JPOR-313; No.
of Pages 7
journal of prosthodontic research xxx (2016) xxx–xxx
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: www.elsevier.com/locate/jpor
Original article
Digital assessment of preliminary impression
accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and
working casts
Takashi Matsuda DDSa, Takaharu Goto DDS, PhDa,*,
Kosuke Kurahashi DDSa, Toshiya Kashiwabara DDS, PhDa,
Megumi Watanabe DDS, PhDa, Yoritoki Tomotake DDS, PhDb,
Kan Nagao DDS, PhDa, Tetsuo Ichikawa DDS, PhDa
a
Department of Oral and Maxillofacial Prosthodontics, Institute of Biomedical Sciences, Tokushima University
Graduate School, Japan
b
Oral Implant Center, Tokushima University Hospital, Japan
article info abstract
Article history: Purpose: The aim of this study was to compare 3-dimensional surfaces of study and working
Received 31 July 2015 casts for edentulous jaws and to evaluate the accuracy of preliminary impressions with a
Received in revised form view to the future application of digital dentistry for edentulous jaws.
17 December 2015 Methods: Forty edentulous volunteers were serially recruited. Nine dentists took prelimi-
Accepted 24 December 2015 nary and final impressions in a routine clinical work-up. The study and working casts were
Available online xxx digitized using a dental 3-dimensional scanner. The two surface images were superimposed
through a least-square algorithm using imaging software and compared qualitatively.
Keywords: Furthermore, the surface of each jaw was divided into 6 sections, and the difference
Digital dentistry between the 2 images was quantitatively evaluated.
Complete denture Results: Overall inspection showed that the difference around residual ridges was small and
Edentulous that around borders were large. The mean differences in the upper and lower jaws were
Impression 0.26 mm and 0.45 mm, respectively. The maximum values of the differences showed that
Scanner the upward change mainly occurred in the anterior residual ridge, and the downward
change mainly in the posterior border seal, and the labial and buccal vestibules, whereas
every border of final impression was shortened in the lower jaw. The accuracy in all areas
except the border, which forms the foundation, was estimated to be less than 0.25 mm.
Conclusion: Using digital technology, we here showed the overall and sectional accuracy of the
preliminary impression for edentulous jaws. In our clinic, preliminary impressions have been
made using an alginate material while ensuring that the requisite impression area was covered.
# 2016 Japan Prosthodontic Society. Published by Elsevier Ltd. All rights reserved.
* Corresponding author at: Department of Oral and Maxillofacial Prosthodontics, Institute of Biomedical Sciences, Tokushima University
Graduate School, 3-18-15 Kuramoto-cho, Tokushima 770-8504, Japan. Tel.: +81 88 633 7347; fax: +81 88 633 7461.
E-mail address: [email protected] (T. Goto).
http://dx.doi.org/10.1016/j.jpor.2015.12.007
1883-1958/# 2016 Japan Prosthodontic Society. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
2 journal of prosthodontic research xxx (2016) xxx–xxx
modified slightly using wax, according to the individual
1. Introduction dentist’s judgment, and as little pressure as possible was
applied while taking the impression. A study cast was
Impression taking is the first process in the fabrication of fabricated according to the manufacturer’s instructions, using
complete dentures and is important for achieving retention, a dental plaster (Zostone, Shimomura Gypsum Co., Ltd.,
support, and stability of complete dentures [1]. Conventional- Saitama, Japan). This study cast was used to make a
ly, after the preliminary impression is taken with a ready- customized acrylic tray in a conventional manner, with
made tray and the study cast is fabricated, a final impression is blockout and relief in the necessary parts and without a
made with a customized tray. The final impression can also be spacer. The final impression was made using this tray and a
made by bite registration or using a trial wax denture as a silicone material (Examixfine Regular, GC Co., Tokyo, Japan)
‘‘bite-seating impression’’. after muscle trimming. A working cast was fabricated in a
Although impression accuracy has often been discussed for same way as for the study cast.
