RESEARCH AND EDUCATION

Three-dimensional differences between intraoral scans and

conventional impressions of edentulous jaws: A clinical study
Lucio Lo Russo, DDS, PhD,a Giammarco Caradonna, DDS,b Giuseppe Troiano, DDS, PhD,c
Angelo Salamini, Eng,d Laura Guida, DDS,e and Domenico Ciavarella, DDS, PhDf

ABSTRACT
Statement of problem. Using intraoral scans for removable dentures has been questioned because of a suggested lack of accuracy. However,
data regarding the accuracy of digital intraoral complete-arch scans are sparse, present some methodological issues, and mostly come from
in vitro studies on dentate casts, which are very different from edentulous arches.
Purpose. The purpose of this clinical study was to evaluate 3D differences between intraoral scans (IOS) and conventional impressions of
edentulous arches by means of digital analysis.
Material and methods. Ten maxillary and 10 mandibular edentulous arches were investigated. For each of them, IOS was performed, and a custom
tray was digitally designed based on these scans. Trays were built by using a 3D printer and used to make a conventional impression with a
polysulfide impression material. The conventional impression was scanned immediately by using the same intraoral scanner and by the same dentist.
Standard tessellation language (STL) files of IOS and the scans of the corresponding conventional impressions (CIS) were superimposed with a
2-phase best-fit alignment in a reverse engineering software program. The corresponding full-scan, 3D mean distance was measured. This procedure
was repeated after trimming the IOS and CIS to eliminate peripheral areas not present in both files, as well as nonmatching areas caused by practical
aspects related to obtaining the IOS (mobile tissue stretching) and the conventional impressions (mobile tissue compression and folding at the
margin of impression), which could have impaired alignment and, consequently, measurement accuracy. The mean distance between the full
and trimmed IOS and CIS was statistically investigated, and subgroup analysis was performed for the maxillary and mandibular arches. The
statistical significance of the differences between the 2 impression methods was also investigated.
Results. The full-scan mean distance between the IOS and CIS (−0.19 ±0.18 mm) was significantly different from that of the trimmed scan mean
distance (−0.02 ±0.05 mm), with no significant differences for maxillary and mandibular arches. The differences between the IOS and CIS were
statistically significant for full scans; they were not significant for trimmed scans, except for the maxillary subgroup.
Conclusions. The mean distance between the IOS and CIS may be significantly different if they are not properly superimposed. The mean
distance (−0.02 ±0.05 mm) between the IOS and CIS falls within the range of mucosa resilience. Thus, 3D differences between the IOS and CIS can
be attributed to the different physics behind the 2 impression methods and not to defects in accuracy of one method compared with the other.
The size of the measured difference between the 2 impression methods was not statistically significant and was not clinically significant for
removable denture fabrication. (J Prosthet Dent 2019;-:---)

The application of digital computer-aided design and intraoral scans, which are routinely applied in fixed res-
computer-aided manufacturing (CAD-CAM) technolo- torations,3 may offer further advantages. The production
gies to the design and fabrication of removable dentures1 of a removable denture in a fully digital approach,
may provide numerous advantages. Current CAD-CAM without a physical cast, eliminates patient discomfort
denture systems start with digitizing impressions or related to conventional impression making, problems of
definitive casts2 by using laboratory scanners. The use of potential allergies to impression materials, errors related
a
Associate Professor of Oral Diseases, Department of Clinical and Experimental Medicine, School of Dentistry, University of Foggia, Foggia, Italy.
b
Resident, Department of Clinical and Experimental Medicine, School of Dentistry, University of Foggia, Foggia, Italy.
c
Resident, Department of Clinical and Experimental Medicine, School of Dentistry, University of Foggia, Foggia, Italy.
d
Co-founder, Sintesi Sud srl, Ariano Irpino (AV), Italy.
e
Private practice, Salus Oris srl, Vallesaccarda (AV), Italy.
f
Aggregate Professor of Oral Diseases, Department of Clinical and Experimental Medicine, School of Dentistry, University of Foggia, Foggia, Italy.

