By Dr. Mehmood (BDS, M Phil, Ph.D.)
Besides visual examination, dentists rely heavily on radiographs to diagnose dental problems, prepare treatment plans, and perform therapeutic procedures. However, radiographic interpretation is a time-consuming process. More importantly, there is significant variation between the radiographic interpretation of dental professionals, causing treatment complications, such as delayed healing, and undesired aesthetic and therapeutic outcomes.
Nevertheless, humans have always been able to find solutions to pressing issues. As such, what if dental x-rays could be read and interpreted by computers? This would help save dentists' time and let them focus on more important things. Furthermore, computer-aided x-ray interpretation can help reduce diagnostic variability and human bias during diagnosis. (Bhimavarapu & Battineni, 2022)
It seems far-fetched, but it is possible. Dental informatics is a new field that helps improve diagnostic accuracy, save dentists' time, and improve patients' overall health and quality of life (Almalki et al., 2022). Thanks to machine learning, computers can be taught to detect patterns, interpret changes, and recommend solutions. The same principle is used to interpret dental x-rays. The technology making that possible is known as artificial intelligence (AI).
How Does it Work?
The first step in "training" a device is to feed it with a dataset of x-rays. Once the radiographs are fed into the system, dental professionals are asked to "mark" or annotate the diseased areas (caries, gum disease, abscess, cysts, etc.) and categorize them. This allows the software to "learn" about each pathology.
As more and more datasets are fed, the diagnostic accuracy of the software increases (Schwendicke et al., 2022). However, it must also be noted that diagnostic accuracy also depends on the skill and experience of the dentist. To overcome this issue, all dental professionals assigned to annotate dental x-rays are standardized, so they have little difference in their diagnoses.
From Testing to Commercialization
Now, here’s the next question. "Ok, all this is great, but what is the practical use of AI in the dental practice?"
While most of the AI research in dentistry is currently in the theoretical phase, a recent breakthrough has brought a ray of hope. Pearl, an AI-based company offering dental care and vision-related solutions, has recently introduced a software product called Second Opinion®. This software remains true to its name: it gives dentists a diagnostically accurate second opinion in identifying various dental conditions.
Second Opinion® has recently obtained approval from the U.S. Food and Drug Administration (FDA). This approval came after evaluating the software across four clinical studies to determine its accuracy and specificity. Also, SOTA, a digital imaging software, has partnered with Pearl to offer native support for AI-based detection using Second Opinion®.
The software is designed to detect caries, radiolucency marginal discrepancies, periodontal lesions, root canals, fillings, implants, and dental bridges. Second Opinion® has joined the family of FDA-cleared Computer Assisted Detection (CADe) Devices, which are already used for radiologically-related tasks in other areas of medicine.
The Future of Dentistry
The opportunities are enormous with this emerging technology. The commercial introduction of Second Opinion® is the first step in digitizing automatic dental tasks and taking a considerable load off dentists. With further research, we can expect AI to help dentists perform dental procedures and predict treatment outcomes.
If you're not yet subscribed to receive the Incisor newsletter, filled with cutting-edge dental news sent directly to your inbox twice a month, you can do so here.
Author: Dr. Mehmood Asghar is a dentist and Assistant Professor of Dental Biomaterials at the National University of Medical Sciences, Pakistan. Dr. Asghar received his undergraduate and postgraduate dental qualifications from the National University of Science and Technology (NUST). He has recently received a Ph.D. in Restorative Dentistry from the University of Malaya, Malaysia. Besides his hectic clinical and research activities, Dr. Asghar likes to write evidence-based, informative articles for dental professionals and patients. Dr. Asghar has published several articles in international, peer-reviewed journals.
Almalki, Y. E., Din, A. I., Ramzan, M., Irfan, M., Aamir, K. M., Almalki, A., Alotaibi, S., Alaglan, G., Alshamrani, H. A., & Rahman, S. (2022). Deep Learning Models for Classification of Dental Diseases Using Orthopantomography X-ray OPG Images. Sensors, 22(19), 7370. https://www.mdpi.com/1424-8220/22/19/7370
Bhimavarapu, U., & Battineni, G. (2022). Automatic microaneurysms detection for early diagnosis of diabetic retinopathy using improved discrete particle swarm optimization. Journal of Personalized Medicine, 12(2), 317.
Schwendicke, F., Cejudo Grano de Oro, J., Garcia Cantu, A., Meyer-Lückel, H., Chaurasia, A., & Krois, J. (2022). Artificial intelligence for caries detection: value of data and information. Journal of dental research, 101(11), 1350-1356.