Deep Learning based Object Detection Algorithm for the Detection of Dental Diseases and Differential Treatments
Athira Thulaseedharan, Lal Priya P S
Abstract
With a wide range of applications in cosmetic dentistry treatments, and treatment planning, the use of Artificial Intelligence (AI) and Machine Learning is having a growing influence in the field of dentistry and is enhancing the growth of digital tools and technology. The surgical sector, automatic disease diagnosis, and recently established customized medicine are just a few of the disciplines that might benefit from AI support. Personalized medicine provides the inclination to diseases, diagnosis, and selection of the best treatment for a specific individual. In fields like image processing, deep learning has been able to match and improve human performance for exceedingly complicated tasks such as object detection models with classification, and prediction capabilities and for the identification of numerous dental diseases and their differential treatments. Many of these neural network applications face significant challenges due to the lack of labeled medical datasets, and oral disease datasets are no exception. This pilot study makes it possible to identify various dental diseases and the treatments for them by utilizing a dataset of 664 dental panoramic X-ray images which is considered a small number. These images have been used for training after being manually labeled for 9 separate classes. You Only Look Once (YOLO) version 5, a one-step search method primarily utilized for real-time object detection, is the deep neural network that was used in this study. Bounding boxes helped identify six differential treatments and three classes of dental diseases. Results show that this strategy was able to attain a Precision of 1.00, F1 score of 0.74, Recall of 0.83, and mAP of 72.4 percent by training with this smaller number of images, which shows tremendous potential for dentists and radiologists.