Deep Learning Approach for the Classification and Detection of Dental and Craniofacial Conditions

Artificial Intelligence (AI) provides everyday functions along with higher applications in the domain of medicine, particularly medical imaging. Due to the advancement of technologies and imaging tools, AI-powered machine learning models will soon be used on a routine basis in medical diagnostics and treatment. The development involves Deep Learning (DL) algorithms and Convolutional Neural Networks (CNN) which will be trained using a dataset of disease-related images. AI gains increasing significance in our lives and medical research. In craniofacial imaging, CNNs have gained popularity and are employed in numerous scientific studies. This research introduces a DL model that utilizes a dataset containing five distinct categories: cavities, crown fractures, gum diseases, malalignment, and receding gums. These conditions have been classified and detected using the pre-trained Mobile-Net model. Notably, this model demonstrates a high training and validation accuracy reaching 99,9% and an incredibly low error rate of 0,001%.