Artificial Intelligence Image Analysis Software In Periodontal Resident Training

To investigate whether AI software with traditional case-based learning improve residents' evaluation skills regarding alveolar bone resorption ①A 4-week trial involving 32 dental residents compared AI-assisted learning (AI group) with conventional training (control group). ②Both groups treated periodontal patients 5 days/week and attended weekly case-based learning sessions. The AI group additionally used AI software for panoramic radiograph analysis during clinical practice. ③After the intervention, an 11-case online radiograph interpretation exam was administered. ​④Multivariate Analysis of Variance (MANOVA) was used to assess between-group differences, with statistical significance set at p<0.05. ①Demographics (age/sex/role) didn't affect training outcomes. ②The AI group showed superior performance in assessing alveolar bone resorption (196.2 ± 60.1 vs. 310.8 ± 112.6, p = 0.001), identifying periapical lesions (66.2 ± 12.8 vs. 51.9 ± 14.7, p = 0.011), and diagnosing hopeless teeth (76.5 ± 17.7 vs. 60.4 ± 18.8, p = 0.027). ③Participants specifically praised AI's accuracy for evaluating bone levels and identifying hopeless teeth. The application of AI significantly enhances residents’ ability to assess alveolar bone resorption, identify periapical lesions, and diagnose teeth with poor prognoses.