Measurement of the accuracy of dental working casts using a coordinate measuring machine.

Background/Aim
Dental impressions present a negative imprint of intraoral tissues of a patient which is, by pouring in gypsum, transferred extraorally on the working cast. Casting an accurate and precise working cast presents the first and very important step, since each of the following stages contributes to the overall error of the production process, which can lead to inadequately fitting dental restorations. The aim of this study was to promote and test a new model and technique for in vitro evaluation of the dental impression accuracy, as well as to asses the dimensional stability of impression material depending on the material bulk, and its effect on the accuracy of working casts.


Methods
Impressions were made by the monophasic technique using the experimental master model. Custom trays with spacing of 1, 2 and 3 mm were constructed by rapid prototyping. The overall of 10 impressions were made with each custom tray. Working casts were made with gypsum type IV. Measurement of working casts was done 24 h later using a co-ordinate measuring machine.


Results
The obtained results show that the working casts of all the three custom trays were in most cases significantly different in the transversal and sagittal planes in relation to the master model. The height of abutments was mainly unaffected. The degree of convergence showed certain significance in all the three custom trays, most pronounced in the tray with 3 mm spacing.


Conclusion
The impression material bulk of 1–3 mm could provide accurate working casts when using the monophasic impression technique. The increase of the distance between abutment teeth influences the accuracy of working casts depending on the material bulk. [Projekat Ministarstva nauke Republike Srbije, br. TR 35020: Research and development of modelling methods and approaches in manufacturing of dental recoveries with the application of modern technologies and computer aided systems]


Introduction
Dental impressions present a negative imprint of intraoral tissues of a patient which is, by pouring in gypsum, transferred extraorally on the working cast.From the laboratory perspective of dental restorations manufacturing, casting an accurate and precise working cast presents the first and very important step, since each of the following stages contribute to the overall error of the production process and can lead to inadequately fitting of dental restorations 1 .
Making an impression is a clinically challenging procedure which is influenced by numerous factors of the oral environment, as well as the properties of the material itself.Examining the factors that influence the accuracy of dental impressions can be conducted in two ways, by direct mesurement of the impression, or by measurement of the appropriate working cast.Both methods have their advantages and limitations.The first method provides direct data about the condition of the impression material, thus avoiding superimposition of further errors.Limitations are related to the use of contactless measurement, which is affected by a small measurement field, software processing and adequate optical characteristics of the impression material [2][3][4] .The second method, measurement of the working casts, provides a wider range of possibilities regarding the measurement technique.Both contact and contacless measurement can be used, with ease of access for manipulation with measurement object 5 .Superimposition of errors, when casting a gypsum model, can be minimised by fixing the experimental conditions with equal casting protocols for all of the investigated dental impressions.Although making of the gypsum working cast prolongs the time and effort needed to obtain necessary data, it is a reference base for manufacturing of dental restorations and as such, provides better insight when assesing the discrepancies of future dental restorations.
Determination of accuracy of dental impressions requires a complex model that can replicate in vivo conditions of making of dental impressions, with the accuracy and precision of in vitro investigation.Earlier studies included a variety of models that consisted of custom blocks, cylinders, single or several abutment teeth and complete edentulous jaws of various materials [5][6][7][8][9][10][11] .Regarding this, there is the need for a reliable experimental model, which combined with a corresponding measurement instrument, could overcome the difficulties of data interpolation between laboratory conditions and clinical practice.The two should complement each other, forming a complex model for data acqusition.From this point of view, the use of a coordinate measuring machine (CMM) was chosen due to the possibility of the accurate and precise three-dimensional (3D) measurement 12,13 .
Previous studies that investigated the influence of material bulk on the accuracy of working casts were conducted using a measurement microscope and included only two dimensional measurements [14][15][16] .Also, the models used consisted of one to several abutment teeth, while as to our knowledge, the influence of material bulk on partially edentulous dental arch model has not been made [14][15][16][17][18] .A more de-tailed analysis requires a model more similar to the intraoral conditions, since the accuracy of impressions depends on the material bulk which changes with spacing between the remaining teeth 19 .The possibility of an independent 3D analysis by CMM for each of the segments of the working casts was considered to be an improvement compared to the previous studies.
Additionally to the previous studies, we aimed to incorporate another use of rapid prototyping (RP) technology in the field of dentistry.Making impressions with RP made custom trays is a new attempt to investigate the use of computer-aided design/computer-aided manufacturing (CAD/CAM) technology into a broader perspective of dental practice.The rapid prototyping technique was chosen with the aim of avoiding possible distortions of the material that can occur in standard acrylic custom trays, and providing a stable base for the impression material 20 .Thus, the aim of this study was to promote and test a new model and technique for in vitro evaluation of the dental impression accuracy, as well as to asses dimensional stability of elastomeric impression material in reference to the material bulk, by examining working casts with CMM.

