Abstract
No study has investigated the effect of learning curves on the accuracy of dental implant navigation systems. This study evaluated the accuracy of the dental implant navigation system and established the learning curve according to operation site and operating time. Each dental model was used for drilling 3 missing tooth positions, and a patient tracking module was created. The same dentist performed the drilling test for 5 sets of dental models. CT back scanning was performed on the dental models. Customized implants based on the drilled holes were inserted. The relative error between the preoperative planning and actual implant was calculated. Using the dental navigation system could help dentists position implants more accurately. Increasing the frequency with which a dentist used the navigation system resulted in shorter operations. Longitudinal and angular deviation were significantly (P < 0.0001 and P = 0.0164). We found that the same level of accuracy could be obtained for the maxilla and mandible implants. The Student's t test demonstrated that the longitudinal error, but not the total or angular error, differed significantly (P = 0.0012). The learning curve for the dental implant navigation system exhibited a learning plateau after 5 tests. The current system exhibited similar accuracy for both maxillary and mandibular dental implants in different dental locations. The one-way ANOVA revealed that the total, longitudinal, and angular errors differed significantly (P < 0.0001, P < 0.0001 and P = 0.0153). In addition, it possesses high potential for future use in dental implant surgery and its learning curve can serve as a reference for dentists.