AI can help dentists detect cavities more reliably

AI use for dental X-rays
Photo: iakovenko 123rf

Researchers in the Czech Republic have developed an artificial intelligence (AI)-based method for more reliable detection of tooth decay. The new method, which is based on deep learning, was able to match the accuracy of experienced dentists in diagnosing cavities from X-ray images.

The research by a team from the Czech Technical University (CTU) and Charles University (CU)—and published in two parts (1 & 2) in Clinical Oral Investigations—compared the performance of the automatic caries detection method with that of seven dentists, three beginners and four experienced professionals. 

“Their task was to mark tooth decay on a hundred X-rays,” Professor Jan Kybic said. 

“The agreement between the experts was relatively low. As a result, the automatic method, which uses deep learning, consistently outperformed the novices and achieved similar or better results than even very experienced dentists.”

“Unlike other medical disciplines, dentists make and interpret most of their own X-rays themselves,” Dr Antonín Tichý added.

“However, the evaluation of X-rays is to some extent subjective and can vary considerably between doctors, which was confirmed in our study. In this respect, AI-based automatic caries detection can reduce the risk of overlooking caries, for example, and visualisation of caries on the image can facilitate communication between the doctor and the patient.”

A dataset of image data and a neural network from almost four thousand anonymised X-rays with more than seven thousand tooth decays was created and then trained to detect tooth decays as accurately as possible. The entire process based on AI tools took about two years of work.

The study has shown that the automatic method can improve the accuracy and repeatability of caries detection and thus provide a useful second opinion even for experienced dentists.

“In radiology, the practice is that two doctors look at the images, and if they disagree, a third can come in,” Professor Kybic said. 

“But a dentist in a typical practice does not have this option, as he is usually alone, so our method could help him the most, especially if he is a beginner with less clinical experience.”

Previous articleRunning a fly-in fly-out dental practice
Next articleStudy reveals anaesthetic injections main stressor in teen dental patients


Please enter your comment!
Please enter your name here