crown and bridge fabrication, there is little information on the The study and working casts were digitized using a dental
accuracy of impressions for edentulous jaws, except in terms 3-dimensional scanner (Dental Wings 7Series, Dental Wings
of the influence of various impression methods on the outline Inc., Montreal, Canada) so that all of the impression surface
of the denture foundation [2], as the mucosa freely changes was covered. This device provides five-axis processing and a
under various types of compressions while taking an impres- nominal resolution of 15 mm. The error value of the
sion. The quality of complete dentures fabricated using two 3-dimensional scanner was determined to be within
different techniques, i.e., traditional impression involving 15 mm by the manufacturer through the repeated measure-
taking a preliminary impression using a stock tray and a final ment of the shape and superimposition. This error value was
impression using a customized tray, vs. and a single confirmed using a similar approach in our experiment. The
impression taken with a stock tray, has been compared in accuracy of this scanner was thus enough to evaluate the
terms of patient satisfaction [3,4]. However, the differences in impression accuracy of the edentulous ridge. Three-dimen-
the 3-dimensional surfaces between impressions taken with sional scanning was done in a constant environment: at a
these two approaches have been not investigated. temperature of (25 3) 8C, humidity of (60 10) %, and
Recently, the development of a scanner has made it (1014 10) hPa atmosphere. The scanner was calibrated with
possible to take a tooth impression for crown and bridge the aluminum calibration plate (125 mm 10 mm 100 mm,
fabrication, and it will soon be possible to take a direct digital Dental Wings Inc., Montreal, Canada) according to the
impression of edentulous jaws. This scanner also allows easy manufacturer’s instruction. Parts other than the impression
measurement of 3-dimensional surfaces and quantitative surface in the 3-dimensional images were trimmed and
evaluation. edited on the monitor by a single operator and a single
In this study, the shapes of study and working casts were inspector. The two surface images of the study and working
quantitatively evaluated and compared using a 3-dimentional casts were superimposed through a least-square algorithm
scanner and specialized software, in order to clarify the using imaging software (Gom Inspect V7 SR2, GOM mbH,
accuracy of the preliminary impression, with a view to future Braunschweig, Germany), such that the surface image of the
taking of direct digital impressions for edentulous jaws. working cast approximately matched that of the study cast
with the least-square error, and were examined qualitatively
on the screen.
2. Materials and methods Furthermore, the surface of each jaw was divided into six
sections according to the functional significance reported by
This research was conducted with the approval of the Ethics Boucher [5,6], and the differences between the two images
Committee of the Tokushima University Hospital (No. 1475). were quantitatively evaluated for each of these areas. The six
Forty edentulous volunteers (35 upper jaw cases and 30 sections in each jaw were defined as shown in Fig. 1. The
lower jaw cases; mean age: 75 7 y) were investigated as boundary of the sections was determined through inflection
subjects in this study. All of these individuals visited the points by agreement between two examiners. A downward
Prosthodontic Department of the Tokushima University direction, which indicates that the surface of the working cast
Hospital for fabrication of new complete dentures from April, was settled relative to that of the study cast, was defined as
2013 to March, 2015, and for whom informed consent for ‘‘plus’’; conversely, the upward direction, which indicates that
participation in this study was obtained from both the patients the surface of working cast was lifted relative to that of the
and attending dentists. Individuals with symptoms of stoma- study cast, was defined as ‘‘minus’’. The maximum values of
tognathic disorder, ulcers, and mucosal abnormalities were the difference in upward and downward directions (minimum
excluded. The 9 dentists attending to the 40 edentulous value for the downward direction), and the distribution ratio of
volunteers had clinical experience from 1 to 31 years (mean three categories (less than the absolute difference of 0.25 mm,
age: 39 11 y), and three of them were board certified between 0.25 and 0.5 mm, and more than 0.5 mm) were
prosthodontists. quantitatively examined for each section.
Preliminary impressions were taken using a ready-made The shape of residual ridge in each jaw was classified
tray and an alginate impression material (ALGINoplast EM according to the treatment difficulty indices developed by the
normal, Heraeus Kulzer Japan Co., Ltd., Tokyo, Japan). Dentists Japan Prosthodontic Society [7], and the differences between
were required to use requisite anatomical landmarks for the two images were also evaluated for classification of
complete denture fabrication: the ready-made tray could be residual ridge shapes.
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
journal of prosthodontic research xxx (2016) xxx–xxx 3
Fig. 1 – The six sections that were compared in the 3-dimensional surfaces between the study and master casts.
Fig. 2 – Superimposition of images of the study and working casts.
2.1. Statistical analysis red, and blue colors refer to the coincidence between the two
images, the lower part of the working cast relative to the study
The Mann–Whitney U-test and Bonferroni’s post hoc test was cast, and the upper part relative to the study cast, respectively.
used for statistical analysis, and was conducted using SPSS1 Overall inspection showed that the difference between the
version 22.0 (SPSS Co., Chicago, IL, USA). Significance was cases was the smallest in the median palatine raphe; the
accepted at p < 0.05. difference increased in the order: posterior residual ridge,
anterior residual ridge, secondary retentive and stress-bearing
areas of the palate, posterior palatal seal, and labial and buccal
3. Results vestibules. Green to yellow colors were largely found around
the posterior residual ridge, as the primary stress-bearing
Fig. 2 shows typical images superimposed with the area, and yellow to red colors were found around the anterior
3-dimensional images of the study and working casts. Green, residual ridge.
