An in-depth learning algorithm successfully detects periodontal disease from 2D bitewing x-rays, according to research presented at EuroPerio10, the world’s leading congress on periodontology and implant dentistry organized by the European Federation of Periodontology (VET).
Our study shows the potential of artificial intelligence (AI) to automatically identify periodontal pathologies that might otherwise be lost. This could reduce radiation exposure by avoiding repeated assessments, prevent the silent progression of periodontal disease and allow for earlier treatment. “
Dr. Burak Yavuz, author of the study, Eskisehir Osmangazi University, Turkey
Previous studies have examined the use of AI to detect caries, root fractures, and apical lesions, but there is limited research in the field of periodontology. This study assessed the ability of deep learning, a type of AI, to determine periodontal status on bite x-rays.
The study used 434 bite x-rays of patients with periodontitis. Image processing was performed with the u-net architecture, a convolutional neural network used to segment images quickly and accurately. An experienced specialist doctor also evaluated the images using the segmentation method. Assessments included total alveolar bone loss around the lower and upper teeth, horizontal bone loss, vertical bone loss, fork defects, and calculus around the maxillary and mandibular teeth.
The neural network identified 859 cases of alveolar bone loss, 2,215 cases of horizontal bone loss, 340 cases of vertical bone loss, 108 cases of fork defects, and 508 cases of dental calculus. The success of the algorithm for identifying defects was compared with the physician’s assessment and was reported as sensitivity, accuracy, and F1 score, which is the weighted average of sensitivity and accuracy. For sensitivity, accuracy and F1 score, 1 is the best value and 0 is the worst.
The results for sensitivity, accuracy, and F1 score for total alveolar bone loss were 1, 0.94, and 0.96, respectively. The corresponding values for horizontal bone loss were 1, 0, 92 and 0, 95, respectively, while AI was unable to identify vertical bone loss. For the dental calculation, the results of sensitivity, precision and F1 score were 1, 0, 0, 7 and 0, 82, respectively, and for the furring defects the corresponding values were 0, 62, 0, 71 and 0, 66, respectively.
Dr. Yavuz said: “Our study shows that artificial intelligence is able to detect many types of defects from 2D images that could help in the diagnosis of periodontitis. More comprehensive studies are needed on larger data sets to increase the success of models and expand their results. use for 3D x-rays “.
He concluded: “This study provides an insight into the future of dentistry, where AI automatically evaluates images and helps dental professionals diagnose and treat the disease sooner.”
Source:
European Federation of Periodontology (VET)