Do digital impressions have a greater accuracy for full-arch implant-supported reconstructions compared to conventional impressions? An in vitro study

Abstract

Aim:

The purpose of this study was to compare the accuracy of conventional implant impressions with digital impression techniques made using two different intraoral scanners.

Setting and Design:

In-Vitro study.

Material and Methods:

A scan of master cast containing four implants was made using two intraoral scanners: CEREC Primescan (Dentsply Sirona, USA) and 3Shape Trios (Copenhagen, Denmark) with PEEK scan bodies attached to the implants. Model was scanned ten times using different scanners. The accuracy of the chairside scanners was compared with highly accurate laboratory scanner. The scans were transferred into the software (Geomagic Control X 20, 3D Systems, Rock Hill, SC, USA) for analysis. The linear deviations and the angular deviations between the scans (scan of each model made using high-definition scanner and the master model scan) were calculated to determine the accuracy. Trueness was used as a parameter to compare the accuracy of different scanners (comparing test and reference).

Statistical Analysis:

Analysis of variance was performed with Bonferroni's post hoc test for multiple group comparisons.

Results:

Distribution of the mean overall absolute linear deviation was significantly lower in the conventional impression group compared to the CEREC Primescan scanner group and 3Shape Trios group (P < 0.05 for both). Distribution of the mean overall absolute linear deviation was significantly lower in the CEREC Primescan scanner group compared to the 3Shape Trios group (P < 0.05). Distribution of the mean overall absolute angular deviation did not differ between the three groups (P > 0.05 for all).

Conclusion:

Conventional impressions showed significantly greater accuracy compared to the digital impressions made with both the above intraoral scanners for implant-supported restoration of an edentulous arch. In addition, the digital impressions with the CEREC Primescan scanner showed greater accuracy as compared to the 3Shape Trios scanner.

INTRODUCTION

Dental implants are commonly used as an alternative to conventional partial or complete dentures for individuals with missing teeth, as they offer higher function, retention, and ease of use.[1,2]

Models made out of gypsum, poured from a physical elastomeric impression material, have been employed commonly to make implant-retained prostheses. To achieve an accurate prosthetic fit, the transfer of implant angulation and position is imperative.[3] Over the years, different techniques and materials have been used and evolved to improve and accurately replicate the implant position from the intraoral situation to the master cast.[4] If this step is not performed accurately, then it could lead to duplication of errors in the following steps of prosthesis fabrication.[5,6,7,8,9,10]

Digital implant dentistry has transformed the way impressions are recorded, and more importantly, the laboratory protocols followed thereafter. With the advent of digital impressions, the workflow of prosthetic reconstruction has been simplified by elimination of multiple steps such as tray selection and shipping to the laboratory. This has reduced the treatment time and has improvised patient compliance.[11,2,13]

A digital impression file eliminates storage issues as it is stored in digital library, which enables an efficient record keeping with a paper-free practice. Other than the learning curve in learning and using the new technology, there are financial limitations like the purchasing cost of an intraoral scanner. There are many scanners available for making digital impressions. These work on different image-capturing principles, and hence, their accuracies may not be the same.[14,16,17,19,18] The CEREC Primescan (Dentsply Sirona, USA) is an intraoral scanner that uses a white light for pattern projection onto an object; this concept is known as active triangulation. The images are captured in color continuously, eliminating the need of contrast spraying.[19] The Trios scanner (3Shape, Copenhagen, Denmark) is designed on the concept of confocal microscopy that records images from different positions in a continuous manner to create a 3D image. The latest model records color data without contrast spraying.[19] Accuracy has been described in the literature using two parameters such as trueness and precision (ISO 5725-1). Trueness describes the closeness to the actual dimensions of the object.[20,21,22] Precision is represented by the degree of reproducibility between repeated measurements.

Accuracy of scanners and conventional impressions have been previously described in the literature.[6,7,8] However, studies comparing the accuracy of the scanner with working on different principles that as optical triangulation and confocal microscopy on axial and tilted implants, used to restore edentulous arches, are still not reported adequately in the literature.

