Accuracy of impressions for multiple implants: A comparative study of digital and conventional techniques

Intraoral scanning has benefits over conventional impression making, but whether scanning is sufficiently accurate for multiple implants is unclear.

The purpose of this in vitro study was to compare the trueness of digital scans acquired by using intraoral scanners from a small range to a complete arch with the conventional impression technique and to determine the influence of 2 different evaluation methods (best-fit algorithm versus absolute linear deviation) on the outcomes of accuracy assessment.

A mandibular model with 8 implants (A-H) around an edentulous arch was used as the master model. Open-format standard tessellation language (STL) data sets (1 reference file from a highly accurate dental laboratory scanner, 10 files from an intraoral scanner, and 10 files from digitized conventional impressions at room temperature) were imported to a metrology software program, and 5 groups of scanning ranges (AB, FGH, CDEF, BCDEFG, and ABCDEFGH) were identified simulating different clinical situations. Two evaluation methods-root mean square values calculated from the best-fit algorithm and average value of linear discrepancies from absolute linear deviation-were used to describe the trueness values. The impacts of different scanning or impression methods, ranges, and evaluation methods were tested by using a 3-way ANOVA. The effect of the scanning range on accuracy was further identified with 1-way ANOVA. The paired-sample t test was used to determine the differences of trueness values between the 2 methods in different groups.

The trueness values of the implant impressions were significantly affected by different scanning or impression methods (P<.001), evaluation methods (P<.001), and scanning ranges (P<.001) as independent variables. With use of the best-fit algorithm, deviations from the digital scans were significantly greater than those from the conventional impressions in cross-arch situations (groups CDEF, BCDEFG, and ABCDEFGH). With use of the absolute linear deviation method, statistically significant lower accuracy was found when larger areas were encountered (groups BCDEFG and ABCDEFGH). Use of the absolute linear deviation method resulted in a higher mean score of inaccuracy than that from the best-fit algorithm method in most situations.

Scanning or impression methods, ranges, and evaluation methods affected the dimensional accuracy (trueness) of scans or impressions with multiple implants. Digital scans had worse trueness values compared with those made with the conventional splinting open-tray technique when cross-arch implant impressions were acquired.