The present study intends to explore different Deep Learning algorithms to identify dental implant brands, to assist in the rehabilitation process. Data Augmentation Techniques were implemented and explored, along with a VGG (Visual Geometry Group) style model, models implemented using Transfer Learning, and an approach from another academic study. A dataset was developed through contact with several dental clinics. Given the nature of the data that makes up the dataset, it was not possible to obtain a large amount of data, as well as an egalitarian distribution between classes. In view of the above, the Stratified k-fold cross-validation method was used during the training process to minimize potential problems. The model’s performances were evaluated according to various statistical measures, during the training process and during the testing process. During the training process, the metrics used were entropy loss and accuracy. In the testing process, the models were evaluated for accuracy, recall, specificity and F1 Score. It is concluded that the best models are the model belonging to the academic study and the model in the VGG style, since both perform well at the level of the training process and at the level of the testing process.