Metrological Protocol for Comparison

of digital and analogic Articulators for complete Dentures

Mattia Maltauro1*[0000-0002-8339-9306], Lorenzo Menarini2[0000-0001-8428-1919],

Roberto Meneghello1[0000-0002-8099-9795], and Leonardo Ciocca2[0000-0002-2127-484X]
1 Department of Management and Engineering, University of Padova, Stradella San Nicola 3,
36100 Vicenza Italy
2 Department of Biomedical and Neuromotor Sciences, Alma Mater Studiorum University of

Bologna, Via S. Vitale 59, 40125 Bologna Italy

mattia.maltauro@phd.unipd.it

Abstract. This paper proposes a methodology to compare the trajectories from

different articulators - both physical and digital - during lateral and protusive
movements. In the case of digital articulators, the articulated models are digitally
moved and exported in position; in the case of mechanical articulators, the mod-
els are locked into position and 3D scanned. The digital models in position, both
digital and scanned, are aligned to a common reference system and a maxilla-
based reference system is tracked. The trajectories are defined as interpolating
splines through the maxilla-based reference system origins. A Gerber mechanical
articulator and an “Artex CR adjustable” virtual articulator were compared. The
repeatability of the mechanical trajectory is found to be less than 184 microns.
The resting position of the two articulators is found significantly different mean-
ing that a bias is introduced by the operator in the analogic protocol. The trajec-
tories have significantly different shapes as expected coming from two different
articulator models. The proposed methodology proved to be a valid means to
compare different articulators.

Keywords: Complete Dentures, Articulators, Trajectory Reconstruction, 3D

Scanners.

1 Introduction

During the last years, CAD-CAM procedures have been introduced in many fields of
dentistry, in particular, digital workflows are now used to produce also complete den-
tures with remarkable advantages: increased biomechanical properties, reduced work-
ing times and costs, and fewer appointments at the office [1–4].
Among other digital design tools, digital articulators were introduced aiming to pro-
vide an alternative to the current golden standard analogic (mechanical) articulators [5].
A widely used mechanical articulator for the design of complete dentures is the Gerber
semi-adjustable articulator, for which a set of specific intermaxillary occlusal registra-
tions, measured from the patient, can be set to approximate a correct occlusion [6].
2

The use of digital articulators is still not completely independent, and some analogic
clinical steps are necessary in digital workflows and cannot be substituted by digital
steps so far, for example, the initial impressions, the gothic arch tracing, or the use of a
facebow [7]. Nonetheless, the digital protocol allows for fewer steps and, therefore,
should concatenate fewer uncertainty sources leading to a better design. Moreover,
nowadays, the use of the digital facebow is a promising option for the evolution of such
measurements.
There are still not many insights regarding digital articulators in the literature. Some
studies looked at the influence of different parameters on the behaviour of digital artic-
ulators, such as the vertical dimension [8], or the inclination of the sagittal condylar
path [9]. Another study compare the maximum intercuspal between classically articu-
lated casts and digitally articulated ones [10], and the effect of the different settings for
the digital articulator was tested on the occlusal morphology of restoration done with
CAD/CAM technologies [11]. These results, with some limited disagreements, confirm
the virtual articulator as a reliable tool to design denture bases.
However, these studies consider cases of patients with natural teeth; with edentulous
cases, it is not trivial to define reliable landmarks on which to base the comparison.
Moreover, to the best of our knowledge, there are no studies that specifically investi-
gated if the simulation of occlusion of the digital articulator is comparable to the one
from the semi-individual mechanical articulators, specifically for edentulous cases.
To fill this gap, this study aims to provide a methodology to compare the different
simulations of occlusion given by different articulators for completely edentulous
cases. The comparison is based on the actual trajectory of the casted or digital model
during three movements: left and right lateral, and protusive movement. The procedure
can work both digital vs analogic, digital vs digital or analogic vs analogic articulators.
In this contribution, the comparison between an analogic and a digital articulator is
provided.

2 Materials and methods

The overall workflow used for the study can be seen in Fig. 1. Starting from the
same impression and intermaxillary occlusal registrations, obtained from a simulated
completely edentulous patient, two parallel clinical workflows are followed, the digital
one obtaining the digitally articulated STL models, and the analogic one obtaining the
plaster model mounted and articulated in the Gerber articulator.
For both cases, the trajectory of the maxilla is recorded by mapping a reference sys-
tem built on the maxilla from a second common reference system built on the mandible.
The trajectories are then compared to assess whether the twos are the same or are sig-
nificantly different.
Three different movements are analyzed: left and right lateral and protrusive move-
ments.
 3

Fig. 1. Overall workflow of the study.