THE JOURNAL OF PROSTHETIC DENTISTRY 1

2 Volume
Clinical Implications
Intraoral scans of edentulous jaws are feasible, and
their accuracy is not significantly different from that
of conventional impressions, either clinically or
statistically. Thus, intraoral scans are a valid option
for removable denture fabrication and can be
integrated into digital denture workflows for
improved efficiency.
to the distortion of impressions materials, and the need
to store physical impressions and gypsum casts.4
Intraoral scans for removable dentures have been
questioned5 because of a suggested lack of accuracy and
difficulties in capturing soft tissues. However, data
regarding the accuracy of digital intraoral complete-arch
scans are sparse and present some methodological
issues. The precise measurement of the accuracy of
intraoral scans is problematic and requires a direct
3D comparison of intraoral scans with the anatomy of the
denture-bearing area. In contrast, the typical reported
experimental design is a comparison of intraoral scans
with casts obtained from a conventional impression6-9
under the unreliable assumption that the process of
production of such reference casts does not affect their
trueness. In addition, most of the available data come
from in vitro studies on dentate casts,9-11 which is
different from edentulous arches. Finally, a definition of a
cutoff value for the accuracy of intraoral scans to be
considered adequate for removable denture fabrication is
lacking. However, the proof of concept of the suitability
of intraoral scans for the fabrication of a functional
denture has recently been reported.12
To provide data useful to address these issues, the 3D
differences between intraoral scans and conventional im-
pressions of edentulous jaws were investigated in a clinical
study by means of 3D digital analysis. The null hypothesis
was that no difference would be found between them.
MATERIAL AND METHODS
Institutional review board approval was received for the
study. Sixteen consecutive edentulous patients requiring
complete dentures were selected and included. All par-
ticipants gave informed consent to be enrolled in the
study. The edentulous maxillary and mandibular arches of
every patient were captured by using an intraoral scanner
(Fig. 1) (TRIOS 3; 3Shape) by 1 trained dentist (L.L.R.).
The scans were imported into a software program (Dental
System; 3Shape) and used to design a custom tray with
the corresponding workflow. The space for the impres-
sion material was set to 1 mm.13 The trays were built by
using a 3D printer (M200; Zortrax) and used to make a
conventional impression with a polysulfide impression
THE JOURNAL OF PROSTHETIC DENTISTRY
- Issue -
material (Permlastic; Kerr Corp) following the manufac-
turer’s instructions (Fig. 2). The obtained conventional
impression was immediately scanned by using the same
intraoral scanner and by the same dentist. A direct scan of
the conventional impression was chosen because the
absence of voids and undercuts in edentulous arches was
expected to minimize the shadowing effect.4
For each patient, the intraoral scan (IOS) and the scan
of the corresponding conventional impression (CIS) were
exported by using the intraoral scanner’s software to the
standard tessellation language (STL) format, compatible
with and importable into other reverse engineering soft-
ware programs for further analyses. For each arch of every
participant, the 3D distance between IOS and the corre-
sponding CIS was measured by using the following
procedure: IOS and CIS STL files were imported into a
software program (Geomagic Wrap 2017; 3D Systems
Inc) to perform a 2-phase best-fit alignment, assuming
IOS as reference and CIS as test object. An initial gross
alignment with symmetry evaluation was carried out by
using 300 points for the comparison; then, IOS and CIS
alignment was optimized with a fine adjustment using
15 000 sample points. Once aligned, the 3D distance be-
tween IOS and CIS was calculated (mean distance be-
tween full scans) and recorded in a data-extraction sheet.
This procedure was repeated after trimming the IOS and
CIS to eliminate peripheral areas not present in both files,
as well as their nonmatching areas caused by practical
aspects related to obtaining the IOS (mobile tissue
stretching) and the conventional impression (mobile tis-
sue compression and folding at the margin of impression)
(Fig. 3), which could have impaired alignment and,
consequently, measurement accuracy. The mean distance
between the trimmed IOS and trimmed CIS was
measured by using the same procedure for full scans. In
addition, a color difference map of each superimposition
was obtained for visualization and analysis of the devia-
tion pattern before and after trimming (Figs. 4, 5).
Distributions of measurement data were analyzed to
determine normality (Shapiro-Wilk test); then, differ-
ences between the means for the analyzed parameters
(mean distance between full scans, mean distance be-
tween trimmed scans) were assessed by appropriate tests
detailed in Table 1. Such differences, expressing the di-
versity between the 2 impression methods, were also
investigated for statistical significance by testing the
following hypothesis: if there was no difference between
IOS and CIS, the distance between them would be zero.
Hence, the 1-sample t test was used to answer the
following question: Is the observed mean distance
significantly different from zero? (Table 2) Subgroup
analysis for maxillary and mandibular arches was also
performed (a=.05 for all tests) by using a statistical
software program (PAST3; https://folk.uio.no/ohammer/
past/, accessed October 30, 2018).
Lo Russo et al
- 2019 3

Figure 1. Intraoral scans. A, Maxilla. B, Mandible.

Figure 2. Conventional impressions. A, Maxilla. B, Mandible.

Figure 3. Superimposition of intraoral scan (green) and conventional impression scan (blue) showing peripheral nonmatching areas. A, Maxillary arch.
B, Mandibular arch.

