An AI-based diagnostic system has been pitted against dentists to evaluate the effectiveness of the system and to see how accurately it was able to diagnose tooth decay from radiographs.
The AI system, developed by Pearl, was used on a set of 8,767 bitewing and periapical radiographs and was compared with the evaluations of three experienced dentists. The main aim of the study was to validate the effectiveness of the AI system; however, Pearl found that the AI system diagnosed tooth decay with greater consistency and accuracy.
Another finding of the study was there was often a lack of consensus between dentists. There was only 79% unanimous agreement about the lack of tooth decay, and only unanimous agreement about the presence of tooth decay in just 4.2% of the total. In almost one in five cases where two of the dentists identified tooth decay, the other did not.
The AI-system was found to be as accurate as the diagnostic conclusion of any single dentist, with the system tending to agree more with the human dentists than they did with each other, which Pearl attributes to the sensitivity of the system at detecting tooth decay.
“Our intention in producing this study was simply to demonstrate the efficacy of computer vision machine learning diagnostics in dental radiology, but the secondary findings dredge up a major deficiency in dental healthcare,” said Pearl CEO Ophir Tanz. “While diagnostic inconsistency may be a natural byproduct of human professionals making case-by-case judgements, at scale inconsistency in the standard of care results in suboptimal patient treatment with broader implications for population health at large.”
The system is not intended to replace diagnoses by dentists, but to act as a diagnostic tool to improve the accuracy of diagnosis. While the findings of the study validate the effectiveness of the system at detecting tooth decay, further studies are required to determine whether the system is better than human dentists at detecting tooth decay, especially for varying degrees of decay. With more data on the effectiveness of the solution it would be easier to convince dentists to invest in AI technology.