Novel method to diagnose extraction patterns with the artificial intelligence decision-making model using neural network

Abstract

Introduction: The diagnosis of extractions in the orthodontic treatment is important and difficult, because that decision has tendency to be based on the practitioners experiences. The purpose of this study was to construct an artificial intelligent expert system for the diagnosis of extraction using neural network machine learning (NNML) and to evaluate performance of this model. Methods: The subjects consisted of 156 patients in total. Input data consisted of 12 cephalometric variables and additional six indices. Output data consisted of three bits to divide extraction patterns. Four NNML models for the diagnosis of extractions were constructed using backpropagation algorithm, and were evaluated. Results: The success rates of the models showed 93% for the diagnosis of extraction versus non-extraction, and showed 84% for the detailed diagnosis of the extraction patterns. Conclusions: This study suggests that artificial intelligent expert systems using neural network machine learning could be useful in orthodontics. Improving performance was achieved by the components such as proper selection of the input data, appropriate organization of the modeling, and preferable generalization.