Measurement of the experimental master model and construction of custom trays
For the purpose of this study, an experimental metal master model which consisted of six abutment teeth was constructed.It presented the upper jaw with two central incisors, canines and first molars (Figure 1).Dimensions of the teeth were taken from literature and reduced by the amount of tooth substance expected to be removed with grinding 21 .The taper was set at 6°.The master model consisted of an assembly with machined components.Geometry of the master model and generation of computer numerical control machines (CNC) code was designed by the CAD/CAM system. the resulting mean values of parameters of measurement were calculated.Nominal dimensions of the CAD model were corrected using mean values and were subsequently used to create a custom tray model as a negative of the previous CAD model.Physical models of custom trays with 1, 2 and 3 mm spacings for the impression material, were constructed by rapid prototyping (Z310 plus, 3D Systems, USA).The powder used was gypsum based (zp 131), with a binder (zb 60) and two component epoxy resin as the filler (s5000), presented at Figures 2 and 3.

Impression technique
Impressions were made using a monophasic technique with silicone addition.Due to the micromechanical retention of the impression material to a custom tray, which is the result of the successive layering technique of rapid prototyping, the tray adhesive was not used.The impression material was automixed (Elite Hd + regular, Zhermack, Italy), working time was set to 60 sec.Making of impressions was conducted at room temperature (23°C), retention of the tray was done with the weight of 1 kg.Setting time was set at 10 min, overall of 10 impressions were made with each custom tray.Due to the expected viscoelastic recovery of the impression material, pouring of gypsum was delayed by 30 min.Working casts were made with the gypsum type 4 (Elite rock, Zhermack, Italy) and were allowed to set for 1 h before the impression was removed.Measurement of the working casts was done 24 h later on CMM.

Measurement of working casts on a coordinate measuring machine
Measurement of working casts was conducted under the same conditions as those of the master model.Inspection was conducted conformant to the new generation of product geometry specification (GPS) 22 .This was based on the fact that the model consisted of geometric primitives (cones, cylinders and planes).Each geometric primitive was measured in a finite number of discrete points randomly arranged on the primitive surface.The measurement strategy of the cone (abutment) contained 100 measurement points, 50 points for measurement of cross-section plane between the cone and the cylinder (abutment and chamfer) and 50 for measurement of the cylinder (chamfer).The output of this measurement included the coordinates of all measurement points which were subsequently used to generate associative geometry of the primitives.Furthermore, the software analysis was used to determine all geometric characteristics (size, angle, form, orientation, location) according to the specification requirements.
Parameters, X 1-3 and Y 1-6 , represented the axial distances which were derived features from the cone.The axial distances were observed on the cross-sectional plane which passed through the base of the abutments.Parameters denoted as Z 1-6 , were determined as the distance between two planes which limited the abutment vertically, while α 1-6 represented the degree of convergence and was directly derived from the abutment measurement (Figure 4).
The presence of measurement uncertainty was disregarded due to the fact that all working casts were measured in  identical laboratory conditions: 24 h after casting, using the same position on machine table, by the same operator and using identical inspection strategies and stylus 23,24 .

Statistical analysis
Statistical analysis of the results obtained from the measurement of the working casts was conducted by the twosample t-test, which determined whether there was a statistically significant difference between the designated distances of working casts compared to the master model.One-way analysis of variance (ANOVA) was also conducted to establish if there were any significant differences in the results obtained by the type of custom tray for the considered parameters.