2.1 Clinical protocol

A completely edentulous patient is simulated using two edentulous plaster models

already mounted in a mechanical articulator, considered as the patient's oral cavity. The
standard protocol to obtain the final impressions and the intermaxillary occlusal regis-
trations for edentulous patients is followed. The intermaxillary occlusal registrations
are set as the clinical average value for all settings, these values are used for both pro-
tocols. The final results are the articulated impressions, which were scanned using a E3
scanner obtaining a digital articulated impression used in the digital protocol, and then
were cast with type IV plaster obtaining the physical model used in the Gerber articu-
lator. The digital articulator “Artex CR Adjustable” in the Exocad software is used for
the digital protocol. The digitally articulated impressions are imported. The software
creates the two digital models already in position. The models, together with the digi-
tally articulated impressions are aligned using the information contained in the inter-
maxillary occlusal registration following the standard clinical protocol. The plaster
models are mounted into the Gerber articulator following the standard clinical proce-
dure that involves the separation between the upper and lower impression and their
recombination once the mandible model is fixed in position with the use of the face
bow. Further details about the clinical protocol and the parameters available in analog
and digital articulators can be found in [12].

2.2 Movement mapping

To map the movement described by the maxilla in the articulator three incremental
points were sampled per each movement plus the common resting position. Thus, each
movement is characterized by a total of four points.
The digital articulator has no physical end to each movement: the operator can
choose whatever displacement value for each movement and the maxilla is moved into
position. Once in position, the two digital models can be exported into an STL file. The
three incremental points were chosen to be 3mm apart from one another.
4

The Gerber articulator has a physical endpoint for each movement that is chosen as
the last point to be sampled; two intermediate points were then sampled. Once locked
in position the articulator was scanned as described in the following.
The scanning procedure is needed since a physical mechanical articulator is consid-
ered, Fig. 2. Since the two casted models are close to each other, it is impossible to
acquire the actual oral cavity.

Fig. 2. Gerber articulator ready to be scanned.

To address this issue, a strategy commonly used with oral scanners is used: the two
models are individually scanned, and then the articulated models are scanned obtaining
mainly the external surface which is used to correctly align the individual scans. Five
spherical landmarks were added to each model to increase the alignment accuracy.
To test the reproducibility of the workflow applied to the physical articulator, each
position was scanned ten times, thus obtaining ten trajectories.
To compare the two articulators the same reference system was established on the
mandibular model and the position of the articulated maxilla model was tracked through
the position of three fixed points reconstructing the rigid body trajectory of the simu-
lated occlusion. The three points define a cartesian reference system.
Starting from the digital models, two simplified bases were defined and three spher-
ical markers were added, Fig. 3. The CAD reference system of the lower base was used
as a common reference system for the comparison, therefore the scanned model or the
exported model were first aligned to this base in GOM Inspect obtaining the aligned
model that was exported.

Fig. 3. Simplified bases with markers.

At this point, each newly exported model shares the same origin and orientation.
This file is imported in GOM Inspect and the upper base is aligned to the mucosa. The
locations of the three spheres are recorded and elaborated in Excel to define the maxilla-
 5

based reference system. The origin is defined as the centroid of the spheres, the x-y
plane as the plane through the spheres, and the x-axis point towards the central sphere
and z-axis pointing upwards.
This procedure was performed per each sampled point in the analogic and digital
protocol. The trajectories are calculated in a mathematical graphical tool: GeoGebra.
This tool was chosen since it allows real-time testing and modification of the procedure.
The details of the maxilla-based reference systems were imported into the 3D space and
the trajectories are defined as cubic interpolating splines through the origins. For the
digital articulator, only one repetition was done since the exporting is by definition
perfectly repeatable. At the same time, the scanning procedure of the mechanical artic-
ulator introduces variability due to the 3D Scan itself and due to the alignment. For this
reason, the Gerber articulator was mapped ten times to determine the repeatability.
Among the ten trajectories derived from the mechanical articulator, the average tra-
jectory had to be defined to define the repeatability. The intermediate points that were
sampled cannot be considered repeatable because of the uncertainty in the looking of
the articulator, but the length of the trajectory is theoretically the same. Therefore, con-
gruent points along the trajectories were defined as points at the same percentual dis-
tance from the resting position, and points at 0%, 30%, 60%, 90%, and 100% of the
trajectory length were sampled. The average trajectory was defined by averaging the
position of these points. The dispersion of the mechanical articulator is defined by the
distance of each trajectory to the average one, the distance is defined, point by point, in
the plane normal to the average trajectories, Fig. 4.

