RESEARCH AND EDUCATION

Influence of abutment tooth geometry on the accuracy

of conventional and digital methods of obtaining
dental impressions
Jeison B. Carbajal Mejía, DDS, PhD,a Kazumichi Wakabayashi, DDS, PhD,b Takashi Nakamura, DDS, PhD,c and
Hirofumi Yatani, DDS, PhDd

ABSTRACT
Statement of problem. Direct (intraoral) and indirect (desktop) digital scanning can record abutment tooth preparations despite their ge-
ometry. However, little peer-reviewed information is available regarding the influence of abutment tooth geometry on the accuracy of digital
methods of obtaining dental impressions.
Purpose. The purpose of this in vitro study was to evaluate the influence of abutment tooth geometry on the accuracy of conventional and
digital methods of obtaining dental impressions in terms of trueness and precision.
Material and methods. Crown preparations with known total occlusal convergence (TOC) angles (−8, −6, −4, 0, 4, 8, 12, 16, and 22 degrees)
were digitally created from a maxillary left central incisor and printed in acrylic resin. Each of these 9 reference models was scanned with a highly
accurate reference scanner and saved in stereolithography (STL) format. Then, 5 conventional polyvinyl siloxane (PVS) impressions were made
from each reference model, which was poured with Type IV dental stone scanned using both the reference scanner (group PVS) and the desktop
scanner and exported as STL files. Additionally, direct digital impressions (intraoral group) of the reference models were made, and the STL files
were exported. The STL files from the impressions obtained were compared with the original geometry of the reference model (trueness) and
within each test group (precision). Data were analyzed using 2-way ANOVA with the post hoc least significant difference test (a=.05).
Results. Overall trueness values were 19.1 mm (intraoral scanner group), 23.5 mm (desktop group), and 26.2 mm (PVS group), whereas overall
precision values were 11.9 mm (intraoral), 18.0 mm (PVS), and 20.7 mm (desktop). Simple main effects analysis showed that impressions made
with the intraoral scanner were significantly more accurate than those of the PVS and desktop groups when the TOC angle was less than 8
degrees (P<.05). Also, a statistically significant interaction was found between the effects of the type of impression and the TOC angle on the
precision of single-tooth dental impressions (F=2.43, P=.002). Visual analysis revealed that the intraoral scanner group showed a
homogeneous deviation pattern across all TOC angles tested, whereas scans from the PVS and desktop scanner groups showed marked
local deviations when undercuts (negative angles) were present.
Conclusions. Conventional dental impressions alone or those further digitized with an extraoral digital scanner cannot reliably reproduce
abutment tooth preparations when the TOC angle is close to 0 degrees. In contrast, digital impressions made with intraoral scanning can
accurately record abutment tooth preparations independently of their geometry. (J Prosthet Dent 2016;-:---)

Direct digital methods of obtaining dental impressions
were introduced into the dental field in the 1980s.1 These
systems evolved rapidly because of improvements in
computing processing power, graphic rendering, and
display screens that are now capable of capturing and
storing high quality 3-dimensional (3D) virtual images of
tooth preparations or dental casts from which restorations
can be fabricated.2 Also, a number of companies offer
intraoral scanners that are increasingly user-friendly,
perceived as pleasant for the patient,3,4 accurate when
recording short-span areas,5-9 time efficient,10 and capable
of producing well-fitting restorations.11 However, among

Intraoral scanner provided by the Japanese branch in Kyoto of 3Shape Copenhagen, Denmark. Previously presented as a poster and awarded first prize
at the Annual Meeting of the Korean Academy of Esthetic Dentistry, Seoul, South Korea, December 2015.
a
Doctoral student, Department of Fixed Prosthodontics, Osaka University Graduate School of Dentistry, Osaka, Japan.
b
Associate Professor, Department of Fixed Prosthodontics, Osaka University Graduate School of Dentistry, Osaka, Japan.
c
Associate Professor, Department of Fixed Prosthodontics, Osaka University Graduate School of Dentistry, Osaka, Japan.
d
Professor and Chairman, Department of Fixed Prosthodontics, Osaka University Graduate School of Dentistry, Osaka, Japan.

