Impact of different scan strategies and implant angulation on impression accuracy of full arch multiple implant: an in vitro study

Objective: to evaluate the impact of three different scan strategies and implant angulation on impression accuracy of an intraoral scanner for full-arch multiple implant scan. Material and Method: A maxillary edentulous model with six implant analogs served as a reference model. The four anterior analogs were positioned parallel to each other, the distal right and the distal left was placed with an angulation of 15o and 20o, respectively. Thirty impression were performed using an intraoral scanner (CEREC Primescan). The master cast was digitalized with an industrial reference scanner (ATOS Core 80). All scans were converted to standard tessellation language (STL), superimposed on the reference scan with a 3d inspection software (GOM Inspect Professional 2019) and then analyzed. Results: All linear distances presented equivalence [p<0.01] to those found on the reference scan for all scan strategies. All scan strategies presented a tendency of negative means for linear distances except for d4 in strategy C. All angular distances did not present equivalence [p=0.05] to those found on the reference scan. Significant 3D deviations [p<0.05] were found between strategy B (0.02 ± 0.01) and C (0.05 ± 0.04) for d1. In all others linear and angular distances no statistically significant difference was found between strategies A, B and C. Conclusions: There was no statistically significant difference between strategies A, B and C except for d1 in strategy B and C; Implant angulation did not affect the accuracy of the CEREC Primescan IOS.


INTRODUCTION
Digital impressions present some clinical impacts with respect to conventional impressions, such as lower patient discomfort, time efficiency and simplified procedures for the clinician, especially when multiple implants are present [1,2]. Bohner et al. [3] reported the acceptable accuracy of the current scanning technologies for specific applications and this accuracy depends especially on the scanner technology, object shape and scanning strategy. It was also reported, by the authors, the clinical challenge regarding the scanning of the edentulous arch.
The accuracy of an Intraoral Scanner (IOS), as defined by ISO-5725-1:1994, is a combination of trueness and precision. Trueness describes the deviation of scans from the true dimensions of the object, while precision describes how much separate scans of the same object differ from each other [4]. Data from literature indicate that the accuracy of fully edentulous scans is more difficult to be achieved due to the lack of reference points and the distance between the scan bodies (SB) [5,6]. Once the virtual model accuracy is affected by the camera movement, the correct scan strategy has an import role on fullarch multiple scan accuracy [2,3]. Nevertheless, it remains unclear which is the best scanning strategy of digital impressions [7]. This in-vitro study aimed to assess the impact of three different scan strategies and implant angulation on impression accuracy of an intraoral scanner in a full-arch multiple implants maxilla. The null hypothesis was that there are no differences between the scan strategies and implant angulation for accuracy.

Master cast
A type IV stone master cast (Fujirock; GC America, USA) of an edentulous maxilla with six internal connection analogs (Bone Level NC; Straumann, Switzerland) was fabricated. The median four implants analog were parallel to each other. The distal right was placed with an angulation of 15 o and the distal left of 20 o . The master cast scan was obtained with a high accuracy industrial reference scanner (ATOS Core 80, GOM GmbH, Germany).

Scan strategies
Polyetheretherketone (PEEK) scan bodies (NC scan body; Straumann, Switzerland) were connected to the analogs on the master cast with 10Ncm. Three scan strategies (A, B and C) were used ( Figure 1) and ten repeated impressions were performed (n = 10 per group) with an intraoral scanner (CEREC Primescan; Sirona, Germany). In scan strategy A, first, the occlusal surfaces were scanned. Second, the buccal surfaces from the distal area toward the middle line. Third, the palatal surfaces moving towards the other side of the arch and lastly, buccal lingual rotation are made on the scan bodies in premolar area. Strategy B started at the distal-palatal surface moving towards the other side of the arch. Second, the occlusal surface and returning via the buccal surfaces. Strategy C started at the buccal surfaces at middle line towards the distal area, for both sides, continuing on the occlusal surfaces and returning on the palatal surfaces. All the scans were performed by the same operator.

Comparative analysis
In order to calculate the deviations, six linear and two angular distances were set ( Figure 2). Reference points were related to the cross-section of the center of the scan body cylinder of the master scan and the experimental scans. STL data from group A, B and C were subsequently superimposed on the reference scan with a 3d inspection software (GOM Inspect Professional 2019, GOM GmbH, Germany) using the best fit alignment to assess the three-dimensional (3D) deviation.

