A Retrospective Cohort Study on Scan Quality of Implant Scanbodies Matched With CAD Libraries

ABSTRACT

Purpose

To assess the effect of scanbody (SB)‐type, edentulous site, and restoration‐type on the scan quality of SBs used in the treatment of short‐span edentulism.

Materials and Methods

The cohort consisted of SBs with different specifications connected to bone‐level implants for intraoral digitalization in the fabrication of fixed restorations. SBs matched with library CAD files for digital implant position transfer into dental CAD software were enrolled in the study group. Intraoral implant digital records were categorically evaluated to assess the quality of SB scans. In statistical analyses, the chi‐squared test was used to describe the clinical variables, and logistic regression models were constructed to reveal the association between the clinical variables and SB scan quality.

Results

A total of 243 SBs were eligible for scan quality evaluation. Scan quality did not differ statistically (p > 0.05) in the SB reference area, while texture in the representation of SB was significantly affected (p < 0.05) by the variables SB‐type and edentulous‐site. Cylindrically designed SBs without specific geometrical features presented remarkably higher risks for reduced scan quality in SB representation.

Conclusion

SBs successfully aligned with library CAD files based on a software algorithm may not consistently present similar scan quality. Intraoral scanning of a SB is highly vulnerable with regard to scan deterioration in texture and geometry.

1. Introduction

Intraoral scanners are commonly used in the treatment of short‐span edentulous areas for both tooth‐ and implant‐supported fixed dental prostheses (FDP) [1, 2]. However, the digitization of dental implants significantly differs from tooth surfaces not in scanning principles but in requirements for appropriate data acquisition. This is basically related to the necessity to accurately transfer the 3‐D spatial position of the implant in the jaw. To facilitate this, a transfer device known as scanbody (SB) is used to capture and integrate the implant's position into the dental arch scan [3]. Fundamentally, this is a data exchange process carried out in the dental CAD software interface to initialize restorative design. SBs are characterized by a variety of features that are related to both hardware and software specifications.

The hardware‐related specifications of a SB can be described in three sections. The scan region and the base present the functionality, which the former serves as the primary reference area for dental CAD software matching, and the latter connects SB either to implant or abutment. Third, the body provides the connection between the scan region and the base [3].

The software specification feature involves activating the SB library through precise alignment between the scan file and its geometrically perfect CAD file [4, 5]. Subsequent to automated virtual 3D implant positioning, restoration design can be executed utilizing the relevant library elements.

User defined factors affecting the accuracy of intraoral scanning data are fundamentally scanning strategy [6], ambient lighting conditions [7], and operator experience [8]. In addition to these general factors, the impact of SB hardware specifications coupled with implant location, angulation, and number [9, 10] on scanning accuracy remains debatable.

SBs can be designed as either single‐piece or two‐piece configurations. Single‐piece designs are typically composed of metal or PEEK, whereas two‐piece designs involve a combination of two materials, often with a metal implant connection in such designs. The stability of the implant‐SB connection has been hypothesized to influence 3D positional transfer [3]. Conversely, Sawyers and Baig [11] reported that reuse of single‐piece PEEK SBs may not compromise registration accuracy. Apart from this, it has been documented that the machinability of the material utilized in production influences the accuracy of alignment between the SB scan file and the CAD library file [12, 13].

Moreover, SBs can exhibit a variety of geometric configurations including cylindrical, conical, stepped, rectangular, or combinations thereof [14]. It has been documented that SBs with simpler geometric shapes are easier to scan, minimizing the potential scanning errors [15], and leading to a manageable processability of the acquired images [16]. Scientific knowledge on this topic is mostly derived from experimental studies lacking in simulating intraoral conditions. Therefore, optimal design regarding the geometrical features of a SB for accurate scanning remains unclear.

The SB scan file, simultaneously created by the intraoral scanner's algorithm, is matched with the library file in the dental CAD software using a best‐fit algorithm. The amount of surface available for the best fit can be decided and adjusted by a special function in order to improve the overlap. Depending upon the area for best fit alignment, the position of the virtual analogue may change. Beyond this, as intraoral scanning of an SB technically presents imperfections in quality, the overlapping of the generated stl file in comparison to its library CAD stl file would likely be affected.

The current literature lacks information on the scan quality of SBs matched with the corresponding digital libraries. This retrospective cohort study, therefore, aimed to evaluate the scan quality of the SBs matched with their designated libraries in understanding the vulnerability in SB scanning. The null hypothesis of the study posits that the parameters of SB type, edentulous site, and restoration type do not affect the scan quality.