THE JOURNAL OF PROSTHETIC DENTISTRY 1

2 Volume - Issue -

Clinical Implications
Undercuts are not a restricting factor for intraoral
digital impressions as the computer-aided design
software can manage them. In a fully digital
workflow, clinicians should be aware of this and
manage undercuts with care; however, their effect
on the definitive restoration is still unknown.

their limitations are high initial investment cost, limited

ability to make complete arch impressions12 and implant
impressions, and limited ability to record preparation
margins obscured by saliva, blood, or soft tissue.2,13,14
Dental impressions, either conventional or digital, are
intended to obtain a copy (imprint) of 1 or several prepared Figure 1. Overall shape of preparations with different convergence an-
teeth, adjacent and antagonist with the interocclusal gles of mesial and distal walls.
relationship; therefore, their accuracy is crucial to the
fabrication of precisely fitting dental restorations. Accuracy
Undercuts should be avoided during tooth preparation
consists of both precision and trueness (International
or blocked out before making a conventional impression to
Organization for Standardization [ISO] standard 5725-
prevent distortion of the impression when it is removed
1),15 with precision defined as the deviation of multiple
from the mouth. However, undercuts are not a restricting
measurements and trueness as how far a measured result
factor for digital impressions because the scanner will
deviates from the actual size of the measured object.
record them as long as it is rotated to completely visualize
The digital restorative workflow aims to avoid the error-
the undercuts. In addition, minor undercuts can be easily
prone multistep process of conventional impressions,
managed in the CAD stage (for example, by changing the
gypsum cast production, and extraoral digitalization16 by
path of insertion of the crown or digitally blocking
using a more standardized, reliable, and predictable
undercuts) before fabricating the restoration. Despite the
approach.2,17 It starts with the generation of a 3D virtual
progress in technology that has led to current intraoral
replica of the clinical situation,2,10 followed by computer-
scanners, little peer-reviewed information is available
aided design and computer-aided manufacturing (CAD-
regarding the influence of abutment tooth geometry on the
CAM) of a prosthetic restoration. Any error that may arise
accuracy of conventional and digital impression methods.
during this workflow18-21 will have a cumulative effect
The purpose of this study was to evaluate the influence
throughout the production process and, if not controlled,
of abutment tooth geometry on the accuracy of conven-
threshold values22-26 for the marginal and internal fit of
tional and digital methods of obtaining dental impressions
dental restorations could be crossed.
with respect to trueness and precision. The null hypoth-
Errors during data acquisition with intraoral scanners
esis was that no significant differences would be found
can be attributed to the visual interference that occurs
between conventional and digital impression methods.
when the camera tilt angle exceeds the tooth’s axial wall
angle of convergence.27 Such sources of interference are
MATERIAL AND METHODS
less likely with extraoral digitizers because multiple scans
from different directions are required to digitize the A resin maxillary left central incisor (Nissin Dental
scanned body,5,6,28 although the color of the light used Products Inc) was scanned with a desktop scanner (KaVo
when scanning may influence accuracy.29 Arctica Scan; Smartoptics Sensortechnik), and a mesh of
The total occlusal convergence (TOC) angle of a the tooth surface was generated using proprietary soft-
prepared abutment tooth has been defined as the ware. The surface meshing was followed by nonuniform
converging angle of 2 opposite axial walls in a given rational B-spline conversion to obtain a 3D model.
plane.30 Also, the TOC angle influences the retention Starting from the intact tooth model, a total of 9 prepa-
and resistance of dental restorations, with parallel axial rations were created using general purpose CAD soft-
walls providing maximum values, and highly converging ware (Rhinoceros 3D for Windows [Microsoft]; McNeel
walls providing the least values.31 However, producing North America) by removing tooth structure in in-
parallel walls or optimal TOC angles (less than 10 crements with different TOC angles in the mesiodistal
degrees) is a clinically challenging task and prone to aspect, namely −8-, −6-, −4-, 0-, 4-, 8-, 12-, 16-, and 22-
preparation undercuts.32,33 degree preparations (Fig. 1). The preparation was a

