# Comparative Evaluation of Deep Learning Models for the Classification of Impacted Maxillary Canines on Panoramic Radiographs

## TL;DR

This study compares deep learning models for identifying impacted maxillary canines in dental X-rays, finding VGG16 to be the most accurate.

## Contribution

The novel contribution is the comparative evaluation of four CNN models for impacted canine classification using transfer learning.

## Key findings

- VGG16 achieved the highest accuracy (99.28%) and F1-score (99.43%) among the tested models.
- A prototype diagnostic interface was developed to demonstrate potential clinical application.
- The study highlights the need for further validation on diverse, multi-center datasets.

## Abstract

Background/Objectives: The early and accurate identification of impacted teeth in the maxilla is critical for effective dental treatment planning. Traditional diagnostic methods relying on manual interpretation of radiographic images are often time-consuming and subject to variability. Methods: This study presents a deep learning-based approach for automated classification of impacted maxillary canines using panoramic radiographs. A comparative evaluation of four pre-trained convolutional neural network (CNN) architectures—ResNet50, Xception, InceptionV3, and VGG16—was conducted through transfer learning techniques. In this retrospective single-center study, the dataset comprised 694 annotated panoramic radiographs sourced from the archives of a university dental hospital, with a mildly imbalanced representation of impacted and non-impacted cases. Models were assessed using accuracy, precision, recall, specificity, and F1-score. Results: Among the tested architectures, VGG16 demonstrated superior performance, achieving an accuracy of 99.28% and an F1-score of 99.43%. Additionally, a prototype diagnostic interface was developed to demonstrate the potential for clinical application. Conclusions: The findings underscore the potential of deep learning models, particularly VGG16, in enhancing diagnostic workflows; however, further validation on diverse, multi-center datasets is required to confirm clinical generalizability.