2. Materials and Methods

2.1. Study Design

This retrospective cohort study was designed to assess the surface scan quality of library‐matched SBs for single tooth implant crowns (iC) or implant‐fixed dental prostheses (iFDP) with two implants. The evaluation was categorically performed in monochrome view on post‐processed data using the interface of the scanner (3Shape A/S v22.1 TRIOS3, Copenhagen, Denmark) and conducted by the same researcher.

2.2. Data Collection

In the study, SBs connected to the bone‐level dental implants (Straumann, BLT/BL RC/NC, Straumann Holding AG, Basel, Switzerland) placed in the maxilla and mandible for the treatment of short edentulous spans with iC and iFDPs between January 2017 and January 2021 were considered. The included SBs were digitized following a 2‐stage scanning protocol applying cut‐out and rescan procedure, and matched with the appropriate library in dental CAD software (3Shape Dental System v2020, 3Shape A/S) (Figure 1). Intraoral scans were performed by the senior researcher (KA), while the evaluation of SB scan quality was carried out by the other researcher (FD). Prior to quality assessment, randomly selected SB scans (n = 50) not included in the study were exercised between the researchers for intra‐observer (FD) consistency. For the SBs enrolled in the study, the cases that remained with uncertainty were resolved with inter‐observer discussion.

FIGURE 1.

Scanbody matching in dental CAD software (left), successful alignment using the SB reference area with the difference map (right).

2.3. Data Variables

SB type, edentulous site, and restoration type were the considered independent variables. Two different SBs were assessed: a one‐piece PEEK original SB (CARES RC/NC Mono Scanbody; Institut Straumann AG) and a one‐piece aluminum nonoriginal SB (Scanbody; 3Shape A/S) compatible with the implant (Figure 2). Edentulous site was identified as maxilla and mandible, either with anterior or posterior location, and their treatment was categorized restoration types as iC and iFDP (Figure 3).

FIGURE 2.

Original (left)—and non‐original (right)—scanbodies with similar height and diameter featuring distinct circumferential differences in geometry. The non‐original SB is characterized by horizontal diameter shifting with bevel in height. The upper half presents 2grooves and a flat vertical surface, whereas the lower half is parallel‐sided cylindrical.

FIGURE 3.

Clinical variables: I. Scanbody‐type original(left), non‐original scanbody(right), II.Edentulous‐site: Mandibular posterior(left), maxillary anterior(right) and III. Restoration‐type: Single‐tooth implant crowns(left), short‐span implant‐supported fixed dental prosthesis(right).

2.4. Data Evaluation

To appraise SB scan quality, three categorical domains were defined: SB‐reference area, SB‐scan, and SB‐representation (Figure 4). In the qualitative evaluation, to begin with, the scanning in the SB‐reference region was recorded as either complete or incomplete. In addition, the SB‐scan assessment was carried‐out for two sections of SB: the body and the base. Scanning in the body was recorded as complete or incomplete, whereas the base was described with regard to the merging of the rescan into the first scan sufficiently or not. Furthermore, SB‐representation was evaluated in terms of surface reconstruction utilizing the image stitching algorithm. This comprised geometry and texture evaluation as paired or impaired for the former and smooth or rough for the latter. Definitions for the evaluation of SB scan are described in Table 1.

FIGURE 4.

Scan quality assessment: Reference area (black arrow), SB scan‐body (red lined zone), SB scan‐base (black bean‐shape), SB representation‐texture (white star) and SB representation‐geometry (black circle).

TABLE 1.

The novel chart developed to define scan body scan quality for categorical assessment.

Domain	Interest	Definition	Description
Reference	Area	Complete	No missing surface data
		Incomplete	Hole(s) completed with algorithm
Scan	Body	Complete	No missing surface data
		Incomplete	Hole(s) completed with algorithm
	Base	Sufficient	> 50% integration into cut‐out region without overlapping
		Insufficient	< 50% integration, double surface, tunnel/thorn formation
Representation	Geometry	Paired	Deformation free in visual inspection
		Impaired	Apparent distortion
	Texture	Smooth	Regular surface appearance
		Rough	Degradation in texture due to improper image stitching

2.5. Statistical Analysis

Chi‐squared test was applied to compare the frequencies of the clinical variables, and logistic regression models were constructed to find the factor that affects SB scan quality. The candidate variables for multiple logistic regression models were specified based on univariate logistic regression models with p value ≤ 0.20. With those candidate variables, backward elimination was applied to present the final model. In final models, the odds ratio (OR) with a 95% confidence interval (CI) was represented. Kappa statistics were applied to identify inter‐rater reliability. Statistical analysis was performed using a statistical software program (SPSS Statistics, v23.0; IBM Corp, New York, USA). The significance level was set at p < 0.05.