THE JOURNAL OF PROSTHETIC DENTISTRY Carbajal Mejía et al

- 2016 3

complete ceramic crown produced in accordance with

standard guidelines.34 All models were then printed
(ProJet 3500HD Max; 3D Systems) in acrylic resin to
obtain a physical reference model. Each of these 9
reference models was scanned 5 times (n=5) using a
reference scanner (Rexcan DS; Solutionix [manufacturer’s
specifications: accurate to within 10 mm]), and the
resulting reference scan data set (R1-R5) was further
compared with those of all other test groups.
Forty-five conventional 2-step impressions were
made with polyvinyl siloxane (PVS) material (Examixfine
putty/wash; GC Corp), with 5 copies obtained (n=5) for
every reference model (−8-, −6-, −4-, 0-, 4-, 8-, 12-, 16-,
and 22-degree preparations). To simulate the clinical
workflow, all impressions were made with prefabricated,
individual acrylic trays of 4-mm thickness coated with Figure 2. Impression-making and superimposition procedure for
tray adhesive (Adhesive; GC Corp). All impression ma- single-tooth conventional and digital impressions. KAV, extraoral
terials were mixed in standardized proportions according scanned casts with KaVo scanner; PVS, polyvinyl siloxane impressions;
REF, reference; TRI, intraoral digital impressions with TRIOS.
to the manufacturer’s recommendations at room tem-
perature (25 C) and ambient humidity by a single
maximum deviation accepted for the best-fit alignment
investigator (J.C.). The preliminary putty impressions
was set at 0.1 mm for reference scans and 1 mm for test
were made first, using a plastic foil on top of the refer-
scanning data. The 3D deviation spectrum was set at 15
ence model, and left for 8 minutes to polymerize. The
color segments, maximum/minimum critical was set at
second step included removal of the plastic foil, injection
±50 mm, and maximum/minimum nominal was set at
of the wash material, and removal of the impressions
±10 mm. For quantitative analysis of 3D differences, the
after 8 minutes of polymerization time. Once obtained,
mean quadratic deviation (root mean square [RMS])5,35,36
all impressions were disinfected for 10 minutes (MD 520;
of the virtual reference object compared with the test
Dürr Dental AG), cast after 8 hours with a Type IV
objects was registered, while for the visual analysis of
gypsum (New Fujirock; GC Corp), and stored for 48
deviation patterns, color-coded images were saved as
hours. Next, all gypsum casts were scanned once with
screenshots. The trueness values of each test group were
the reference scanner, and data were exported as open-
obtained by superimposing each model scan (n=5) with
format stereolithography (STL) files.
the reference scan data set, whereas precision values
All previously prepared gypsum casts (5 casts per refer-
were calculated as the mean RMS between all superim-
ence model) were scanned once with a desktop scanner
position combinations within 1 test group (n=10) (Fig. 2).
(KaVo Arctica Scan; KAV) according to the manufacturers’
All RMS values were analyzed by usingstatistical
instructions and exported as STL files. This resulted in 5
software (XLSTAT v2014; Addinsoft). The Shapiro-Wilk
extraoral scans (n=5) of the same reference model. For the
test for normality and the Levene test for equality of
direct digital impressions group, a matting powder (Okklu-
variances were performed (a=.05). Statistical differences
sionsspray; Yeti Dental) was used to pretreat the surface of
between the test groups were analyzed using 2-way
the reference model before obtaining 5 optical impressions
ANOVA with 2 factors (type of impression and TOC
per model (n=5) using the TRIOS intraoral scanner (3Shape;
angle), and pairwise post hoc comparisons were done
TRI). Scan data were sent to the manufacturer for post-
using the least significant difference test (a=.05).
processing and then exported as STL data files.
Data handling and computations were performed
RESULTS
using metrology software (Geomagic Control 2014; 3D
Systems). Artifacts from the visualized data sets and An overview of the results and statistics is presented in
unnecessary data below the simulated preparation Tables 1-3 and Figure 3. The overall trueness values of all
margin were removed. To compare the test groups with test groups indicated that the TRI group had the smallest
the reference model, 1 scan was selected randomly from deviation, with a mean RMS value of 19.1 mm, followed
the 5 reference data sets (R1-R5) by using a smart phone by the KAV group (23.5 mm) and the PVS group (26.2
application (Undecided; Deadmans Productions) and mm). Statistically significant differences occurred among
imported into the metrology software along with the data all impression techniques (P<.001). The 2-way ANOVA
sets from the PVS, KAV, and TRI groups. An initial indicated a significant difference in both the main effect
manual alignment of the scans was done with the sub- type of impression and the TOC angle (P<.05). A sig-
sequent use of the software’s best-fit algorithm. The nificant interaction was not indicated (P=.162) (Table 2).

Carbajal Mejía et al THE JOURNAL OF PROSTHETIC DENTISTRY

4 Volume - Issue -

Table 1. Mean ±SD RMS trueness and precision values of conventional and digital impressions (mm)
Total Occlusal Convergence
Impression
Technique −8 degrees −6 degrees −4 degrees 0 degrees 4 degrees 8 degrees 12 degrees 16 degrees 22 degrees Overall mean
PVS Trueness 29.1 ±2.2 26.5 ±8.1 32.6 ±8.4 26.7 ±6.8 27.5 ±4.5 28.7 ±7.4 20.2 ±6.1 22.2 ±1.7 22.4 ±4.7 26.2 ±6.6
Precision 14.6 ±1.1 20.9 ±5.3 21.0 ±4.2 19.4 ±3.3 15.8 ±3.0 17.2 ±4.6 18.7 ±2.3 17.8 ±3.7 16.9 ±2.7 18.0 ±3.9
KAV Trueness 24.3 ±5.0 21.6 ±4.9 26.5 ±4.1 26.1 ±6.7 22.5 ±5.2 25.0 ±8.7 23.1 ±7.4 21.7 ±3.1 20.4 ±2.7 23.5 ±5.5
Precision 21.3 ±5.0 21.6 ±5.6 18.7 ±1.9 21.2 ±3.8 19.0 ±2.8 21.9 ±5.8 21.4 ±4.9 21.7 ±5.1 19.6 ±3.7 20.7 ±4.4
TRI Trueness 19.9 ±1.7 17.9 ±1.1 20.6 ±1.0 18.4 ±1.5 18.1 ±1.7 15.8 ±0.9 20.4 ±1.6 20.3 ±2.1 20.2 ±1.1 19.1 ±2.0
Precision 13.4 ±1.8 11.1 ±3.8 13.4 ±1.8 11.1 ±1.2 12.8 ±1.6 11.7 ±1.9 12.2 ±2.3 10.0 ±2.1 11.0 ±1.6 11.9 ±2.3

