INTRODUCTION

Study models provide a three-dimensional replica of malocclusion during any stage of treatment, as well at the final outcome. Despite all associated benefits, these study models have some disadvantages in terms of storage, durability, and transferability.

Many attempts have been made to replace plaster study models. In the mid 1990s, three-dimensional (3D) scanning technology was introduced, and study models1–7 were transformed into a digital format. Since then, software technology has refined this approach, and digital orthodontic models have become commercially available.8–14 However, this technology is associated with other problems, such as the scarcity of digital model supplier companies, the time required to ship impressions to those companies, and questions on the accuracy of the final digital model.

Currently, five companies worldwide are producing digital models commercially. Three of these companies are in the United States, one is in The Netherlands, and one is in Poland. These companies accept high-quality alginate impressions with 100+ hours' dimensional stability and disposable impression trays.

Besides the digital model supplier companies, some software companies, such as 3Shape (3Shape A/S, Copenhagen, Denmark; scans only stone models), Laserdenta (Laserdenta AG, Basel, Switzerland; scans both stone models and impressions), and INUS Dental Scanning Solution (INUS Technology, Inc, Seoul, Korea), supply 3D model scanners and orthodontic software for individual clinical practice.

As instructed by the digital model supplier companies, alginate impressions are delivered in 1 day, regardless of the location of their origin; in case of any delay, the impressions retain dimensional stability for up to 100 hours, or even up to 10 days. However, the International Standard (IS) for alginate impression materials (ISO 1563: 1990E) contains no specification for dimensional stability and thus places no requirement on manufacturers to state dimensional stability properties on their labels.

Dimensional stability was defined by Nicholls15 as “the ability (of a material) to maintain accuracy over time.” Further, this paper defines the result of loss of accuracy as “distortion,” and “the relative movement of a single point, or group of points, away from some originally specified reference position such that permanent deformation is apparent.” Dental alginates, similar to all hydrocolloids, tend to distort over time as they lose (by evaporation and syneresis) or gain (by imbibition) water, thereby contracting or expanding. Even when stored under conditions of 100% humidity, an alginate impression will contract, indicating that processes other than dehydration, including polymerization and syneresis, are involved.15–19

Therefore, the best results are obtained when dental alginate impressions are poured after 10 minutes, to avoid distortion from initial expansion and elastic deformation, and before 1 hour, to avoid distortion from alginate contraction or expansion due to water movement and syneresis.2021

Another question is the accuracy of digital modeling. Several studies have shown that linear measurements on digital models and plaster models have significant differences, but the magnitude of these differences does not appear to be clinically relevant. The overall conclusion of these studies is that digital models are acceptable alternatives to stone casts for the routine measurements used in orthodontic practice.22–34 On the other hand, no study has evaluated the possible time-dependent deformation of alginate impressions during their shipment to the companies involved.

The two aims of this study are as follows:

• To evaluate the accuracy of digital models produced by the 3Shape system

• To test the dimensional stability of three different brands of alginates for durations of 1, 2, 3, and 4 days in a laboratory environment that simulates shipping conditions

MATERIALS AND METHODS

In this study, a maxillary dental model (D85SDP-200; Kilgore International Inc, Coldwater, Mich) was used as a master model. Five artificial reference points were prepared on the master model to increase the reproducibility of measurements with the use of a conical bur on the following:

• Right and left first molar mesiobuccal cusp tips

• Right and left canine cusp tips

• Left second premolar buccal cusp tip

Three different brands of alginates—(1) Cavex (Holland BV, Haarlem, The Netherlands), (2) Orthoprint (Zhermack Spa, Badia Polesine [RO], Italy), and (3) Tropicalgin (Zhermack)—were used to produce stone models of the master model. Thirty-five impressions with each alginate were taken from the master model, constituting a sum of 105 impressions. All impressions were taken by the same researcher with the use of plastic trays of the same size (DuraLock Plus Impression Trays 501-005U; Ortho Technology, Tampa, Fla). Tray adhesives were not used.