Statistical analysis
The mean and standard deviation values were calculated for each group in each test. Data were explored for normality using Shapiro-Wilk test. The Schuirmann test, a two-one sided test (TOST), was used to assess the equivalence (trueness) between the test groups and the reference scan. A 95% confidence interval of 0.1 mm (100 µm) [8] and 0.1 o was set. The analysis of variance (ANOVA) and the Tukey's test were used to compare the differences (precision) between groups A, B and C. The significance level was set at P ≤ 0.05. All the statistical analysis was performed with the SAS 9.4 software (SAS Institute, Cary, NC, USA).  Tables I and II show the mean, standard deviation and confidence interval from the experimental scans for linear and angular distances, in millimeters and degree respectively. Figure 3 shows the values of 3D deviations from the experimental scans to the reference scan for all linear distances. All linear distances presented equivalence [p<0.01] to those found on the reference scan for all scan strategies. All scan strategies presented a tendency of negative means for linear distances except for d4 in strategy C (0.001 ± 0.01). Figure 4 shows the values of 3D deviations from the experimental scans to the reference scan for all angular distances. All angular distances did not present equivalence [p=0.05] to those found on the reference scan. For the alpha angle, strategies A (-0.04 ± 0.27) and B (-0.28 ± 0.50) presented a tendency of negative means while strategy C (0.12 ± 0.51) presented a tendency of positive means. Strategy A presented the lowest mean and standard deviation values while strategy B the highest one. For the beta angle, all strategies presented a tendency of negative means. Strategy B (-0.24 ± 0.28) presented the lowest mean and standard deviation values, strategy A (-0.39 ± 0.35) the highest one and strategy C presented mean and standard deviation of 0.30 ± 0.32. Figure 5 shows the absolute values of 3D deviations from the experimental scans to the reference scan for all linear distances. Significant 3D deviations [p<0.05] were found between strategy B (0.02 ± 0.01) b and C (0.05 ± 0.04) a for d1. There was no significant difference between strategies A (0.03 ± 0.02) ab and B (0.02 ± 0.01) b and strategies A (0.03 ± 0.02) ab and C (0.05 ± 0.04) a . In all other linear (d2, d3, d4, d5 and d6) and angular (α and β) distances no statistically significant difference was found between strategies A, B and C. Figure 6 shows the absolute values of 3D deviations from the experimental scans to the reference scan for all angular distances.

DISCUSSION
The aim of this study was to compare the accuracy of three different scan strategies and the influence of implant angulation for full-arch multiple implant using the Primescan IOS. The null hypothesis was not rejected.
The reference scanner (ATOS Core 80) is an industrial structured blue light 3D scanner, complies with ISO 12836 and shows accuracy up to 3µm, with a precision of 2µm for jaw-sized scans [9][10][11]. A sample size of 10 scans for each scanning strategy has been demonstrated as sufficient to obtain consistent statistical results by several authors [11][12][13][14][15][16].
Some published reports have evaluated the scan strategy in complete-arch scanning and digital intraoral scanning [7,11,17,18] However, there is still a lack of studies that evaluate the scan strategy in full-arch multiple implant scan [6,[19][20][21]. To the knowledge of the authors, this is the first study using the CEREC Primescan to evaluate the impact of different scan strategies and implant angulation on scan accuracy in full arch multiple implant impression. There are no other studies using the same measuring method. Thus, the lack of available literature makes it difficult to compare the results obtained in this study.
Although a study [22] using a 0.4°as an acceptable angle deviation between implants was found, the clinically acceptable threshold for the angle deviations is not yet defined in the literature [4]. In the present study a confidence interval of 0,1 o was set. The statistical analysis suggest that equivalence would be achieved if a 0.5 o confidence interval was set for all strategies.
An in vitro study investigated the trueness and precision under repeatable conditions for different IOSs (Trios 3, Trios 3 Mono and Itero Element) when scanning fully edentulous arch with multiple implants and reported that precision is low for the tested IOS devices [19]. A more recent in vitro study evaluated the accuracy of digital complete-arch edentulous implant scanning and the influence of the different extents of surrounding movable soft tissue and reported that the accuracy of the IOSs tested (Trios 3, Trios Color, CEREC Omnicam and CEREC Primescan) was comparable with that of the conventional impression technique and that the amount of flexible soft tissue interference affected the accuracy of the digital scans [21]. This is an in vitro study where clinical conditions that could compromise the results of the scanning process such as presence of saliva, tongue, cheek, limited mouth opening and patient movement have not been evaluated. Although there wasn't statistical significance difference between strategies A, B and C, strategy B should be preferred in clinical practice since has fewer steps. Further in vivo studies should be conducted to validate the accuracy of this IOS device under clinical conditions.

CONCLUSION
Within the limitations of this in vitro study, the following conclusions were drawn: 1-There was no statistically significant difference between strategies A, B and C except for d1 in strategy B and C; 2-Implant angulation did not affect the accuracy of the CEREC Primescan IOS.