KAV, extraoral scanned casts with KaVo scanner; PVS, polyvinyl siloxane impressions; RMS, root mean square; TRI, intraoral digital impressions with TRIOS.

Table 2. Results of 2-way ANOVA
Parameter Source df SS MS
Impression technique 2 0.001 0.001
Trueness TOC angle 8 0.000 0.000
Impression technique×TOC 16 0.001 0.000
Impression technique 2 0.004 0.002
Precision TOC angle 8 0.000 0.000
Impression technique×TOC 16 0.000 0.000
TOC, total occlusal convergence. P<.05 indicates significant difference.
The post-hoc least significant difference test showed that
impressions from the TRI group were significantly more
accurate than the impressions from groups PVS and KAV
when the TOC angle was below 8 degrees (P<.05).
Qualitative analysis revealed that casts from the TRI
group showed a very homogenous deviation pattern
(Fig. 4) with no local deviations higher than +45 mm
(yellow to red), whereas groups PVS and KAV had local
deviations at the incisal and proximal areas of the abut-
ment tooth, especially when the TOC was below 0 de-
grees; in some cases (0 and −4 degrees), such deviations
reached up to −100 mm (Fig. 4, navy blue).
Overall precision values were identified as the most
accurate for the TRI group (11.9 mm), followed by the
PVS group (18.0 mm) and the KAV group (20.7 mm).
Statistically significant differences occurred among all
impression techniques (P<.001). The TRI group showed
the highest precision values for all TOC angles tested.
Two-way ANOVA indicated a statistically significant
interaction between the effects of the type of impression
and TOC angle on the precision of single-tooth dental
impressions (P=.002) (Table 2). Visual analysis showed
marked local deviations on the buccal and mesiodistal
aspect of the abutment tooth in the KAV and PVS groups
(Fig. 5). Larger deviations (up to ±50 mm (Fig. 5, red/navy
blue) were visible in the PVS and KAV groups when the
TOC was below 0. Moreover, the deviation pattern was
relatively homogenous across all casts for the TRI group,
except for the incisal edge, where deviations reached up
to +30 mm (Fig. 5, yellow to red).
Mesh topology and triangle density varied markedly
among the scans obtained from different impression
techniques (Fig. 6), with the most accurate group (TRI)
showing the least number of triangles (approximately
17 000 triangles), followed by the KAV group
(approximately 50 000 triangles) and the PVS group
F P (approximately 120 000 triangles).
25.144 <.001
2.083 .044
DISCUSSION
1.386 .162
150.682 <.001 On the basis of the results of the present in vitro study,
1.355 .217 the null hypothesis that the conventional and digital
2.432 .002 methods of obtaining dental impressions are equally
accurate was rejected. The accuracy of the PVS and KAV
groups was affected when the TOC angle was close to
0 degrees, whereas impressions made in the TRI group
were not. This result is in accordance with a previously
published study,28 except that intraoral scanning seems
not to be affected by the TOC angle of a preparation.
In this study, metrology software was used to super-
impose and analyze scans from all impression techniques
tested.5,35,36 For this purpose, all scans were exported as
STL files. In this file format, a scanned surface is approxi-
mated using triangles from the point cloud generated by
the 3D scanner. Scanners with high-definition sensors
generate more points and a shorter point-to-point dis-
tance, thus producing more triangles for surface recon-
struction that would result in a more detailed
representation of the scanned body. In the present work,
triangle density and mesh topology varied markedly
among scans obtained from the different impression
techniques (Fig. 6) and when the triangle density of the test
groups was compared in relation to the reference scanner,
the most accurate group (TRI) had the lowest number of
triangles by a factor of 6.8, followed by the the KAV group
(2.4) and the PVS group (1.0). Nedelcu and Persson9 also
reported that highly accurate scans (3M Lava and CEREC
Bluecam) had a low triangle density; therefore, it can be
hypothesized that scanners with higher-definition sensors
would not necessarily produce highly accurate scans;
instead, technological aspects might have influenced the
final accuracy of the datasets.
Undercuts (negative angles) should generally be
avoided during tooth preparation or blocked out to a
great extent before making a conventional impression to
prevent distortion of the impression when it is removed
from the mouth. However, undercuts neither restrict the
making of intraoral digital impressions nor are they
detrimental to impression accuracy. Furthermore, if a
minor undercut is recorded at the first stage of a fully