Results
The results of the two-sample t-test, with a 95% confidence interval, show that the working casts of all three custom trays were in most cases significantly different in transversal and sagittal plane in relation to the experimental master model (X 1-3 , Y 1-6 ).The height of the abutments (Z 1-6 ) was mainly unaffected, while the degree of convergence (α 1-6 ) increased, especially in the working casts made by custom trays with 3 mm spacing.The results are presented in Table 1.
Most notable deviations for the first tray (T 1 ) were mea-sured in X 3 (20 µm) for the transversal plane, Y 2 (27 µm) for the sagittal plane, Z5 (-13 µm) for the height of the abutments and α1 (-0.09°) for the convergence of axial surface.The dimensions of Z 1 -Z 4 , Z 6 and α 3 , α 4 and α 6 showed no statistical significance in relation to the experimental master model.Most pronounced deviations for the second tray (T 2 ) were measured in X 3 (22 µm) for the transversal plane, Y2 (27 µm) for the sagittal plane, Z5 (-15 µm) for the height of the abutments and α1 (0.11°) for the convergence of axial surface.The dimensions of Y 1 , Y 5 , Z 1 -Z 4 and α 6 showed no statistical significance in relation to the experimental master model.
The third tray (T 3 ) showed largest deviations in X 3 (36 µm) for the transversal plane, Y 2 (24 µm) for the sagittal plane, Z 5 (-7 µm) for the height of the abutments and α1 (-0.18°) for the convergence of axial surface.Dimensions of Y 1 , Y 5 , Y 6 , Z 1 -Z 6 showed no statistical significance in relation to the experimental master model.
Analysis of variance with a 95% confidence interval showed that the type of tray did not have any significant influence (p > 0.05) on the parameters of measurement, except for the X 3 and α1-6.(p < 0.05) Therefore, with respect to α1-6 and X 3 , the third tray yielded the results which were significantly different from the first two trays.The presentation and graphical comparison of the results are shown in Figures 5 and 6.

