold standard. Inter-observer reliability was also poor between all observers, however range of variability was less than 0.5 mm. We referred this result to multidirectional change caused by RME thus 2D assessment on the axial cut didn’t help enough in determination of the direction of the change. Also, the image quality of the axial cut affected the accuracy of the results especially in OnDemand. The qualitative ranking suggested that InVivo VBR method was the best. SBR method of Dolphin overlay method showed an intermediate rank. PBR method of InVivo point, Dolphin side by side and Mimics came at the lower scale. While OnDemand software was the worse. Regarding the qualitative assessment, the VBR method showed the most superior results followed by the SBR method. The PBR method was the most inferior one. In general, InVivo Dental software was the most preferable program for the raters while, Mimics software was the least accepted one regarding 3D superimposition. Total qualitative and quantitative ranking grant the first rank to InVivo VBR method. In conclusion: 1. 2D assessment using the axial cuts of the 3D superimposed models was not accurate in follow up of RME cases. 2. from the quantitative and the qualitative ranking of 3D superimposition methods, voxel based superimposition was superior to the other two methods. 3. Regarding software programs, InVivo software showed the best performance regarding both the quantitative and the qualitative assessments. |