This study analyzes both the linear deviation and the angular deviation to evaluate the difference in the accuracy of conventional implant impressions and digital impression techniques made using these two different intraoral scanners.

MATERIALS AND METHODS

Fabrication of master model and master control STL files

Four dental implants, Bone Level Tapered 4.1 mm × 10 mm (RC, SLActive, Straumann AG, Switzerland, Basel), were placed in a sawdust model of an edentulous mandible. Anteriorly, implants were placed straight, and posteriorly, implants were placed at a 10° distal[11,12,13] angulation. This served as the master model [Figure 1].

Figure 1.

Sawdust model of an edentulous mandible with implants placed (control model)

Four scan bodies (Cares^® RC Mono Scanbody, RC, BLT, Straumann, Basel, Switzerland) were then connected to the implants and tightened as recommended by the manufacturer [Figure 2]. The master model was scanned using a high-definition scanner (Artec 3D, Luxembourg, Europe) to obtain a STL file. This was the control STL file [Figure 3].

Figure 2.

Scan bodies are attached to the analogs (Cares^® RC Mono Scanbody, RC, BLT, Straumann)

Figure 3.

Control STL file

There were three study groups (N = 30):

1. Group A: Impressions made by conventional technique (n = 10)

2. Group B: Impressions made by intraoral scanner CEREC Primescan (Dentsply Sirona, USA) (n = 10)

3. Group C: Impressions made by intraoral scanner Trios 3Shape scanner (3Shape, Copenhagen, Denmark) (n = 10).

Group A

Using the open-tray impression technique, implant-level copings were fixed to the implants on the control/master model. Splinting of the open-tray impression copings was done using self-cure acrylic resin (Pattern Resin LS, GC America). Tray adhesive (Impregum; 3M ESPE, USA) was applied onto the intaglio surface of the custom tray. The impression was made only after drying the tray adhesive for 15 min. Using polyether impression material (Impregum; 3M ESPE, USA), ten impressions per group (A, B, and C) were made following standard procedure. The lab analogs were attached to the copings, and ten models were made. Scan bodies were fixed to each of the analogs, and each model was then scanned with the high-definition scanner (Artec 3D, Luxembourg, Europe), and the data, in the form of 3D images, were created and exported as an open-source STL file.

Group B

Using the scan bodies (Cares^® RC Mono Scanbody, RC, BLT, Straumann, Basel, Switzerland), digital impressions were made with CEREC Primescan (Dentsply Sirona, USA), ten times, according to the manufacturer's instruction and exported as STL files [Figure 4].

Figure 4.

STL file created by CEREC Prime scanner

Group C

Using the same scan bodies (Cares^® RC Mono Scanbody, RC, BLT, Straumann, Basel, Switzerland) in place, ten digital impressions were made by the intraoral scanner 3Shape Trios (Copenhagen, Denmark) and exported as STL files [Figure 5].

Figure 5.

STL file created by 3Shape Trios

Data analysis

All the scans were transferred into the metrology software (Geomagic Control X 20, 3D Systems, Rock Hill, SC, USA) for data analyses. Best fit algorithm was used; the tolerance was set at 1 μm; the control STL file of the master model [Figure 3] was superimposed to the four scan bodies and saved as a new STL file. This method was allowed for comparing the scan bodies only, minus the other irrelevant areas. As Ender and Mehl defined,[23] accuracy comprised the following two parameters: trueness depicts the degree of resemblance between the test scan and the scan taken by the scanner, while precision describes the variation between the test scans. The primary objective was, therefore, to ascertain and evaluate the accuracy, which includes trueness at the level of the scan bodies. Test scans and control scans were superimposed [Figures 6 and 7] using an algorithm with the tolerance set at 1 μm. Following this, a 3D comparison was made, calculating the linear [Table 1] and angular [Table 2] mean deviation from the mean positive and negative deviation using the methodology previously described by Papaspyridakos et al. (2016).[24]

Figure 6.

Superimposition and measurement of linear deviation

Figure 7.

Superimposition and measurement of angular deviation

Table 1.