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
4 journal of prosthodontic research xxx (2016) xxx–xxx
vestibules, while the opposite was true for the <0.25 mm
category.
No significant relationship was found between years of
clinical experience of the practitioner and the mean difference
between the casts in each jaw. Residual ridges were classified
into 4 types in vertical and buccolingual dimensions, as shown
in Table 1. Most cases were categorized as Level I (high and
U-shaped) and Level II (moderate and between U- and
V-shaped). No significant relationship was found between
residual ridge types and the mean difference between the
casts.
4. Discussion
Fig. 3 – Mean differences between the study and working
models in each jaw.
Recent advances in computer and sensor technologies have
allowed technological innovations in the field of dentistry,
which is termed ‘‘digital dentistry’’. These innovations have
In the upper jaw, yellow to orange colors were largely found led to the wider commercial availability of intraoral scanners,
around the rugae as secondary retentive stress-bearing areas and the accuracy of impressions for crowns and bridges using
of the palate, and green to blue colors were mainly found this technology has been reported [8–14].
around the median palatine raphe, as the relief area. Yellow to It has also been reported that the use of digital impressions,
red colors were found around the buccal vestibules. in which the edentulous jaw is directly measured using a
In the lower jaw, the differences between the scans were digital apparatus, is currently being investigated [15]. Al-
low around the residual ridge; conversely the differences though Nokubi et al. [16] and Nagaoka et al. [17] evaluated
around the border seals were large. Green to yellow and green the 3-dimensional morphologies of die casts using digital
to orange colors were often found around the buccal shelf, as technology, the measurement capacity of the technology
the primary stress-bearing area, and the residual ridge, as the used at that time was much lower than those of current
secondary stress-bearing area, respectively. Blue color was technology.
mainly present around the retromolar pad, as the border seal The target for taking an impression of an edentulous jaw is
area. Yellow to red colors were largely found around the the mucosa, which has high displacement properties, and it is
alveolingual sulcus and the buccal vestibules. difficult to determine the denture border and to select the
Quantitatively, the mean differences of the upper and degree of impression pressure, making it difficult to evaluate
lower jaws were 0.26 mm and 0.45 mm, respectively; that of the true values of an impression. Digital impression of
the lower jaw was thus significantly larger (Fig. 3). Figs. 4 and 5 edentulous jaws is expected to be developed soon, but the
show the differences in the six sections of each jaw in terms of accuracy and quality of such impressions remain unclear.
the maximum and actual values of difference, and distribution During the process of taking the final impressions using
ratio of the three difference categories: less than 0.25 mm, custom trays, the custom tray is manually adapted to the
between 0.25 and 0.5 mm, and more than 0.5 mm. Comparing residual ridge as much as possible. In practice, this clinical
the maximum values in the downward and upward direction, procedure is the same approach as the best-fit algorithm used
the location of the working cast relative to that of the study in this study, which entails a numerical superimposition of 2
cast can be estimated. Judging from the maximum values in images through a least-square algorithm. We believe it will be
the upper jaw, upward changes were often found in the possible to predict the standard accuracy of a preliminary
anterior residual ridge, while downward changes were seen in impression for edentulous jaws by comparing the 3-dimen-
the posterior border seal and labial and buccal vestibules. The sional surfaces between study and working casts, although
changes in the maximum values in the upward and downward quantitative values need to be evaluated.
directions were equal when compared to those in the lower The degree of coincidence between 2 impressions was
jaw. The distribution ratios suggested that the >0.5 mm analyzed in each of the 6 sections into which each jaw was
category was found significantly more often in the buccal divided, according to the report by Boucher, which represent
vestibule and posterior border seal, and the <0.25 mm acceptable functional and anatomical criteria. The results of
category was more often applicable in other sections. our superimposition study suggested that the degree of
Judging from the maximum values in the lower jaw, every coincidence was high, whereas the degree of coincidence at
border in the final impression was reduced. The distribution the border area was low. In taking a preliminary impression
ratios suggested that the differences in the residual ridge and with alginate impression material, we have attempted to take
retromolar pad were small; conversely, these were larger in the impression in such a way as to include all the anatomical
the labial vestibule and alveolingual sulcus. The distribution landmarks, rather than to determine the denture border. In
ratios in the lower jaw was more evenly spread compared to taking a final impression using the customized tray, the border
those in the upper jaw. The >0.5 mm category was seldom is set by border molding, termed ‘‘muscle trimming,’’ as there
found in the residual ridge and retromolar pad, and was a large difference was found in this region. It has been reported
frequently present in the alveolingual sulcus and buccal that the border in a compound impression of upper jaws is
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
journal of prosthodontic research xxx (2016) xxx–xxx 5
Fig. 4 – Differences between the upper study and working casts. Maximum difference and distribution ratio of three
difference ranges in the six sections compared.
