Accuracy of auto-identification of the posteroanterior cephalometric landmarks using cascade convolution neural network algorithm and cephalometric images of different quality from nationwide multiple centers.

INTRODUCTION: The purpose of this study was to evaluate the accuracy of auto-identification of the posteroanterior (PA) cephalometric landmarks using the cascade convolution neural network (CNN) algorithm and PA cephalogram images of a different quality from nationwide multiple centers nationwide. METHODS: Of the 2798 PA cephalograms from 9 university hospitals, 2418 images (2075 training set and 343 validation set) were used to train the CNN algorithm for auto-identification of 16 PA cephalometric landmarks. Subsequently, 99 pretreatment images from the remaining 380 test set images were used to evaluate the accuracy of auto-identification of the CNN algorithm by comparing with the identification by a human examiner (gold standard) using V-Ceph 8.0 (Ostem, Seoul, South Korea). Pretreatment images were used to eliminate the effects of orthodontic bracket, tube and wire, surgical plate, and surgical screws. Paired t test was performed to compare the x- and y-coordinates of each landmark. The point-to-point error and the successful detection rate (range, within 2.0 mm) were calculated. RESULTS: The number of landmarks without a significant difference between the location identified by the human examiner and by auto-identification by the CNN algorithm were 8 on the x-coordinate and 5 on the y-coordinate, respectively. The mean point-to-point error was 1.52 mm. The low point-to-point error (<1.0 mm) was observed at the left and right antegonion (0.96 mm and 0.99 mm, respectively) and the high point-to-point error (>2.0 mm) was observed at the maxillary right first molar root apex (2.18 mm). The mean successful detection rate of auto-identification was 83.3%. CONCLUSIONS: Cascade CNN algorithm for auto-identification of PA cephalometric landmarks showed a possibility of an effective alternative to manual identification.

Comparison of slot sizes and parallelism of metal brackets manufactured through metal injection molding and computerized numerical control.

OBJECTIVE: To investigate and compare the slot sizes and parallelism of metal injection molding (MIM) and computerized numerical control (CNC) brackets. METHODS: The following four MIM bracket series with 0.022-inch (in) slots were selected for investigation: Di MIM mini Twin (Ortho Organizers), Mini Diamond Roth (Ormco), Gemini MBT (3M Unitek), and Formula R Roth (Tomy). The following four CNC bracket series with 0.022-in slots were selected for investigation: Econoline MBT (Adenta), Legend mini MBT (GC Orthodontics), Crown mini MBT (Adenta), and Evolve MBT (DB Orthodontics). The slot dimensions were measured using an optical microscope (XTCam-D310M; Mitutoyo) with a resolution of 1 µm. The results were statistically analyzed using one-way analysis of variance and the Tukey post-hoc test with a significance level of 0.05. RESULTS: The results indicated that all the investigated slot sizes were oversized with respect to the manufacturers' specifications (0.022 in). Among the eight bracket series, the Di MIM bracket (MIM) was the most oversized by 10.4%, whereas the Evolve bracket (CNC) was the least oversized by 2.6%. The slots in seven of the bracket series had divergent walls instead of parallel ones. The Evolve bracket alone had parallel slot walls. CONCLUSIONS: Regardless of the manufacturing method, all the slot sizes of the brackets investigated in this study were significantly oversized; most of the slot walls were nonparallel, except for those of the Evolve bracket. This study could not establish that the CNC method was more accurate than the MIM method in manufacturing bracket slots.