Fig. 4. Analog trajectories error definition (a) and dispersion (b).

The dispersion was sampled at steps of 15% of the average trajectory length exclud-
ing the resting position. Assuming a bivariate normal distribution with the same disper-
sion in each direction for the sampled trajectories, and therefore a Rayleigh distribution
for the distances (radial component), the standard deviation of the bivariate distribution
is defined by:

2 ∑𝑛
𝑖=1 𝑑𝑖
𝜎=√
𝜋 𝑛

The dispersion of the resting position is divided into two contributions. For the lat-
eral movement, the aforementioned formula is applied to the projection of the resting
position point into the sagittal plane (lateral plane, x-z); for the protusive movement,
the projection into the frontal plane (y-z) is considered instead.
6

2.3 Movement comparison

The differences between the digital and analogic trajectories are defined by the dis-
tance between the average analogic trajectory - including its dispersion - and the digital
one. Since the analogic trajectory ended up being shorter than the digital one, it was
used as the baseline.
First, the baseline error, define as the distance between the resting positions is cal-
culated and represents the initial bias between the two trajectories.
The digital trajectory is translated of the quantity defined as the baseline error to
check for differences excluding the initial bias.
The distance between the two trajectories is calculated at steps of 15% of the ana-
logic trajectory as the distance in the plane normal to the baseline trajectory.

3 Results and discussion

Regarding the mechanical articulator, the ten trajectories that were sampled and the
average trajectory, with its associated standard deviation, can be seen in Fig. 5.a). The
repeatability test, applied to the mechanical articulator, shows a maximum standard
deviation of 184 microns for the left lateral movement, see Fig. 5.b).

Fig. 5. Result of the ten trajectories sampled with the Gerber articulator (black lines), the average
trajectories (red lines), and the standard deviation along the three movements.

The reparability of the rest position is 123 microns in the sagittal plane and 93 microns
in the frontal plane. On average the repeatability decrease (higher standard deviation)
the further from the resting position; indeed the locking in position adds a level of un-
certainty to the procedure.
The comparison between the two articulators' trajectories can be seen in Fig. 6.a).
At first glance, the difference in length between the two trajectories can be noted. None-
theless, this does not imply an actual difference because the shorter trajectory can still
represent the first part of the longest one.
The first real difference that can be recorded is the initial bias between the two ar-
ticulators, the resting positions are 1,373 mm apart. Considering the dispersion of the
resting position for the mechanical articulator this difference is considered significant.
This means the bias is not due to the measuring uncertainty but is based on an actual
difference. The reason behind this deviation can be traced back to the analogic clinical
 7

protocol when the articulated impressions need to be separated for the alignment in the
Gerber articulator. Maxillary and mandibular casts are then fixed again thanks to the
gothic arch plates and the bite registration during the clinical protocol. This step may
introduce a shift between the two impressions therefore it is the main candidate to ex-
plain the bias.

Fig. 6. Comparison between the analogic and digital trajectories.

To evaluate the actual differences between the two trajectories, excluding the bias,
the digital trajectories are rigidly translated to share the same resting position with an-
alogic ones, Fig. 6.b). Comparing the two trajectories, taking into consideration the
standard deviation associated with the analogic trajectory, it can be said that the trajec-
tories are significantly different. The local error along the trajectories can be seen in
Fig. 6.c).
The local error increase as the distance from the resting position increase meaning
that the two trajectories diverge. Looking at Fig. 6.b) it can be seen that this divergence
is not due to an alignment issue but it come from the trajectories having different
shapes; for instance, the mechanical articulator has a shallow angle between the two
lateral movement when compared to the digital one. The maximum recorded difference
is 1.350 mm for the right lateral movement.
It must be noted that this study does not want to conclude that any of the tested
articulators is better than the other. Differences were expected since the type of articu-
lators is different. The same protocol may be used to check whether a digital articulator
simulating a specific mechanical model and the actual physical counterpart coincide.