THE JOURNAL OF PROSTHETIC DENTISTRY Carbajal Mejía et al

- 2016 5

Table 3. Statistical significance among test groups for trueness measurement according to 2-way ANOVA with post hoc LSD test (a=.05)
PVS KAV TRI
TOC Angle −8 −6 −4 0 4 8 12 16 22 −8 −6 −4 0 4 8 12 16 22 −8 −6 −4 0 4 8 12 16 22
PVS
−8 0 0 0 0 0 X X X 0 X 0 0 X 0 0 X X X X X X X X X X X
−6 X 0 0 0 X 0 0 0 0 0 0 0 0 0 0 X X X 0 X X X X X X
−4 0 0 0 X X X X X X X X X X X X X X X X X X X X X
0 0 0 X 0 0 0 0 0 0 0 0 0 0 X X X X X X X X X X
4 0 X 0 0 0 0 0 0 0 0 0 0 X X X X X X X X X X
8 X X X 0 X 0 0 X 0 0 X X X X X X X X X X X
12 0 0 0 0 X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X 0 0 0
22 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X 0 0 0
KAV
−8 0 0 0 0 0 0 0 0 0 X 0 0 X X 0 0 0
−6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
−4 0 0 0 0 0 0 X X 0 X X X X X X
0 0 0 0 0 0 X X 0 X X X 0 0 0
4 0 0 0 0 0 0 0 0 0 X 0 0 0
8 0 0 0 0 X 0 X X X 0 0 0
12 0 0 0 0 0 0 0 X 0 0 0
16 0 0 0 0 0 0 0 0 0 0
22 0 0 0 0 0 0 0 0 0
TRI
−8 0 0 0 0 0 0 0 0
−6 0 0 0 0 0 0 0
−4 0 0 0 0 0 0
0 0 0 0 0 0
4 0 0 0 0
8 0 0 0
12 0 0
16 0
22

0 indicates no statistical difference; X indicates statistical difference. KAV, extraoral scanned casts with KaVo scanner; PVS, polvinyl siloxane impressions; LSD, least significant difference;
TOC, total occlusal convergence angle (degrees); TRI, intraoral digital impressions with TRIOS.

digital workflow, the clinician can either remove more
dental structure to obtain a tapered preparation or
whenever possible take advantage of the algorithms from
the CAD software to overcome them (noted in KaVo
multiCAD instructions; http://www.kavo.com/de/file/
552/download?token=dLSm_utB; CEREC 3D Prepara-
tion Guidelines, http://www.sirona.com/ecomaXL/files/
cerec_3d_preparation_guidelines_en.pdf&download=1).
Naturally, preserving dental structure without sacrificing
the integrity of the definitive restoration is desirable,
provided that the design software can successfully
manage undercuts.
Differences in trueness and precision were found,
depending on the TOC angle of the abutment tooth and
the impression technique. Impressions made with TRIOS
showed the highest trueness and precision values, even
when negative angles were included in the abutment
tooth. The PVS and KAV groups differed significantly
from the TRI group when the TOC angle was close to
0 degrees. Casts from the KAV group showed the lowest
precision in relation to the TRI group (11.9 mm versus
20.7 mm, respectively). This outcome is in agreement with
that of Vandeweghe et al16 who reported large deviations
in casts digitized with the KaVo scan. One of the ex-
planations for this outcome could be the accumulation of
dimensional errors originating from the impression
material, gypsum, and digitization process. A second
factor could be the nature of the light used by the KaVo
scanner (white-light scanner), as Jeon et al29 showed that
blue-light scanners exhibited greater precision than
white-light scanners. A third explanation may be the
threshold value for this scanner (the manufacturer
specifies an accuracy of at least 20 mm). Contrary to these
possible detrimental factors, if the margins of the
extraoral scanned casts had been ditched, their accuracy
might have improved28 because ditching facilitates the
automatic identification of the preparation margin by the
scanning software. In contrast, Su and Sun6 found that
the precision of single-tooth intraoral impressions
(TRIOS) and extraoral digitization (D800; 3Shape) was
similar in either anterior (13.33 mm versus 14.89 mm) or
posterior abutment teeth (7.0 mm versus 8.67 mm).

