Accuracy of complete‐arch implant impressions obtained with intraoral scanning, calibrated intraoral scan protocol, extraoral and intraoral photogrammetry, and conventional impression techniques: An in vitro study
Jiayi Lu, Jiayi Yu, Ashraf Adi, Berna Saglik, Fei Liu, Junying LiAbstract
Purpose
To compare the accuracy of five techniques for complete‐arch implant impressions: conventional open‐tray splinted impression technique (CNV), direct intraoral scanning (IOS), calibrated intraoral scan protocol (CISP), extraoral photogrammetry (EPG), and intraoral photogrammetry (IPG).
Materials and Methods
A maxillary edentulous model with six implants was used. Scan bodies were connected at the abutment level, and desktop scanning (DSK) data served as the reference dataset. Five impression techniques (CNV, IOS, CISP, EPG, IPG) were performed 10 times each. DSK was repeated 10 times and included as a group only for precision evaluation. STL datasets were aligned using two strategies: alignment of all scan bodies to evaluate overall fitness and alignment of the first and second scan bodies to simulate a virtual Sheffield fit test. Linear deviation, angular deviation, and root mean square (RMS) error were calculated to evaluate trueness and precision. Nonparametric statistical tests were applied ( α = 0.05).
Results
For trueness, when aligned using all scan bodies, EPG demonstrated the highest trueness with the lowest linear deviation (median: 24.40 µm) and angular deviation (median: 0.07°). CISP (median: 38.26 µm) and CNV (median: 39.31 µm) showed comparable linear deviations, both significantly lower than those of IOS (median: 48.37 µm). RMS errors were similar among EPG (median: 19.47 µm), IPG (median: 19.86 µm), and CISP (median: 21.55 µm) and were smaller than those of IOS (median: 27.30 µm). When aligned using SB1 and SB2, linear deviation and RMS error increased toward distal implants in all groups. At SB6, median linear deviations reached 157.82 µm (IOS), 194.13 µm (EPG), and 188.70 µm (IPG), whereas CNV and CISP remained below 100 µm. For precision, EPG showed the highest precision with the lowest linear deviation (median: 7.77 µm), angular deviation (median: 0.02°), and RMS error (median: 4.18 µm), whereas IOS demonstrated the lowest precision (linear deviation: 66.59 µm) using overall fit test. When aligned by SB1 and SB2, EPG showed the highest stability and outperformed the other groups through all scan body sites.
Conclusion
Extraoral photogrammetry demonstrated the highest overall accuracy, whereas the calibrated intraoral scan protocol and conventional impressions showed smaller distal deviations under constrained alignment conditions.