RESULTS Only the mean distance between full scans (−0.19

±0.18 mm) (Table 1) had a nonnormal distribution
Four patients were completely edentulous in both the
(P=.016, Shapiro-Wilk test; null hypothesis: normal dis-
maxillary and mandibular arches; 6 were edentulous in
tribution); it was significantly different (P<.001) from
the maxillary arch, and 6 were edentulous in the
the mean distance between trimmed scans (−0.02
mandibular arch. In total, 10 maxillary and 10 mandibular
±0.05 mm). Such a difference highlights a significant
arches were investigated.

Lo Russo et al THE JOURNAL OF PROSTHETIC DENTISTRY

4 Volume - Issue -

Figure 4. Color deviation map between intraoral scan and conventional impression of maxillary arch. A, Before trimming: mean distance is +0.04 mm.
Gray color shows portions of reference object (intraoral scan) with no corresponding data on test object (conventional impression); blue color indicates
that test object in more apical position. B, After trimming: mean distance is −0.01 mm.

Figure 5. Color deviation map between intraoral scan and conventional impression of mandibular arch. A, Before trimming: mean distance is 0.00 mm.
Gray color shows portions of reference object (intraoral scan) with no corresponding data on test object (conventional impression); red color indicates
that test object in more coronal position. B, After trimming: mean distance is −0.02 mm.

improvement in the overall alignment accuracy after Table 1. Results and statistical analysis for measured parameters
trimming the scans. Subgroup analysis for maxillary and P (Maxillary
Total Maxillary Mandibular versus
mandibular arches showed no significant difference for Variable Arches Arch Arch Mandibular)
either the mean distance between full scans or the mean Number of investigated 20 10 10
distance between trimmed scans. arches
Differences between IOS and CIS were statistically Mean distance between -0.19 ±0.18 -0.11 ±0.09 -0.26 ±0.29 .105c
full scans (mm)
significant (P<.001) for the full scans; this was applicable Mean distance between -0.02 ±0.05 -0.03 ±0.03 -0.02 ±0.07 .766a
to both the whole investigated group and the maxillary trimmed scans (mm)
and mandibular subgroups (Table 2). However, such P (full vs trimmed) <.001d .024b .006b
differences were not significant (P=.064) for trimmed Bold font indicates statistical significance. Two-sample t test. Two-sample paired t test.
a b

scans, except if only the maxillary arch subgroup (P=.02) c

Mann-Whitney test. dWilcoxon test.

was considered.
DISCUSSION
Clinical data regarding the accuracy of the IOS of
edentulous arches are sparse. For digital scans of palatal
soft tissues in dentate patients, a trueness of 130.54
±33.95 mm and precision of 55.26 ±11.21 mm have been
reported when compared in a 3D analysis with a refer-
ence cast obtained by digitizing gypsum casts poured
from a conventional polyvinyl siloxane impression by
using a laboratory scanner.6 The authors attributed these
differences to the flexibility of palatal mucosa, which can
be compressed by the conventional impression tech-
nique. The IOS is a mucostatic impression, whereas every