Alginates were mixed manually, with strict adherence to manufacturers' instructions. Impressions were not rinsed with water or immersed in any disinfecting solutions. The 105 impressions were divided into five equal groups, as follows:

• Immediate group: 7 Cavex + 7 Orthoprint + 7 Tropicalgin = 21 impressions

• First day group: 7 Cavex + 7 Orthoprint + 7 Tropicalgin = 21 impressions

• Second day group: 7 Cavex + 7 Orthoprint + 7 Tropicalgin = 21 impressions

• Third day group: 7 Cavex + 7 Orthoprint + 7 Tropicalgin = 21 impressions

• Fourth day group: 7 Cavex + 7 Orthoprint + 7 Tropicalgin = 21 impressions

The impressions of the immediate group were poured into stone within 1 hour with the use of Fujirock EP type 4 dental stone (GC, Tokyo, Japan) mixed in a vacuum mixer (Twister; Renfert GmbH, Hilzingen, Germany) and were allowed to set, with the tray not inverted. The impressions of the first, second, third, and fourth day groups were placed in sealed plastic bags in accordance with digital model companies' instructions. The brand of alginate, the number of the model, and the day and exact time the impressions were taken were recorded on the sealed bags. The sealed impression bags of all groups were stored in a dark room at standard room temperature (23 ± 2°C [ISO 4823: 1992 E]). The impressions of the first day were poured after 24 hours, the second day after 48 hours, the third day after 72 hours, and the fourth day after 96 hours. At the end of four different storage times, 105 stone models had been obtained.

Measurement Parameters

Seven points (five artificial and two anatomic) and five linear measurements were used to analyze the master, stone, and digital models. The measurement points and the linear measurements are shown in Figure 1. The sum of linear measurements was calculated to create an overall linear measurement for each model.

View Figure

• Download as Powerpoint
• Download Figure

The master model, 21 stone models of the immediate group, and 21 digital models were analyzed by three different researchers to achieve a gold standard, whereas the remaining 84 stone models of the first, second, third, and fourth days were analyzed by two different researchers. Each measurement was repeated three times. Digital models were measured with the measurement tool provided by 3Shape Orthoanalyzer software to the nearest 0.01 mm (10 μm). The master model and the stone models were measured with an electronic caliper (Masel, Bristol, Pa) to an accuracy of 0.01 mm (10 μm) (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

The reliability of the observers revealed high intraclass correlation coefficients in master model measurements, measurements on 21 stone models of the immediate group, and measurements on 21 digital models (Table 1).

Table 1.  Interexaminer Reliability

View Fullscreen

The correlation coefficient for the measurement of the master model and the 21 stone models of the immediate group was 0.975. The absolute difference value between the master model and the 21 stone models was 0.152 ± 0.125 mm, and the absolute percentage was 1.285% ± 1.529% for the overall measurements (Table 2).

Table 2.  Comparison Between Measurements Made on 21 Stone Models and the Master Model

View Fullscreen

The correlation coefficient for the measurement of the master model and the 21 digital models was 0.977. The absolute difference value between the master model and the 21 digital models was 0.095 ± 0.053 mm, and the absolute percentage was 0.695% ± 0.637% for the overall measurements (Table 3).

Table 3.  Comparison Between Measurements Made on Digital Models and the Master Model

View Fullscreen

The correlation coefficient for the measurement of the 21 digital models and the 21 stone models of the immediate group showed a high level of reliability with the value of 0.989 (Table 4). The absolute difference value between the 21 stone models of the immediate group and the 21 digital models was 0.199 ± 0.164 mm, and the absolute percentage was 0.73% ± 0.3% for the overall measurements (Table 4).

Table 4.  Comparison Between Measurements Made on Digital Models and Stone Models of the Immediate Group

View Fullscreen

The differences in overall measurements of the stone models poured from three different alginates immediately and after 1, 2, 3, and 4 days of storage time are shown in Table 5 and Figure 3.

Table 5.  Changes in Overall Measurements of Stone Models Poured into Three Different Brands of Alginates in Five Different Storage Periods

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

A digital model is essentially a replica of a replica. The errors and the problems may be doubled because the digital model process has two phases. The first phase involves taking an impression and pouring a model (mouth-to-stone model phase), and the second phase consists of scanning a stone model (stone-to-digital model phase). In addition to evaluating the accuracy of 3D scanning alone, the accuracy of impressions and different alginate behaviors should be evaluated throughout the digital model process. This study sought to investigate the complete process, from the impression taking stage to the stage of 3D model analyses.

