Prediction and Classification of Single Tooth Recession Using deep neural networks

Abstract

Background: Single-tooth recession is a common dental condition that can lead to aesthetic concerns and sensitivity. The early prediction of single tooth recession can be crucial for implementing preventive measures and conservative treatments.
Materials And Methods: The study used an Orange machine learning tool to analyze intraoral frontal images to predict and detect lower anterior recession. Three algorithms were used: Logistic Regression, Neural Network, and Naïve Bayes. The models were trained on 70 images and evaluated using a confusion matrix.
Results: The machine learning algorithms with CNN embedding successfully detected and predicted a recession. The model showed high accuracy in predicting recession in anterior teeth. For multiple lower anterior recessions, it showed an AUC of 0.964. With the advancements in artificial intelligence (AI) and machine learning, there is growing interest in utilizing these technologies to predict a single-tooth recession.
Conclusion: This article provides an overview of the current state of research in this area, discusses the challenges and opportunities, and suggests future directions for developing AI-based predictive models for single tooth recession.