similar to that in an alginate impression, whereas the border differences in all areas except the border should be the same
in a compound impression of lower jaws is longer than that in as that in the upper jaw.
an alginate impression [18]. This study showed the opposite, The residual ridge shape and clinical experience of the
particularly for the lower jaw, in which the border in the final practitioner did not have a marked effect on the difference in
impression was shorter than that in the preliminary impres- values between study and working casts. If these factors
sion. The difference in these findings may be because the influenced both casts equally, their influence might not have
preliminary impressions made using alginate impression been detected in this study. In the criteria of patient selection,
material in our clinic ensured that the requisite impression the condition of patient visits had priority over the dentist
area was covered. selection. As a result, the 9 dentists who had clinical
The mean difference for the upper jaw was 0.26 mm, but experience from 1 to 31 years were selected with a wide
the difference in all areas except the border, which forms the range of ages and clinical experiences. Although it is reported
foundation of the upper jaw, should be less than 0.25 mm. that the amount of tooth reduction and operation time depend
There was a large difference at the posterior section of the on clinical experience of the practitioner with respect to the
palate, which is associated with denture retention through abutment tooth preparation [19], the result of this study, in
the border seal. The location of the soft palate during the which no relationship between the clinical experience of
impression may affect this difference. The mean difference in practitioners and difference of two casts was found, may be
the lower jaw was 0.45 mm. The maximum values and characterized as the impression of edentulous mucosa.
distribution ratio of differences in the lower jaws were larger However, considering the relative numbers of cases and
than those in upper jaws. This may be because the residual practitioners, the evaluation on the influence of clinical
ridge and palate in the upper jaw involves more flat areas, experience of practitioners may not be always sufficient.
whereas the residual ridge in the lower jaws involve more Further discussion is needed with more case number and
sloped areas and the denture border is longer. However, the standardization of practitioner’s experience or skills.
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
6 journal of prosthodontic research xxx (2016) xxx–xxx
Fig. 5 – Differences between the lower study and working casts. Maximum difference and distribution ratio of three
difference ranges in the six sections compared.
Table 1 – Classification of residual ridge shapes and the mean differences between the study and working models.
Classification of residual ridge shapes (vertical dimension)
Level I Level II Level III Level IV
Maxilla 14 20 1 0
(n = 35) 0.25 0.08 mm 0.28 0.16 mm 0.21 mm –
Mandible (n = 30) 4 17 0 9
0.20 0.03 mm 0.52 0.43 mm – 0.43 0.22 mm
Classification of residual ridge shapes (buccolingual dimension)
Level I Level II Level III Level IV
Maxilla 13 21 1 0
(n = 35) 0.27 0.10 mm 0.26 0.16 mm 0.25 mm –
Mandible (n = 30) 7 19 3 1
0.46 0.46 mm 0.39 0.35 mm 0.88 0.02 mm 0.19 mm
Upper row: matched number; lower row: difference (mean SD).
technology. The accuracy of the preliminary impression in all
5. Conclusion areas except the border area, which forms the foundation on
the mucosa, was estimated to be less than 0.25 mm in the
This study investigated the overall and sectional accuracy of upper jaw. In the lower jaws, the accuracy of all areas other
preliminary impressions of edentulous jaws using digital than the border was the same. The differences in the denture
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
JPOR-313; No. of Pages 7
journal of prosthodontic research xxx (2016) xxx–xxx 7
border in every section were characterized and preliminary the reliability of the treatment difficulty indices developed
impression using an alginate material was made in our clinic, by the Japan Prosthodontic Society. J Prosthodont Res
2012;56:71–86.
ensuring that the requisite impression area was covered.
