[Construction of learning curves for digital impression and their application in prosthodontic education]

Objective: To construct learning curves for different intraoral scanners and compare their scanning performance and operator experience, thereby providing evidence for preclinical training design. Methods: A prospective randomized controlled trial enrolled 30 third-year undergraduate dental students from School of Stomatology, Wuhan University with no prior experience were block-randomized into three groups (each n=10) to perform 15 simulated full-arch scans with scanner A (Trios 5), B (Trios 3) and C (AoralScan3) respectively. Scanning time, number of photograms and difficulty questionnaire data were collected, with subsequent learning curve fitting, scanning accuracy was evaluated using the root mean square error (RMSE) and Spearman correlation analysis. Results: Scanning time decreased significantly with practice in all groups (P<0.001). Groups A, B, and C reached proficiency plateaus after 10, 11, and 12 scans, with mean times of (663±36), (765±43), (821±65) s respectively. There was no significant differences in scanning accuracy for the maxilla or mandibular among the three groups (P>0.05). Questionnaire analysis showed significant changes in self-efficacy, perceived fluency, and perceived difficulty across training sessions (all P<0.001). Correlation analysis revealed a strong negative association between the number of image frames and perceived fluency (A: r=0.65,P<0.001; B: r=-0.66, P<0.001; C: r=-0.75, P<0.001), whereas no stable correlation was observed between the number of image frames and scanning accuracy (all P>0.05). Conclusions: Novices' scanning efficiency improves with practice. Learning patterns differ among scanners, suggesting that preclinical training should be tailored to device-specific characteristics.