To determine the observers' reliability, three sets of measurements on the master model, 21 stone models of the immediate group, and 21 digital models were compared with the use of the intraclass correlation coefficient, the absolute difference, and the absolute percentage. Results showed a high level of interobserver reliability, thus setting up a gold standard (Table 1).

Tomassetti et al22 stated that distinguishing the proper landmarks was difficult with the OrthoCAD (Cadent Inc, Carlstadt, NJ) and Quick Ceph Systems (Quick Ceph Systems Inc, San Diego, Calif). Moreover, in his review, Houston35 points out that the greatest source of random error involves difficulty in identifying a particular landmark. This leads to the issue of reproducibility of the computerized analysis.

Hence, in this study, investigators used a master model to represent the maxillary dental arch, plastic impression trays of the same size, and one brand of dental stone, to lessen the variables; investigators also prepared artificial reference points on teeth to enhance the reproducibility of measurements. Both the measurements made on these reference points (RM-LM, RC-LC, LM-LP) and the measurements made on anatomic points (RM-RMG, RC-RCG) showed high degrees of interobserver correlation, revealing that researchers were fairly accurate in measuring anatomic points as well (Table 1). The artificial reference points did not have an additive effect on accuracy in this study. These results may be due to the high quality of the models (both digital and stone) and the care and skill of the observers, which are important factors, as mentioned in Houston's study.35

Deformation of alginates occurs in multiple linear directions. However, the aim of this study was to show the overall effect of time-dependent deformation of alginate impressions on all linear parameters in digital modeling, rather than clarifying its effects on each parameter separately. Thus, although the statistical results of each linear parameter are shown on each table, together with the overall parameter, the present discussion is founded on its effects on the overall parameter rather than on its effects on each linear parameter separately.

A high correlation between the master model and the stone models of the immediate group shows that the traditional approach to the “mouth-to-stone model phase” is a reliable procedure. The absolute difference in the overall measurements is 0.152 mm, and the mean absolute percentage error is 1.285%. This leads to a 0.11 mm error on the mesiodistal measurement of an 8 mm wide tooth. These values can be accepted as within clinical tolerance in orthodontic practice (Table 2).

A high correlation was also noted between the measurements of 21 stone models of the immediate group and 21 digital models (stone-to-digital model phase). The absolute difference value was 0.199 mm, and the absolute percentage was 0.011% for the overall measurements (Table 4). This result shows that the scanning procedure is much more reliable than the taking of alginate impressions. However, one should still take an impression and pour into the stone model to obtain a digital model.

The differences between measurements of the master model and the digital models were also evaluated and highly correlated. The absolute difference between the digital models and the master model was as small as 0.095 mm, and the mean absolute percentage was 0.695%. It is interesting to note that these results showed that digital models, which we call “the replica of the replica,” were more similar to the master model than were the stone models. This may be due to the evaluation of linear changes in terms of absolute values (Table 3).

The effects of the time-dependent deformation of alginates on digital model accuracy were evaluated throughout the measurements on stone models of all groups because the stone models of the immediate group were highly correlated with their digital replicas. Depending on the measurements of stone models poured from the three brands of alginates, investigators learned that significant deformations occurred during 4 days of storage (Table 5 and Figure 3). These three alginates also showed significant differences among one other at the first, third, and fourth days but not on the second day in terms of overall linear measurements. However, regarding the overall values in Table 5, these differences are very small in terms of millimeters, which can be accepted in clinical tolerance and in orthodontic analyses.

As a result, storing these three brands of alginate impressions in sealed plastic bags up to 4 days had no adverse effect on digital modeling. One can expect that developments in 3D scanning technology will continue to occur until direct 3D scanning of the full dentition becomes a clinically practicable method for orthodontists.

CONCLUSIONS

• Digital orthodontic models are as reliable as traditional stone models and probably will become the standard for orthodontic clinical use.

• Storing alginate impressions in sealed plastic bags for up to 4 days had no adverse effects on the digital model.

• This new technology is still alginate dependent.