Carbajal Mejía et al THE JOURNAL OF PROSTHETIC DENTISTRY

6 Volume - Issue -

30 a 25
b
b
a
25
20
c

Precision, RMS (µm)

Trueness, RMS (µm)

20
15 c

15

10
10

5
5

0 0
PVS KAV TRI PVS KAV TRI
Impression Procedure A Impression Procedure B
Figure 3. Accuracy of conventional and digital impressions (mm) A, Trueness. B, Precision. Groups with different lowercase letters indicate significant
differences at P<.05. KAV, extraoral scanned casts with KaVo scanner; PVS, polyvinyl siloxane impressions; RMS, root mean square; TRI, intraoral digital
impressions with TRIOS.

Figure 4. Typical deviation pattern between test impression and reference model (trueness). Deviation range color coded from +50 mm (dark red)
to −50 mm (dark blue). Maximum/minimum nominal ±10 mm (green). K, extraoral scanned casts with KaVo scanner; P, polyvinyl siloxane impressions;
T, intraoral digital impressions with TRIOS.

Therefore, deviations found in the KAV group must have
been influenced by the type of scanner and the TOC
angle rather than the technique itself.
A comparison of the results of the present study
with previously published research on the accuracy of
single-tooth intraoral digital impressions showed that
previous studies reported trueness values of 27.9 mm
(CEREC),5 19.2 mm (CEREC Bluecam)7 and 6.9 mm
(TRIOS),8 and precision values of 13.3 mm (TRIOS),6
10.8 mm (CEREC Bluecam),7 and 4.5 mm (TRIOS),8
while the present work found trueness values as accu-
rate as 19.1 mm and precision values as accurate as 11.9
mm. Different values reported in different studies can be
explained by the different study design and improve-
ments in the devices.
The results of the present study should be inter-
preted with caution as it has been shown that intraoral
digital impressions are highly accurate when scanning
single or partial fixed prosthesis preparations but their
accuracy is limited for complete arch impressions.12 This
may be due to the lack of fixed references19 when a
surface is scanned. Thus, the first image recorded by the
scanner is the reference, and all subsequent images will
be stitched to the previous one by a best-fit algorithm
that represents the best possible overlap of images. Each
overlap has an inherent error and it can be anticipated
that the longer the scanning field the larger the error
introduced. A clear example of this would be that Ender
and Mehl12 reported deviations of up to 170 mm in the
posterior area during complete arch scanning, whereas

THE JOURNAL OF PROSTHETIC DENTISTRY Carbajal Mejía et al

- 2016 7

Figure 5. Representative deviation pattern between impressions of one test group (precision). Deviation range color coded from +50 mm (dark red)
to −50 mm (dark blue). Max/min nominal ±10 mm (green). K, extraoral scanned casts with KaVo scanner; P, polyvinyl siloxane impressions; T, intraoral
digital impressions with TRIOS.

Figure 6. Differences in mesh triangle density and topology among scans obtained from different impression techniques. A, Intraoral digital
impressions with TRIOS. B, Extraoral scanned casts with KaVo scanner. C, Polyvinyl siloxane impressions.

trueness values as small as 6.9 mm were found for
single-tooth digital impressions.26 Moreover, the accu-
racy of earlier versions of intraoral scanners was shown
to be affected by the tilt angle of the scanner if this
angle exceeded the tooth’s axial wall angle of diver-
gence.27 However, this no longer seems to be a limiting
factor for modern versions especially as intraoral scan-
ners are constantly being updated.
Furthermore, the fully digital workflow is not exempt
from errors,18 and inaccuracies may arise from the digi-
tizing device and its underlying scanning technology,
digitizing environment (blood, saliva, and natural and
external light source), scanning strategy, and scanned
data processing (alignment algorithms). Also, a trans-
formation error (9.7 mm)20 can occur in the CAD stage
when direct digital data acquisition is transferred to the
CAD software. This is related to the software’s ability to
reconstruct missing tooth surfaces by using a biomimetic
approach (with B-Splines, Hermit, or nonuniform
rational B-splines).18 Finally, additional inaccuracies can
be introduced (from 5 to 25 mm)21 in the CAM stage by
the CAM software, the numerically controlled machine
tool, and the milling machine process (including me-
chanical loads, vibrations, and tool wear).
The material-dependent dimensional changes and
error-prone multistep processes involved in the con-
ventional fabrication of dental crowns or the propagation
error throughout the digital restorative workflow will lead
to misfit of the definitive restoration. However, no
consensus has been reached as to what constitutes a
clinically acceptable misfit. In this regard, threshold
values reported in the literature vary from 50 to 200
mm,22-25 although a maximum clinically acceptable mar-
ginal discrepancy according to most researchers is less
than 120 mm.26 This means that the accuracy of any
impression technique must be within this range or even
below. In the present study, all scans made with TRIOS
and many from the KAV and PVS groups met this
requirement, but impressions from the PVS and KAV
groups showed deviations of up to −100 mm when the
TOC angle was close to 0 degrees or negative. However,
the present results on accuracy of impression making do

