AI Detected Periodontitis on Radiographs With High Accuracy Across 50,080 Images

Edited by Dentaltown staff

Artificial intelligence models identified periodontitis on radiographs with pooled sensitivity above 0.90, but the authors said the tools are not ready for routine clinical use, according to a systematic review and diagnostic test accuracy meta-analysis published Sept. 17 in the Journal of Dentistry.

Fourteen studies comprising 50,080 radiographic images met inclusion criteria. Seven made diagnoses at the patient level and seven at the tooth or image level.

At the patient level, pooled sensitivity was 0.93 and pooled specificity 0.88, with an area under the summary receiver operating characteristic curve of 0.96. Tooth-level performance was comparable, at a pooled sensitivity of 0.90, specificity of 0.94, and area under the curve of 0.97.

Meta-regression found that algorithm architecture, imaging modality, and whether a model had undergone external validation did not account for the variation among studies.

The authors described prospective studies with robust clinical reference standards, external validation, and transparent reporting as urgently needed before the tools are integrated into routine dental care. They also noted that the current evidence base underrepresents low- and middle-income countries, where the burden of periodontitis is greatest.

In a note on clinical significance, the authors placed the strongest case for radiographic AI with general dental practitioners and non-specialist clinicians in primary care or resource-limited settings, where access to periodontists is scarce and diagnostic variability is widest. They framed the tools as a way to flag suspected moderate-to-severe disease chairside and speed referral, rather than as a replacement for specialist examination.

Researchers at the School of Stomatology at Lanzhou University in China conducted the review. Searches covered PubMed, Web of Science, and the Cochrane Library through April 2026, and pooled estimates were derived using a bivariate random-effects model.

Sources:
Journal of Dentistry, “Artificial intelligence for detection of periodontitis on radiographic images: a systematic review and diagnostic test accuracy meta-analysis,” published online Sept. 17, 2026: doi.org/10.1016/j.jdent.2026.107054
PubMed record, PMID 42753854: pubmed.ncbi.nlm.nih.gov/42753854