[8] Kuhn K, Ostertag S, Ostertag M, Walter MH, Luthardt RG,
In order for digital direct impressions of edentulous jaws to
Rudolph H. Comparison of an analog and digital
become practically implemented in future, the accuracy of the quantitative and qualitative analysis for the fit of dental
new procedure will have to be compared with the conven- copings. Comput Biol Med 2015;57:32–41.
tional procedure. The quantitative differences indicated in [9] Grünheid T, McCarthy SD, Larson BE. Clinical use of a direct
this study will be significant for validating the accuracy of chairside oral scanner: an assessment of accuracy, time,
complete denture fabrications. Patients and clinicians’ condi- and patient acceptance. Am J Orthod Dentofacial Orthop
2014;146:673–82.
tion, including edentulous alveolar ridges, will be categorized
[10] Flügge TV, Schlager S, Nelson K, Nahles S, Metzger MC.
and digital assessment of edentulous jaws will be established Precision of intraoral digital dental impressions with iTero
through the quantitative difference of more clinical cases in and extraoral digitization with the iTero and a model
future research. scanner. Am J Orthod Dentofacial Orthop 2013;144:471–8.
[11] Patzelt SB, Bishti S, Stampf S, Att W. Accuracy of computer-
aided design/computer-aided manufacturing-generated
Conflict of interest dental casts based on intraoral scanner data. J Am Dent
Assoc 2014;145:1133–40.
[12] Ender A, Mehl A. Accuracy of complete-arch dental
The authors have no conflicts of interest to disclose. impressions: a new method of measuring trueness and
precision. J Prosthet Dent 2013;109:121–8.
[13] Mino T, Maekawa K, Ueda A, Higuchi S, Sejima J, Takeuchi
references T, et al. In silico comparison of the reproducibility of full-
arch implant provisional restorations to final restoration
between a 3D Scan/CAD/CAM technique and the
[1] Carlsson GE. Critical review of some dogmas in conventional method. J Prosthodont Res 2014;59:152–8.
prosthodontics. J Prosthodont Res 2009;53:3–10. [14] Su TS, Sun J. Comparison of repeatability between intraoral
[2] Nishigawa G, Minagi S, Sato T. A study of the area of digital scanner and extraoral digital scanner: an in-vitro
preliminary impression surface of edentulous arch in study. J Prosthodont Res 2015;59:236–242.
relation to impression techniques. J Jpn Prosthodont Soc [15] Patzelt SB, Vonau S, Stampf S, Att W. Assessing the
1992;36:834–41. feasibility and accuracy of digitizing edentulous jaws. J Am
[3] Kawai Y, Murakami H, Shariati B, Klemetti E, Blomfield JV, Dent Assoc 2013;144:914–20.
Billette L, et al. Do traditional techniques produce better [16] Kawabata N, Ono H, Harada O, Tanaka H, Koreeda Y, Saitoh
conventional complete dentures than simplified F, et al. Three-dimensional reconstruction of the cast of
techniques. J Dent 2005;33:659–68. edentulous jaw measuring a shape. J Jpn Prosthodont Soc
[4] Yuzbasioglu E, Kurt H, Turunc R, Bilir H. Comparison of 1994;38:1015–24.
digital and conventional impression techniques: evaluation [17] Kibi M, Ono T, Morimitsu T, Watanabe K, Namba H, Nokubi
of patients’ perception, treatment comfort, effectiveness T. Department of three dimensional shape measurement
and clinical outcomes. BMC Oral Health 2014;14:10. system for residual ridges. J Jpn Prosthodont Soc
[5] Hickey JC, Zarb GA. Rehabilitation of edentulous patients. 1994;38:17–24.
In: Hickey JC, Zarb GA, editors. Boucher’s prosthodontic [18] Sato T, Nishigawa G, Minagi S, Nojima T. Modeling
treatment for edentulous patients. 8th ed. St. Louis: Mosby; compound as a preliminary impression material for
1980. p. 115–487. mandibular edentulous arch: the reasons for the choice.
[6] Boucher CO. Impression building recognizes the Dent Outlook 1993;82:379–89.
importance of mouth anatomy as the determining factor in [19] Noguchi S, Aoki J, Nakamura Y, Kurike H, Hanamura N.
denture form. Dent Surv 1946;22:2062–8. Convergence angle of the abutment preparation on full cast
[7] Kuboki T, Ichikawa T, Baba K, Hideshima M, Sato Y, Wake crowns: Part 2. Difference of experience. J Jpn Prosthodont
H, et al. A multi-centered epidemiological study evaluating Soc 1982;26:710–20.
Please cite this article in press as: Matsuda T, et al. Digital assessment of preliminary impression accuracy for edentulous jaws: Comparisons of
3-dimensional surfaces between study and working casts. J Prosthodont Res (2016), http://dx.doi.org/10.1016/j.jpor.2015.12.007
