Improving the oral health of the public and the professional development of ASDA members

News

Can AI be used to detect periapical radiolucencies?

Reprinted from ADA Morning Huddle, June 27, 2024:

Investigators have examined whether artificial intelligence may help dentists detect periapical radiolucencies on imaging, according to a systematic review and meta-analysis published in the Journal of Dentistry.

Previous research has found inconsistencies in the accuracy of diagnosing periapical radiolucencies.

In the study, the investigators used the PubMed/MEDLINE, ScienceDirect and Institute of Electrical and Electronics Engineers Xplore databases to analyze the data of 24 articles focused on periapical radiolucency detection. They noted that 23 of the studies utilized a convolutional neural network to assist in diagnosing the radiolucencies.

Among the four articles included in the meta-analysis, AI demonstrated a pooled sensitivity of 0.94 and specificity of 0.96 — representing the potential to support dentists in periapical radiolucency detection.

However, the investigators emphasized that more diverse studies, such as prospective, real-life randomized controlled trials, may be needed to further understand the benefit of using AI for this application.

Billy Tarpley

About Billy Tarpley

Serve as chief lobbyist for the dental association, which was established in 1887. Oversee all legal and regulatory matters, as well as publications, continuing dental education, conference planning and governance activities. Between 2005 and 2009, spent 4 years as Director of Development at Henderson State University (alma mater), but returned to the association in 2009.