4 Conclusions

This paper aims to propose a methodology to compare the dynamic movements of dif-
ferent articulators used in dentistry. The methodology is designed having in mind to
give the possibility to compare mechanical vs mechanical, mechanical vs digital, and
digital vs digital articulators. To prove its usability the methodology was tested by com-
paring a mechanical articulator (Gerber) to a digital one (Artex CR Adjustable” in the
Exocad). Therefore differences were expected.
As expected significant differences were found proving the overall methodology to
be a valid means to compare different articulators. A far more interesting comparison
for future works will be the study of corresponding digital and mechanical articulators
 8

to test whether the digital articulator can provide a precise replica of the, so far, standard
clinical procedure based on mechanical articulators.
The late development of digital face bows capable of tracking the actual 3D jaw
movement of the patient without any approximation (needed using an articulator both
digital or mechanical based on a set of values) opens the possibility to compare the
actual patient movements to the one approximated by the articulators.
Further studies will also be necessary to automate the procedure and test a clinically
significant population to provide clinicians with useful data to improve dentures design.

References
1. Baba, N.Z., Goodacre, B.J., Goodacre, C.J., Müller, F., Wagner, S.: CAD/CAM Complete
Denture Systems and Physical Properties: A Review of the Literature. J. Prosthodont. 30,
113–124 (2021). https://doi.org/10.1111/jopr.13243.
2. Goodacre, B.J., Goodacre, C.J., Baba, N.Z., Kattadiyil, M.T.: Comparison of denture tooth
movement between CAD-CAM and conventional fabrication techniques. J. Prosthet. Dent.
119, 108–115 (2018). https://doi.org/10.1016/j.prosdent.2017.02.009.
3. Srinivasan, M., Cantin, Y., Mehl, A., Gjengedal, H., Müller, F., Schimmel, M.: CAD/CAM
milled removable complete dentures: an in vitro evaluation of trueness. Clin. Oral Investig.
21, 2007–2019 (2017). https://doi.org/10.1007/s00784-016-1989-7.
4. AlHelal, A., AlRumaih, H.S., Kattadiyil, M.T., Baba, N.Z., Goodacre, C.J.: Comparison of
retention between maxillary milled and conventional denture bases: A clinical study. J.
Prosthet. Dent. 117, 233–238 (2017). https://doi.org/10.1016/j.prosdent.2016.08.007.
5. Bisler, A., Bockholt, U., Kordass, B., Suchan, M., Voss, G.: The virtual articulator. Int. J.
Comput. Dent. 5, 101–6 (2002).
6. White, G.E.: The Gerber articulator and system of full denture construction. Dent. Tech. 26,
12–20 (1973).
7. Úry, E., Fornai, C., Weber, G.W.: Accuracy of transferring analog dental casts to a virtual
articulator. J. Prosthet. Dent. 123, 305–313 (2020).
https://doi.org/10.1016/j.prosdent.2018.12.019.
8. Lee, H., Burkhardt, F., Fehmer, V., Sailer, I.: Accuracy of Vertical Dimension Augmentation
Using Different Digital Methods Compared to a Clinical Situation-A Pilot Study. Int. J.
Prosthodont. 33, 380–385 (2020). https://doi.org/10.11607/ijp.6402.
9. Hsu, M.R., Driscoll, C.F., Romberg, E., Masri, R.: Accuracy of Dynamic Virtual Articulation:
Trueness and Precision. J. Prosthodont. 28, 436–443 (2019).
https://doi.org/10.1111/jopr.13035.
10. Arslan, Y., Bankoğlu Güngör, M., Karakoca Nemli, S., Kökdoğan Boyacı, B., Aydın, C.:
Comparison of Maximum Intercuspal Contacts of Articulated Casts and Virtual Casts
Requiring Posterior Fixed Partial Dentures. J. Prosthodont. 26, 594–598 (2017).
https://doi.org/10.1111/jopr.12439.
11. Olthoff, L., Meijer, I., de Ruiter, W., Bosman, F., van der Zel, J.: Effect of virtual articulator
settings on occlusal morphology of CAD/CAM restorations. Int. J. Comput. Dent. 10, 171–
85 (2007).
12. Goldstein, G., Goodacre, C.: Selecting a Virtual Articulator: An Analysis of the Factors
Available with Mechanical Articulators and their Potential Need for Inclusion with Virtual
Articulators. J. Prosthodont. 32, 10–17 (2023). https://doi.org/10.1111/jopr.13517.