Professional Documents
Culture Documents
Precision and Accuracy of A Digital Impression Scanner in Full-Arch Implant Rehabilitation
Precision and Accuracy of A Digital Impression Scanner in Full-Arch Implant Rehabilitation
Purpose: To evaluate the accuracy and precision of a digital scanner used to scan four
implants positioned according to an immediate loading implant protocol and to assess the
accuracy of an aluminum framework fabricated from a digital impression. Materials and
Methods: Five master casts reproducing different edentulous maxillae with four tilted implants
were used. Four scan bodies were screwed onto the low-profile abutments, and a digital
intraoral scanner was used to perform five digital impressions of each master cast. To assess
trueness, a metal framework of the best digital impression was produced with computer-
aided design/computer-assisted manufacture (CAD/CAM) technology and passive fit was
assessed with the Sheffield test. Gaps between the frameworks and the implant analogs were
measured with a stereomicroscope. To assess precision, three-dimensional (3D) point cloud
processing software was used to measure the deviations between the five digital impressions
of each cast by producing a color map. The deviation values were grouped in three classes,
and differences were assessed between class 2 (representing lower discrepancies) and the
assembled classes 1 and 3 (representing the higher negative and positive discrepancies,
respectively). Results: The frameworks showed a mean gap of < 30 μm (range: 2 to 47 μm).
A statistically significant difference was found between the two groups by the 3D point cloud
software, with higher frequencies of points in class 2 than in grouped classes 1 and 3 (P < .001).
Conclusion: Within the limits of this in vitro study, it appears that a digital impression may
represent a reliable method for fabricating full-arch implant frameworks with good passive fit
when tilted implants are present. Int J Prosthodont 2018;31:171–175. doi: 10.11607/ijp.5535
Fig 1 The five casts used for the in vitro investigation and the aluminum frameworks realized on the base
of the most complete impression.
A previous in vitro study by the same team of au- Materials and Methods
thors6 compared traditional and digital impression
techniques using a simplified model with multiple im- Five master casts made with plaster type 4 (Fujirock
plants. The intraoral scanner appeared to be a viable EP, GC) were poured, reproducing five different eden-
alternative to traditional impressions, using different tulous maxillae with four tilted implants rehabilitated
materials and techniques (open-tray, closed-tray, and according to the Columbus Bridge Protocol9 (Fig 1).
splinting of impression copings with resin). Similar re- The implants presented different inclinations and dif-
sults were obtained by another group of researchers, ferent inter-implant distances (range: 13 to 27 mm).
who concluded that digital full-arch implant impres- Four implant low-profile analogs (diameter 4 mm;
sions were significantly more accurate than conven- low-profile abutment non-hexed temporary cylinder,
tional impressions.7 Biomet3i) were used (Fig 1). Four prototype scan
In 2015, Ender and Mehl8 introduced the terms bodies designed by Simbiosi were screwed onto the
“trueness” and “precision” as different measures of low-profile abutments. The scan bodies (Fig 2) had
accuracy. Trueness is defined as the comparison be- a cylindrical body with semi-spherical extremities
tween a control STL dataset and a test STL dataset. aimed at a more precise digital repositioning of the
Precision is defined as a comparison between differ- scan bodies themselves. They have been specifically
ent datasets obtained using the same digital scanner. designed for edentulous patients because they are
In light of the previous outcomes and definitions, the shorter than the usual scan bodies to allow for easier
aims of the present study were (1) to assess the preci- impression taking, even at the distal implants.
sion of a dental scanner used to scan models repro- An intraoral scanner system (True Definition Scanner,
ducing real clinical situations including tilted implants 3M ESPE) was used to perform five digital impressions
using prototype scan bodies; and (2) to assess the fit of each master cast (total of 25 digital impressions). Five
of a metal framework produced from a digital impres- different operators who had never used an intraoral
sion with computer-aided design/computer-assisted scanner before were selected and participated in a day
manufacture (CAD/CAM) technology. of training with the digitizing system before the start of
Fig 2 Specially designed scan bodies. Fig 3 Digital reproduction of the model and color scale, indicating the
value of discrepancy compared to the control impression (the red color
indicates extreme positive values and the blue color indicates extreme
negative values; green points were the most accurate).
the study. The training session undertaken was com- Table 1 D
iscrepancy Values Subdivided into Three
prised of a 4-hour explanation of the system and its Classes, with Class 2 Representing Smaller
application followed by 4 hours of practice using a Deviation from the Reference Impression
test cast. The clinicians complied with a defined pro- Class Class start (μm) Class end (μm)
tocol: before taking the impression, four scan bodies 1 –0.099895410240 –0.024934630841
were screwed onto the implant analogs; subsequent- 2 –0.024934630841 0.025039222091
ly, the cast was powdered with a homogenous layer
of dust (3M High-Resolution Scanning Spray, 3M 3 0.025039222091 0.100000001490
ESPE) for scanning procedures. The scanning proce-
dure started from implant 26 and proceeded in con-
tinuous mode, making circular movements, around
all the scan bodies. This differed in comparison to the between the other framework-analog interfaces. If the fit
technique described by Gimenez et al,7 allowing a is not sufficient, the superstructure will be lifted when a
faster digital impression process. The entire digitizing screw is tightened, creating a gap at the level of one or
process had to be performed in less than 7 minutes. more abutment analogs.10,11
At the end of the impression procedure, the operat- A stereomicroscope [Wild M3Z, Wild Heerbrugg] with
ing system confirmed the suitability of the scan or ×40 magnification was used to record the value of gaps
highlighted the need to repeat the scan in order to at the framework-abutment analog interfaces (on the
acquire missing or incorrectly digitized areas. vestibular side). Eight measurements were recorded for
each metal framework: four screwing the framework at
Fit Assessment implant site 2.6, and another four screwing the frame-
work at implant site 1.6.