THE JOURNAL OF PROSTHETIC DENTISTRY Lo Russo et al

- 2019
Table 2. Statistical significance of differences between IOS and CIS
Total
Arches Maxillary
Variable P Arch P
Mean distance between full scans <.001a .003b
Mean distance between trimmed .064 b
.02
scans
Bold font indicates statistical significance. aOne-sample Wilcoxon rank test; null
hypothesis: median distance between IOS and CIS is zero. bOne-sample t test; null
hypothesis: mean distance between IOS and CIS is zero.
impression material exerts some pressure on the oral
mucosa. In addition, because of the difference in the
morphological and histological features of edentulous
arches, tissues offer a different resistance to pressure
exerted by the impression material and, thus, can be
displaced to a different extent in different areas.14 This is
relevant for a balanced evaluation of results provided by
in vivo studies evaluating accuracy and comparing IOS
with CIS.
The results of the present study show that, from a
methodological point of view, the differences measured
between IOS and CIS may vary if IOS and CIS are not
properly matched with regard to their surface area and
alignment. After trimming the IOS and CIS to eliminate
nonmatching peripheral areas, the results were signifi-
cantly different: the mean distance between the IOS and
CIS was reduced by almost 10-fold (−0.19 ±0.18 mm to
−0.02 ±0.05 mm).
The size of the measured mean distance (−0.02 ±0.05
mm) between the IOS and CIS falls within the range of
mucosa resilience reported by Picton and Wills.15 The
minus sign means that the CIS surface is, on average,
located in a more apical position than the IOS, which is
explained by the compression of tissues. The 3D differ-
ence between the IOS and CIS can be related to the
different physics behind the 2 impression methods and
not to the error or defects in accuracy of one method
compared with those of the other. What has been
measured can be attributed to mucosal resilience (the
amount of mucosa displacement under the pressure
applied by the impression material) and is not statistically
significant. This is consistent with another clinical study16
showing that, for the maxillary edentulous arch, IOS and
impressions made with other conventional materials had
similar deviations and may yield clinically acceptable
results.
CONCLUSIONS
Based on the findings of this clinical study, the following
conclusions were drawn:
1. The different physics behind CIS (compression and
tissue displacement) and IOS (no compression at
Lo Russo et al
5
all, with no tissue displacement) makes the com-
parison between them inappropriate from a meth-
Mandibular
Arch P odological point of view; thus, any conclusion
.005b stating which is more precise should be regarded
b
.4b with caution. They are different approaches
(mucostatic versus mucocompressive).
2. The difference between CIS and IOS (-0.02 mm)
was not statistically significantly different and was
not clinically significant for removable denture
fabrication.
REFERENCES
1. Bilgin M, Baytaroglu E, Erdem A, Dilber E. A review of computer-aided
design/computer-aided manufacture techniques for removable denture
fabrication. Eur J Dent 2016;10:286-91.
2. Steinmassl PA, Klaunzer F, Steinmassl O, Dumfahrt H, Grunert I. Evaluation
of currently available CAD/CAM denture systems. Int J Prosthodont 2017;30:
116-22.
3. Chochlidakis KM, Papaspyridakos P, Geminiani A, Chen CJ, Feng IJ,
Ercoli C. Digital versus conventional impressions for fixed prosthodon-
tics: A systematic review and meta-analysis. J Prosthet Dent 2016;116:
184-90.e12.
4. Quaas S, Rudolph H, Luthardt RG. Direct mechanical data acquisition of
dental impressions for the manufacturing of CAD/CAM restorations. J Dent
2007;35:903-8.
5. Patzelt SB, Vonau S, Stampf S, Att W. Assessing the feasibility and accuracy
of digitizing edentulous jaws. J Am Dent Assoc 2013;144:914-20.
6. Gan N, Xiong Y, Jiao T. Accuracy of intraoral digital impressions for whole
upper jaws, including full dentitions and palatal soft tissues. PLoS One
2016;11:e0158800.
7. Ender A, Attin T, Mehl A. In vivo precision of conventional and digital
methods of obtaining complete-arch dental impressions. J Prosthet Dent
2016;115:313-20.
8. Kuhr F, Schmidt A, Rehmann P, Wostmann B. A new method for assessing
the accuracy of full arch impressions in patients. J Dent 2016;55:68-74.
9. Zhang F, Suh KJ, Lee KM. Validity of intraoral scans compared with plaster
models: an in-vivo comparison of dental measurements and 3D surface
analysis. PLoS One 2016;11:e0157713.
10. 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.
11. Hayama H, Fueki K, Wadachi J, Wakabayashi N. Trueness and precision of
digital impressions obtained using an intraoral scanner with different head
size in the partially edentulous mandible. J Prosthodont Res 2018;62:
347-52.
12. Lo Russo L, Salamini A. Single-arch digital removable complete denture: A
workflow that starts from the intraoral scan. J Prosthet Dent 2018;120:20-4.
13. Razek MK. Assessment of tissue conditioning materials for functional im-
pressions. J Prosthet Dent 1979;42:376-80.
14. Chang Y, Maeda Y, Wada M, Gonda T, Ikebe K. Influence of mandibular
residual ridge morphology on pressure distribution during impression pro-
cedures: a model experiment. Int J Prosthodont 2018;31:370-4.
15. Picton DC, Wills DJ. Viscoelastic properties of the periodontal ligament and
mucous membrane. J Prosthet Dent 1978;40:263-72.
16. Chebib N, Kalberer N, Srinivasan M, Maniewicz S, Perneger T, Muller F.
Edentulous jaw impression techniques: An in vivo comparison of trueness.
J Prosthet Dent 2019;121:623-30.
Corresponding author:
Prof Lucio Lo Russo
c/o Salus Oris srl
Via Serro D’Annunzio, 20
83050 Vallesaccarda (AV)
ITALY
Email: lucio.lorusso@unifg.it
Copyright © 2019 by the Editorial Council for The Journal of Prosthetic Dentistry.
https://doi.org/10.1016/j.prosdent.2019.04.004
THE JOURNAL OF PROSTHETIC DENTISTRY