Automatic Detection of Dental Cysts in Panoramic Radiography Images using Preprocessing Techniques and Convolutional Neural Networks

Mouth-related pathologies represent an important challenge for public authorities. To develop a methodology, through studies on Computer Vision techniques, for the automatic identification of dental cysts in panoramic radiography images, providing Dental professionals with an alternative to aid in the interpretation of these images. For this purpose, two CNN architectures were analyzed for classification and experimentation using image pre-processing techniques. One such proposal, using morphological contrast, had a better performance, with a precision of 0.937 and an F1 score of 0.847.