Sheffield Test. For each master cast, the best digi-
tal impression was identified according to three pa- Precision Assessment
rameters: (1) presence of scattering; (2) complete
at the scan body level; and (3) file dimension (to Superimposition Test. Impressions taken with the
facilitate the software in the repositioning process). True Definition scanner were imported with a 3D
This impression was used as a control impression point cloud (and triangular mesh) processing software
for evaluating both trueness and precision. The (Cloudcompare, Danielgm.net). All five digital impres-
control impressions were used to fabricate five alu- sions for each model were analyzed to perform com-
minum frameworks with CAD/CAM technology in parisons between 3D point clouds. The best selected
a specialized laboratory (Simbiosi). Passive fit was impression for each model, as described above, was
assessed with the Sheffield test, screwing each compared with the other four impressions. The soft-
framework onto the corresponding master cast. ware produced a color map demonstrating the size of
According to this test, a framework is considered the deviations between the control impression and the
well fitting if it is possible to tighten a single screw other impressions (Fig 3). Values of discrepancies were
on a distal model abutment without creating a gap grouped in three classes (Table 1).
Table 2 S
heffield Test Findings for Mean, Standard use of an intraoral scanner represents a viable
Deviation (SD), Minimum (Min), and Maximum alternative to traditional impression materials for
(Max) Gap Values for the Five Models
the fabrication of full-arch, implant-supported
Framework Mean ± SD (μm) Min (μm) Max (μm) prostheses provided with a satisfactory passive
Model 1 24 ± 19 3 44 fit. In particular, comparing the impression with a
Model 2 22 ± 14 3 47 reference image obtained with a 3D Coordinate
Model 3 27 ± 15 3 45 Measurement Machine (CMM), they found that
Model 4 21 ± 12 3 37
the digital impression performed the best, while
traditional techniques revealed a greater variability
Model 5 21 ± 16 2 46
in the results. It seems that traditional techniques
presented a greater risk of error. The purpose of
the present study was therefore to assess in vitro
Table 3 O
verall Differences Between Class 2 and
the precision of digital implant impressions of com-
Grouped Classes 1 and 3 plete-arch restorations with tilted implants using
Group Mean SD Median Minimum Maximum
prototype scan bodies and a simplified methodol-
ogy compared to previous studies.
Class 2 33,459.4 4,488.5 32,041.5 28,931.0 41,904.0
According to the present results, all the frame-
Classes 3,587.4 3,512.5 2,550.5 167.0 11,141.0 works presented a vestibular gap of < 47 μm as
1 and 3
evaluated in the Sheffield test. Jemt and Lie14 de-
P ≤ .001
fined the passive fit as an accuracy level that does
not cause any long-term clinical complications and
discrepancies of up to 150 μm as acceptable. In
addition, the superimposition test showed great
Statistical Analysis precision of the five impressions performed by five
different clinicians for each cast: The higher fre-
A statistical analysis was performed comparing the quencies of points were found in class 2, represent-
number of points included in each class of discrepancy. ing a gap between –0.025 and 0.025 μm. Moreover,
Descriptive analysis was reported as mean and stan- it must be underlined that points in classes 1 and
dard deviation (SD) and median with range (minimum 3 were not at the scan body level, but only in the
to maximum). gingival area; in fact, the greatest differences be-
Statistical tests were performed using SPSS v 20 tween the five impressions were at gingival level,
(IBM). Class 2 (representing lower discrepancies) fre- and some of the clinicians did not take a complete
quencies were compared with the other assembled impression of all the gingiva. On the contrary, all
groups (1 and 3) using nonparametric Mann-Whitney U clinicians focused their attention at the scan body
test. P ≤ .05 was considered statistically significant. level, which is the most important part for the real-
ization of the framework.
Results In the present study, the fits of milled frame-
works were tested in a more clinical way compared
The findings of the Sheffield test are reported in Table 2. to other papers in which a coordinate-measuring
The frameworks showed a mean gap value < 30 μm machine (CMM) was used.7 Clinicians cannot use
(range: 2 to 47 μm). a CMM inside the mouth to test trueness; the only
A statistically significant difference regarding preci- exam clinically possible to test framework precision
sion was found between the two groups, with higher fre- is the Sheffield test. The results of the Sheffield test
quencies of points in class 2 (lower discrepancies) than in the present study showed that with a full digital
in the other classes (1 and 3) (P < .001) (Table 3). process (digital impression, best-fit repositioning,
milled framework), the same precision as with the
Discussion classic impression technique can be obtained but
in a faster way, even in the challenging situation
The accuracy and rigidity of prosthodontic frameworks reproduced in the present study, where implants
have been reported to be fundamental prerequisites for with different angulations were used in full-arch
the predictable osseointegration of dental implants for rehabilitations.
immediate loading.12,13 Given the increasing interest in these devices,
In a previous study, Menini et al6 evaluated the ac- several studies have been published in recent
curacy of different impression techniques (traditional years testing the accuracy of different intra-
vs digital) on multiple implants. They reported that the oral scanners. However, only a few articles have