Carbajal Mejía et al THE JOURNAL OF PROSTHETIC DENTISTRY

8 Volume
not provide any information about its influence on the fit
of definite restorations.
This study has several limitations. Idealized prepara-
tions of a maxillary left central incisor were used to make
both conventional and digital impressions. The 3D-prin-
ted resin models differ from human enamel and dentin in
terms of hardness, wettability, surface roughness, and
light reflection. The absence of sulcular fluid, blood,
saliva, patient movements, and temperature-related dis-
tortions differ significantly from the clinical setting. In
addition, as this study used only 1 elastomeric impression
material, 1 desktop scanning system, and 1 intraoral
scanner, general conclusions should be drawn carefully.
Finally, this study is a part of a larger investigation to
evaluate the reliability of algorithms present in the CAD
software to deal with undercuts, their effects on the
marginal and internal fit of dental restorations, and their
influence on the physical properties of milled prosthetic
devices. Those questions will be addressed in future
studies from the present research group.
CONCLUSIONS
Within the limitations of this in vitro study, the following
conclusions were drawn:
1. Elastomer-based dental impressions alone or
further digitized with an extraoral digital scanner
cannot reliably reproduce abutment tooth prepara-
tions when the TOC angle is close to 0 degrees.
2. In contrast, the TRIOS intraoral scanner can accu-
rately record abutment tooth preparations inde-
pendently of their geometry.
REFERENCES
1. Mörmann WH. The evolution of the CEREC system. J Am Dent Assoc
2006;137(suppl):7S-13S.
2. Alghazzawi TF. Advancements in CAD/CAM technology: options for prac-
tical implementation. J Prosthodont Res 2016;30:72-84.
3. Yuzbasioglu E, Kurt H, Turunc R, Bilir H. Comparison of digital and conven-
tional impression techniques: evaluation of patients’ perception, treatment
comfort, effectiveness and clinical outcomes. BMC Oral Health 2014;14:10.
4. Burhardt L, Livas C, Kerdijk W, Van der Meer WJ, Ren Y. Treatment comfort,
time perception, and preference for conventional and digital impression
techniques: a comparative study in young patients. Am J Orthod Dentofacial
Orthop 2016;150:261-7.
5. Luthardt RG, Loos R, Quaas S. Accuracy of intraoral data acquisition in com-
parison to the conventional impression. Int J Comput Dent 2005;8:283-94.
6. Su TS, Sun J. Comparison of repeatability between intraoral digital scanner and
extraoral digital scanner: an in-vitro study. J Prosthodont Res 2015;59:236-42.
7. Mehl A, Ender A, Mörmann W, Attin T. Accuracy testing of a new intraoral
3D camera. Int J Comput Dent 2009;12:11-28.
8. Hack GD, Patzelt SB. Evaluation of the accuracy of six intraoral scanning
devices: an in-vitro investigation. ADA Professional Product Review 2015;10:
1-5.
9. Nedelcu RG, Persson AS. Scanning accuracy and precision in 4 intraoral
scanners: an in vitro comparison based on 3-dimensional analysis. J Prosthet
Dent 2014;112:1461-71.
10. Joda T, Brägger U. Digital vs. conventional implant prosthetic workflows: a
cost/time analysis. Clin. Oral Impl Res 2015;26:1430-5.
11. Abdel-Azim T, Rogers K, Elathamna E, Zandinejad A, Metz M, Morton D.
Comparison of the marginal fit of lithium disilicate crowns fabricated with
CAD/CAM technology by using conventional impressions and two intraoral
digital scanners. J Prosthet Dent 2015:554-9.
THE JOURNAL OF PROSTHETIC DENTISTRY
- Issue -
12. Ender A, Mehl A. Accuracy of complete-arch dental impressions: a new
method of measuring trueness and precision. J Prosthet Dent 2013;109:121-8.
13. Ting-Shu S, Jian S. Intraoral digital impression technique: a review.
J Prosthodont 2015;24:313-21.
14. Logozzo S, Zanetti EM, Franceschini G, Kilpelä A, Mäkynen A. Recent
advances in dental optics-Part I: 3D intraoral scanners for restorative
dentistry. Opt Lasers Eng 2014;54:203-21.
15. International Organization for Standardization. ISO 5725-1:1994 Accuracy