3. Results

Table 2 presents the descriptive statistics of 243 SBs enrolled in the cohort. The categorical assessments of SB scan quality and Cohen's κ‐coefficients are shown in Table 3. SB scan quality at SBs' three major sections, reference area, body, and base, was exceptionally high 97,5%, 92,6%, and 100%, respectively. Undecided SB representation geometry and texture resolved upon inter‐observer discussion were six and one, respectively.

TABLE 2.

Frequency (percent %) of enrolled scanbodies with regard to edentulous site into the cohort for implant fixed restorations.

Variable	Original SB	Non‐original SB	p
Mandible
Anterior	5 (31.25%)	11 (68.75%)	0.065
Posterior	62 (55.86%)	49 (44.14%)
Maxilla
Anterior	7 (38.89%)	11 (61.115%)	0.178
Posterior	55 (56.12%)	43 (43.88%)

Abbreviation: SB, scanbody.

TABLE 3.

Frequency and percentage of SB scan quality records and Cohen's κ coefficients with regard to the defined categorical domains.

	Domain	Interest	Definition	Frequency	Percent (%)	Kappa
Scanbody	Reference	Area	Complete	237	97,5	1.00
			Incomplete	6	2,5
	Scan	Body	Complete	225	92,6	1.00
			Incomplete	18	7,4
		Base	Sufficient	243	100	1.00
			Insufficient	0	0
	Representation	Geometry	Paired	204	84	0.50
			Impaired	39	16
		Texture	Smooth	135	55,6	0.78
			Rough	108	44,4

The logistic regression analysis outcomes are displayed at Table 4. The generated models revealed that SB type significantly affects the SB scan quality in the SB representation domain for both geometry (< 0.001) and texture (< 0.001), and the relationship between jaw and SB representation in geometry and texture is significant as well (0.003 and < 0.001, respectively). Restoration type, crown versus FDP, was significantly effective on scan quality at SB's scan at the body region (0.004).

TABLE 4.

Generated logistic regression models to study the effect of clinical factors on SB scan quality.

Response	Variable	Univariate logistic regression		Multiple logistic regression
		OR (95% CI)	p	OR (95% CI)	p
SB reference‐area	Type (original/non‐original)	4.556 (0.524–39.593)	0.169	—	—
	Maxilla/mandibula	2.232 (0.401–12.421)	0.359	—	—
	Anterior/posterior	1.236 (0.14–10.919)	0.849	—	—
	Crown/FDP	0.577 (0.104–3.212)	0.53	—	—
SB scan‐body	Type (original/non‐original)	1.113 (0.424–2.925)	0.827	—	—
	Maxilla/mandibula	0.677 (0.253–1.81)	0.437	—	—
	Anterior/posterior	1.252 (0.342–4.575)	0.734	—	—
	Crown/FDP	6.598 (1.858–23.432)	0.004	6.598 (1.858–23.432)	0.004
SB representation‐geometry	Type (original/non‐original)	4.191 (1.838–9.558)	0.001	4.462 (1.928–10.327)	< 0.001
	Maxilla/mandibula	0.319 (0.148–0.688)	0.004	0.296 (0.134–0.653)	0.003
	Anterior/posterior	0.663 (0.22–2.001)	0.466	—	—
	Crown/FDP	0.887 (0.445–1.768)	0.732	—	—
SB representation‐texture	Type (original/non‐original)	5.437 (3.105–9.522)	< 0.001	6.663 (3.617–12.276)	< 0.001
	Maxilla/mandibula	0.32 (0.188–0.543)	< 0.001	0.245 (0.134–0.449)	< 0.001
	Anterior/posterior	0.552 (0.256–1.191)	0.13	—	—
	Crown/FDP	1.74 (1.043–2.902)	0.034	—	—

Abbreviations: CI, confidence internal; OR, odds ratio.

In evaluation of SB‐type variable, cross tabulation presented that 79.5% and 75% of the impairment in geometry and roughness in texture, respectively, is related to original SBs' scan quality. Further comprehension along with the logistic regression model is that the risk of impaired geometry for original SBs is 4.462 times higher than for nonoriginal SBs, and the occurrence of surface roughness is 6.663 times higher as well (Table 4).

Regarding the edentulous‐site variable, SB scan quality is being negatively affected in the mandible than in the maxilla by 3.378 (1/0.296) and 4.081 (1/0.245) times for representation in geometry and texture, respectively (Table 4).

Taking the restoration‐type variable into consideration, 83.3% of incompleteness of SB scanning in the body region for iCs compared to 16.7% for iFDPs is noticeable with cross tabulation. This remained statistically significant in the logistic regression model, with the odds ratio indicating that the risk of incompleteness for iCs was 6.598 times higher compared with iFDPs (Table 4).

4. Discussion

In this retrospective cohort study, the effect of SB type, edentulous site, and restoration type on scan quality of SB‐reference area, SB‐scan, and SB‐representation was categorically evaluated. As the outcomes revealed that at least one domain in the assessment of SB scan quality was significantly affected by one of the variables, the null hypothesis was rejected.

