PeriaTrace generates tooth-level, patient-friendly reports from dental X-rays
A three-stage AI framework links periapical radiograph findings to image evidence and plain-language explanations, according to an arXiv abstract describing internal and external evaluation.
In plain words
Dentists use periapical X-rays, close-up images of individual teeth, to diagnose problems, but writing reports takes time and the technical wording can confuse patients. PeriaTrace is an AI system that identifies findings at the tooth level, marks where they appear in the image, and adds plain-language explanations and care recommendations. In the reported evaluation, most reports needed only minor review by specialists.
What was announced
The authors describe PeriaTrace as a three-stage framework combining tooth-level diagnosis, visually grounded report generation, and patient-friendly explanations with care recommendations. It was developed and evaluated on 12,877 periapical radiographs from an internal dataset plus two external clinical cohorts. Internal results included a macro Dice coefficient of 85.35%, BLEU-4 of 0.6504, an F1 score of 94.37% for joint tooth localization and FDI identification, and a grounding-tag F1 of 77.93%.
In the two external cohorts, senior endodontists judged 91.16% and 84.52% of reports suitable for retention or revision without complete rewriting, and generation averaged 4.41 seconds per radiograph, which the authors describe as several tens of times faster than manual drafting. In a survey of 196 participants without medical backgrounds, 96.4% preferred reports with plain-language explanations and care recommendations, and 84.7% found visual localization cues helpful. The authors state the system may support clinical review and patient communication; the published material is an abridged abstract, so full methods and peer review status are not shown.
What it does not show
- Evidence is limited to an abridged arXiv abstract; full methods, datasets details and peer review are not available in the supplied passage.
- Report quality assessments come from the system's developers and survey participants, not from independent confirmation.
Sources
Source-support review completed 19 Nov 2026. This story was drafted with AI assistance and checked against retained source passages. Automated checking is not independent verification.
Coverage: arXiv: abstract.
Updates and corrections
No corrections so far.
First published 19 Nov 2026
New to the subject? AI from the Ground Up starts from the beginning.