Distribution of the mean absolute linear deviations in different groups studied

Implant positions studied	Group A Conventional impression technique (n=10)		Group B Scan using CEREC Primescan (n=10)		Group C Scan using 3Shape Trios (n=10)
	Mean	SD	Mean	SD	Mean	SD
A–B	0.059	0.038	0.095	0.038	0.073	0.047
B–C	0.038	0.037	0.127	0.057	0.248	0.107
C–D	0.061	0.023	0.057	0.031	0.050	0.037
A–D	0.020	0.060	0.226	0.063	0.316	0.129
A–C	0.131	0.048	0.173	0.029	0.217	0.061
B–D	0.113	0.041	0.174	0.034	0.351	0.046
Average	0.101	0.032	0.142	0.018	0.209	0.034

Absolute deviation in mm. SD: Standard deviation

Table 2.

Distribution of the mean angular deviations in different groups studied

Implant positions studied	Group A Conventional impression technique (n=10)		Group B Scan using CEREC Primescan (n=10)		Group C Scan using 3Shape Trios (n=10)
	Mean	SD	Mean	SD	Mean	SD
A–B	0.664	0.492	0.436	0.603	1.446	0.807
B–C	0.751	0.711	2.066	0.876	0.576	0.424
C–D	0.791	0.363	0.843	0.520	1.544	0.327
A–D	0.855	0.765	1.594	0.942	1.309	0.985
A–C	0.838	0.571	1.219	0.452	2.042	0.872
B–D	0.857	0.588	1.229	0.823	1.541	0.654
Average	0.793	0.329	1.231	0.309	1.409	0.752

Absolute deviation in degrees. SD: Standard deviation

RESULTS

Distribution of the mean overall absolute linear deviation was statistically significantly lower in the conventional impression group as compared to the CEREC Primescan scanner and 3Shape Trios groups (P < 0.05 for both) [Table 3 and Figure 8].

Table 3.

Intergroup statistical comparison of distribution of the mean absolute linear deviation in different groups studied

Intergroup comparisons (P)
Conventional versus CEREC Primescan	Conventional versus 3Shape Trios	CEREC Primescan versus 3Shape Trios
0.012*	0.001***	0.001***

P-value by ANOVA with Bonferroni’s post hoc test for multiple group comparisons. P<0.05 is considered statistically significant. *P<0.05, ***P<0.001. ANOVA: Analysis of variance

Figure 8.

Distribution of the mean linear deviations in different groups studied (absolute deviation in mm)

Distribution of the mean overall absolute linear deviation was statistically significantly lower in the CEREC Primescan scanner group as compared to the 3Shape Trios group (P < 0.05).

Distribution of the mean overall absolute angular deviation did not differ significantly across the three types of scanner groups in intraoral model (P > 0.05 for all) [Table 4 and Figure 9].

Table 4.

Intergroup statistical comparison of distribution of the mean angular deviation in different groups studied

Intergroup comparisons (P)
Conventional versus CEREC Primescan	Conventional versus 3Shape Trios	CEREC Primescan versus 3Shape Trios
0.355 (NS)	0.094 (NS)	0.999 (NS)

P-value by ANOVA with Bonferroni’s post hoc test for multiple group comparisons. P<0.05 is considered statistically significant. NS – Statistically nonsignificant

Figure 9.

Distribution of the mean angular deviations in different groups studied (absolute deviation in degrees)

DISCUSSION

This study evaluated the linear and angular deviations produced by the three groups by comparing them to a master model which underwent scanning by a laboratory scanner (Artec 3D Space Spider, Luxembourg, Europe). Su and Sun compared the accuracy of 3ShapeTrios scanner and with a laboratory scanner by evaluating the precision between the two (Lava Scan ST). They found that the precision was significantly lower for 3Shape Trios, and the deviation was directly proportional to the number of teeth scanned during the procedure.[22]

The results of this study demonstrated the highest linear deviation for 3Shape Trios, followed by CEREC Primescan with the conventional impressions showing the least deviation in comparison to master model. A statistically significant difference was noted between the conventional impression group and CEREC Primescan (P < 0.05), between conventional impressions and 3Shape Trios, and between CEREC and 3Shape Trios (P < 0.001) groups regarding linear deviation. 3Shape Trios demonstrated the highest angular deviation at impression and scan level, followed by CEREC Primescan and conventional impression, respectively. Comparison of angular deviation at impression and scan level was found to be statistically insignificant in this research.