The eligible SBs for scan quality assessment were scanned with a staged approach in which implant digitalization was finalized following implant region cut‐out from the initial arch scan, and then rescan to merge SB scan into the arch scan. Outcomes of the studies on scanning accuracy for cut‐out and rescan protocol are inconclusive [17, 18]. In the current study, all scan quality evaluations for SBs' base region presented sufficient integration between the two scans for the SB type variable. Therefore, we may imply that SBs differing in material, original SB with peek versus non‐original SB with aluminum, are predictably scannable in terms of the SB base to follow the 2‐stage scan protocol.

SB representation was the most affected domain in scan quality evaluation as roughed texture and impaired geometry, which were recorded at 44.4% and 16%, respectively. In essence, it is not surprising due to challenges, particularly for the image‐stitching algorithm in intraoral scanning of geometric and monochromatic surfaces. With this, one may assume SB representation would be poorer in cases with extended edentulous spans, literally presenting reduced commonalities for the image‐processing algorithm.

It has been presented that precision had been significantly affected by the SB design and geometry under extraoral scanning conditions [19]. Moreover, a recent systematic review based on in vitro studies reported that SB geometry affected scanning accuracy [20]. In this cohort, impaired geometry in SB representation was significantly different for the SB type variable and roughed texture as well. Accordingly, various geometrical features in the non‐original SB were helpful for deformation‐free SB scan in comparison to the original SB, which is limited to cylindrical shape with beveled reference scan region (Figure 2). Authors believe that the irregularities, even geometrical, facilitate image stitching in the algorithm with the scanner using optical sectioning technology.

Park and Choi [5] addressed deficiencies in SB scanning in an in vitro study. The SB scan data was modified to simulate incomplete scanning prior to surface alignment with the CAD library. They concluded that based on the created defect size, the implant positioning accuracy in the dental CAD software decreases. Finally, they posited that the restorations fabricated upon that data would remain beyond clinically acceptable limits. In the current study, SB scan quality was categorically evaluated, and incomplete reference and scan‐body scanning clinically reduced the scan quality of the SBs by 9.9%. Furthermore, the outcomes clearly demonstrated that SB scan quality would be dramatically decreased when the challenges in representation for deformation‐free SB scanning are considered. Therefore, simply matching the SB scan with its library CAD file may not assure predictable restorative outcomes. As the focus on the target for the current study was to understand how clinical variables interact with SB scan quality, further studies are needed to explore the interplay between SC scan quality and restorative outcomes.

The effect of the restoration type variable on scan quality categorically for SB scan at body was remarkable with the studied regression model. It was predicted that incompleteness in SB scan at body is 6.598 times more prevalent for iCs compared with iFDPs. This outcome may be explained by the limitation of the scanner's field of view due to the proximity of SB to the adjacent teeth for iC restorations. Having significantly favorable outcomes at SB representation for non‐original SB, we recommend the use of a SB with geometrical features for single‐tooth spaces in clinical practice.

It is essential to acknowledge that intraoral scanners have the potential to introduce scanning errors in the stitching of images due to a lack of stable reference points [21]. This fact can explain why the mandible demonstrated a significantly higher risk for roughed texture at SB representation. This challenge may become even more complicated in cases with free‐end edentulism [22, 23]. Remembering that the scanning of non‐anatomical surfaces couples the defined difficulties, it is recommended to use appropriate tools to move away movable tissues such as the buccal and lingual areas from the teeth and to keep the scanning field as dry as possible. For such challenging cases, the use of SBs with geometrical features may lower the probability of reduced scan quality specifically for SB representation.

The non‐homogeneity in the number of SBs between anterior and posterior edentulous sites represents a potential limitation of this study. Another limitation may be addressed as information lacking in stability of SB to implant connection with regard to multiple use of SBs. Furthermore, additional information on how scan quality affects the restorative outcomes would be invaluable for correct clinical interpretations.

5. Conclusion

Scanbody‐, restoration‐type, and edentulous site are the clinical variables presenting significant influence on SB scan quality in which SB representation is the most affected. It may be suggested to prefer an SB with distinct geometrical variations for an implant when there are options to choose from.

Author Contributions

K.A. conceived the ideas, administered the project, wrote the original draft, reviewed, and edited the manuscript, approval of article. F.D.B. collected the data, wrote the draft, reviewed, and edited the manuscript. K.A. and F.D.B. analyzed and interpreted the data.

Ethics Statement

This study was approved by the NonInterventional Clinical Research Ethics Committee of Hacettepe University (Protocol Code: GO 21/848).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data supporting the findings of this study are not currently available anywhere.