Discussion
The results of this study show that both the experimental master model and the technique used for making the dental impressions have proven to be a reliable method for assessment of the accuracy of dental impression materials.Working casts obtained from all the three custom trays were, in most cases, significantly different from the experimental master model.Most evident deviations were recorded in the sagittal (Y 2 ) and transversal plane (X 3 ) for all of the working casts.These were also the greatest distances among the measured abutments, so the differences were most likely to be observed.The measurement of all of the working casts was done by the principles of 3D measurement.Each surface of the abutments was measured with 50-100 surface points.The coordinates of surface points served as a base for measuring the geometry of the abutments.Due to a large number of data and easier analysis of measurement, the results were presented as a cross-section of the model in one plane, which in this case passed through the base of the abutments.This is why the results are presented as the classical 2D measurement, but the principles of measurement are related to the entire surface of the measured model and additional data can be withdrawn for each additional crosssection 25 .
The impression material used in this study had a defined polymerization contraction of ≤ 0.2%, while expansion of gypsum was claimed to be 0.19% by the manufacturer.Expansion of gypsum would thus compensate for polymerization contraction of the impression material, but it should be noted that these values refer to the two sizes that are not equal, ie the thickness of the impression material in accordance with the height and width of the gypsum abutments 26 .Thickness of the impression material varied from 1 to 3 mm from each side of the abutment, while the height and width of the abutments varied between 4.5 and 7.5 mm.In relation to this, expansion of gypsum should be a more dominant factor, which was proven by this study.Additionally, the influence of the micromechanical retention of the impression material to a custom tray should also be considered.In this case, the expansion of gypsum will be superimposed with the impression materials contraction towards the tray's walls, so that the combined effect will be presented by even wider abutments.As for the height of the abutments, where the upper surface was smaller than the side surface of the custom tray, in most cases the height of the abutments was smaller than that of the master model (Z 1-6 ).This can also be explained by impression materials contraction towards the walls.As the side surface of the tray had a bigger contact area than the upper one, the impression material was pulled down, which resulted in shorter abutments.These results are in accordance with other studies where tray adhesive was used in their research protocols [27][28][29] .
The previously described process indicates that the silicone addition had a good retention with a custom tray made of epoxy resin by the RP technique.Although the RP technique is not widely used in the field of dentistry, it shows promise for manufacturing of dental devices.Its use in everyday practice, as a rational method for fabrication of custom trays, requires advanced systems for intra-or extraoral scanning, together with an adequate software connected to an RP machine, which is currently a limiting factor.Additive technologies such as RP are mainly being used for the production of dental copings for fixed dental restoration, fabrication of surgical guides in implant dentistry and in reconstructive maxillofacial surgery 30,31 .As there is a large variety of materials that can be used for the production of custom trays with RP technology in different working regimes, these results are just a starting point and require additional investigation.
The analysis of convergence of axial surfaces showed that the largest deviations were observed in the third custom tray (α 1 = -0.18°),while deviations of the second tray (α 1, 2 = -0.11°)and the first tray (α 1 = -0.09°)were smaller.The angle of convergence was measured by scanning the axial surface and further software processing.Although small, deviations detected were oriented towards the increase of convergence.Convergence of the axial wall of the master model abutments was set to 6° (total of 12° when observing both axial planes), while gypsum abutments were detected to have more of a taper.The increase of the convergence angle reduces the overall surface of the abutment, which can influence the retention of the dental restoration 32 .While the width of the abutments was wider in the base cross-section, the axial surface has proven to have more taper.This could be explained by the increase of thickness of the impression material towards the upper base of the abutment and lower retention to the impression tray.Also, during removal of the impression the material deforms elastically.The resulting effect will be more pronounced in the axial surface, because of the overall increase of contact surface and direction of removal force away from the base of the abutment.This may all together explain the behaviour of the material to contract towards the walls of the tray in the base area, while slightly contracting inwards as the height progresses.
The results of this study show that all three custom trays performed satisfactory, but slightly better results were obtained with custom trays with 1 or 2 mm spacing.This is contrary to the claim that the thickness of the impression material should be at least 3 mm to prevent distortion of the material 33 .Due to a problem of correct positioning of the impression tray when thickness of the material is low, the custom tray with 2 mm spacing should be recommended for use in the clinical practice.This is considered to be beneficial regarding the accuracy of the impression, ease of handling and reduction of the quantity of material.Making a precise impression is especially important when constructing long-span bridges, because the accuracy of fit is harder to achieve as the number of abutment teeth and the distance between them increases.In this case, custom trays should be recommended, because they provide uniform thickness of impression material that can influence the accuracy of working casts.The issue of impression material bulk and its effect on accuracy of working casts has been previously adressed.Plausible results were obtained up to 5 mm of the impression material 14-18, 34, 35  Difficulty of comparison of our own results with other studies, lies in the complexity of the methodological procedure.Even comparison with studies that used the monophasic impression technique, as it was in this study, should be observed not only by input, but also by output parameters, which are obtained through the use of the measurement instrument.As most of the studies used different measurement procedures, a relevant overview is hard to achieve, so we will restrain from further data comparison.
Limitations of this study are related primarily to the method of measurement.Although CMM is an instrument of great accuracy and precision, it is necessary to know the geometry of the measurement object in order to obtain reliable results 3 .Because the measurement is done by a relatively small amount of surface points (in this study up to 100), the measurement object has to be clearly defined before the measurement takes place.Since the master model of known dimensions and shape was used, measuring in this study is characterized by considerable accuracy and precision.
Based on the obtained data, future studies would be based on determing of accuracy of the rest of the production process of dental restorations, including production of wax patterns and the casting procedure.Detection of flaws in each of the stages of the production procedure, will provide useful guidelines for the dentists and dental technicians for the improvement of the quality of dental restorations.

Conclusion
The results of this study show that the impression material bulk of 1-3 mm could provide accurate working casts when using a monophasic impression technique.The increase of the distance between abutment teeth influences the accuracy of working casts.Custom trays produced by rapid prototyping can be successfully used for dental impressions.The experimental master model, combined with a CMM, is a reliable tool for assessment of dental impression accuracy.

Fig. 1 -
Fig. 1 -The experimental master model.Measurement of the experimental master model was performed with the CMM (Contura G2, Carl Zeiss, Germany) equipped with a contact probe.Maximum permissible error for size measurement (MPE E ) of this CMM is 1.9 + L/330 µm.The master model was measured five times and

Table 1 Results of the measurement
.