Digital impressions may have varied accuracy levels which largely depend on multiple factors such as scanning technique, size of the scan field, the angulation, number of implants, and the scan body fit.[25,26] The results of this study are in agreement with the studies conducted by Papaspyridakos et al. and Ender and Mehl where the authors conclude that there was statistically no significant difference observed between the accuracy of conventional and digital impressions.[23,24] However, studies conducted by Giménez et al. showed that the scanner recorded the first quadrant rather accurately, whereas for the second quadrant, the trueness significantly decreased.[27] Stimmelmayr et al. noted that there was a statistically significant difference in the scan body fit between laboratory analogs and implants.[28]

In this study, the conventional impressions were noted to be more accurate as compared to intraoral scans. This could be attributed to the fact that the open-tray splinted impressions have known to have a higher accuracy as compared to other impression techniques. Similar results were noted in several studies that reported high accuracy of open-tray splinting impression technique for internal connection implants.[29,30,31,32] The splinting of open-tray posts to each other does not permit any movement of the posts while making or retrieving the impression. Scan was made with high accuracy lab scanner. Furthermore, it has been observed that digital workflow has its own operator-based challenges. A study conducted by Giménez et al. reported that digital impression making has its own learning curve and the clinician needs adequate practice to reproduce or make precise intraoral scans.[33] This study compares both the liner deviation and the angular deviation of the impressions made using conventional method and digital intraoral scanners.

The limitations of this study are as follows: (i) owing to its in vitro design, this study oversimplifies impression making as the scans are recorded from a simplified model, where the implants were placed linearly; and (ii) intraoral scanning may show increased inaccuracies intraorally owning to the highly contrasting environments.[34]

The other difference lies in the stability of the surfaces scanned. The soft-tissue texture and form varies depending on the patient's jaw movements, thereby complicating the procedure of the scanning because it depends on the presence of reference points which are fixed (Andriessen et al.).[35] Similarly, it has been observed that an increase in interimplant distance along with a flat and dynamic mucosal surface results in an insufficiency of definitive reference points to enable accurate stitching Giménez et al.[33] In this research, the implant positions were near to one another. The implication, therefore, would be that the interimplant distance is directly proportional to the scanning difficulty, and therefore, indirectly proportional to the accuracy. Clinically, biological factors such as saliva, gingival fluid, blood, breathing pattern, and movements of the tongue are some of the factors that contribute to reduction in accuracy.[34] Furthermore, the use of high-definition scanner for the master model and the conventional impression was a confounding factor in the study.

In addition, another limitation is that only a single implant system was used. Further studies should be carried out in a clinical setup with different implant systems and scanners of different technology specifications as well before clinical recommendations can be made for the treatment of an edentulous patient. Future studies should evaluate the accuracy of implants placed with higher angulation.

Pertaining to the clinical scenario, intraoral scanners show a great potential to physical impressions for implant prosthesis. However, for full edentulous situations, especially with a greater interimplant distance, a conventional open-tray impression with splinted impression posts may be the most accurate solution as the intraoral scanners do not get enough reference points in the edentulous arch and this leads to further inaccuracies. Furthermore, virtual images obtained can be printed or milled into physical models to draw a comparison with stone models which help establish a framework for the assessment of the clinical results. The ITI consensus statements also state that for edentulous impressions, the use of scans is not still recommended.[36]

CONCLUSION

The following can be concluded based on the research performed in this study:

1. The conventional impressions showed a high level of accuracy for implant-supported restoration of an edentulous arch

2. Digital impressions made using the scanner that works on optical triangulation principle and uses white LED light had a greater accuracy as compared to impressions made using the scanner working on the principle of ultrafast optical scanning and confocal microscopy

3. When all the three impression techniques were compared, conventional impressions showed significantly greater accuracy compared to the digital impressions made with both the above intraoral scanners for implant-supported restoration of an edentulous arch.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.