(trueness and precision) of measurement methods and results? Part 1:
General principles and definitions. Geneva: ISO; 1994. Available at: www.iso.
org/obp/ui/#iso:std:iso:5725:-1:ed-1:v1:en. Accessed October 19, 2016.
16. Vandeweghe S, Vervack V, Vanhove C, Dierens M, Jimbo R, De Bruyn H.
Accuracy of optical dental digitizers: an in vitro study. Int J Periodontics
Restorative Dent 2015;35:115-21.
17. Christensen GJ. Will digital impressions eliminate the current problems with
conventional impressions? J Am Dent Assoc 2008;139:761-3.
18. Tapie L, Lebon N, Mawussi B, Fron-Chabouis H, Duret F, Attal JP. Under-
standing dental CAD/CAM for restorations e accuracy from a mechanical
engineering viewpoint. Int J Comput Dent 2015;18:343-67.
19. Gimenez B, Ozcan M, Martinez-Rus F, Pradies G. Accuracy of a digital
impression system based on parallel confocal laser technology for implants
with consideration of operator experience and implant angulation and depth.
Int J Oral Maxillofac Implants 2014;29:853-62.
20. Song YL, Li J, Yin L, Huang T, Gao P. The feature-based posterior crown
design in a dental CAD/CAM system. Int J Adv Manuf Technol 2007;31:
1058-65.
21. Lebon N, Tapie L, Duret F, Attal JP. Understanding dental CAD/CAM for
restorations e dental milling machines from a mechanical engineering
viewpoint. Part A: chairside milling machines. Int J Comput Dent 2016;19:
45-62.
22. Christensen GJ. Marginal fit of gold inlay castings. J Prosthet Dent 1966;16:
297-305.
23. Björn AL, Björn H, Grkovic B. Marginal fit of restorations and its relation to
periodontal bone level. II. Crowns. Odontol Rev 1970;21:337-46.
24. Mitchell CA, Pintado MR, Douglas WH. Nondestructive, in vitro quantifi-
cation of crown margins. J Prosthet Dent 2001;85:575-84.
25. Kohorst P, Brinkmann H, Li J, Borchers L, Stiesch M. Marginal accuracy of
four-unit zirconia fixed dental prostheses fabricated using different
computer-aided design/computer-aided manufacturing systems. Eur J Oral
Sci 2009;117:319-25.
26. McLean JW, von Fraunhofer JA. The estimation of cement film thickness by
an in vivo technique. Br Dent J 1971;131:107-11.
27. Parsell DE, Anderson BC, Livingston HM, Rudd JI, Tankersley JD. Effect of
camera angulation on adaptation of CAD/CAM restorations. J Esthet Dent
2000;12:78-84.
28. Chan DC, Chung AK, Haines J, Yau EH, Kuo CC. The accuracy of optical
scanning: influence of convergence and die preparation. Oper Dent 2011;36:
486-91.
29. Jeon JH, Choi BY, Kim CM, Kim JH, Kim HY, Kim WC. Three-dimensional
evaluation of the repeatability of scanned conventional impressions of pre-
pared teeth generated with white- and blue-light scanners. J Prosthet Dent
2015;114:549-53.
30. The glossary of prosthodontic terms. J Prosthet Dent 2005;94:10-92.
31. Jørgensen KD. The relationship between retention and convergence angle in
cemented veneer crowns. Acta Odontol Scand 1955;13:35-40.
32. Tiu J, Al-Amleh B, Waddell JN, Duncan WJ. Clinical tooth preparations and
associated measuring methods: a systematic review. J Prosthet Dent
2015;113:175-84.
33. Marghalani TY. Frequency of undercuts and favorable path of insertion in
abutments prepared for fixed dental prostheses by preclinical dental students.
J Prosthet Dent 2016;116:564-9.
34. Rosenstiel SF, Land MF, Fujimoto J. Contemporary fixed prosthodontics. 5th
ed. St. Louis: Mosby; 2016. p. 184-208.
35. Cho SH, Schaefer O, Thompson GA, Guentsch A. Comparison of accuracy
and reproducibility of casts made by digital and conventional methods.
J Prosthet Dent 2015;113:310-5.
36. Schaefer O, Schmidt M, Goebel R, Kuepper H. Qualitative and quanti-
tative three-dimensional accuracy of a single tooth captured by elasto-
meric impression materials: an in vitro study. J Prosthet Dent 2012;108:
165-72.
Corresponding author:
Dr. Jeison B. Carbajal Mejía
Department of Fixed Prosthodontics
Osaka University Graduate School of Dentistry
1-8 Yamadaoka, Suita-city
Osaka 565-0871
JAPAN
Email: carbajalds@gmail.com
Copyright © 2016 by the Editorial Council for The Journal of Prosthetic Dentistry.
Carbajal Mejía et al