| dc.description.abstract | Introduction: Determining the right time for orthodontic treatment is one of the most important factors affecting the treatment plan and its outcome. The aim of this study was to estimate the mandibular growth stage based on cervical vertebral maturation (CVM) in lateral cephalometric radiographs using artificial intelligence. An attempt was made by not using the usual naming of CVM stages ,but instead, the cervical vertebrae was directly related to the growth slope of the mandible.Therefore more agreement between CVM and the growth slope of the mandible was achieved.
Methods and materials: To conduct this study, first, information of people that was archieved in American Association of Orthodontics Foundation (AAOF) growth centers was assesed and after considering the entry and exit criteria, a total of 200 people, 108 women and 92 men, were included in the study. Then, the length of the mandible in the lateral cephalometric radiographs that were taken serially from the patients was calculated, and the corresponding graphs were labeled based on the growth rate of the mandible in 3 stages; before the growth peak of puberty (pre-pubertal), during the growth peak of puberty (pubertal) and after the growth peak of puberty (post-pubertal). A total of 720 images were selected for evaluation with an artificial intelligence model. These images were evaluated with different artificial intelligence models and using the diagnostic parameters of sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV).
Results: In the diagnosis of pre-pubertal and pubertal stages, the neural network designed for this study has the highest sensitivity (0.76, 0.7 respectively) and NPV (0.87, 0.84 respectively) and ResNet-50 neural network has the highest specificity (0.85, 0.95 respectively), PPV (0.77, 0.84 respectively) and accuracy (0.81, 0.79 respectively). To detect the post-pubertal stage, the highest sensitivity (0.66) with the ResNet-50 neural network and the highest specificity (0.64). NPV (0.65), PPV (0.57), and accuracy (0.76) were obtained with the network designed for this study. The highest overall diagnostic accuracy was also obtained using the network designed for this study and equal to 67%. Also, by assigning more data to the training phase, this accuracy is up to 72% increased.
Conclusion: The artificial intelligence model trained in this study can receive images of cervical vertebrae and predict mandibular growth status by classifying it into one of three groups; before the growth spurt (pre-pubertal), during the growth spurt (pubertal), and after the growth spurt (post-pubertal). This model has more diagnostic accuracy to determine the pre-pubertal and pubertal stages. | en_US |
