Statement of problem
Although studies have reported the trueness and precision of intraoral scanners (IOSs), studies addressing the accuracy of IOSs in reproducing inlay preparations are lacking.
The purpose of this in vitro study was to compare the accuracy of representative IOSs in obtaining digital scans of inlay preparations and to evaluate whether the IOSs had sufficient depth of field to obtain accurate images of narrow and deep cavity preparations.
Material and methods
Digital scans of a bimaxillary typodont with cavity preparations for inlay restorations on the maxillary first premolar, first and second molar, mandibular second premolar, and first molar were obtained using 6 IOSs (CEREC Omnicam, E4D, FastScan, iTero, TRIOS, and Zfx IntraScan). Standard tessellation language (STL) data sets were analyzed using the 3-dimensional analysis software (Geomagic Verify). Color-coded maps were used to compare the magnitude and pattern of general deviation of the IOSs with those of a reference scan. Each tooth prepared for inlay restoration was digitally cut out, and the trueness and precision of each IOS were measured using the superimposition technique. Statistical analyses were conducted using statistical software (α=.05).
The trueness values were lowest with the FastScan (22.1 μm), followed by TRIOS (22.7 μm), CEREC Omnicam (23.2 μm), iTero (26.8 μm), Zfx IntraScan (36.4 μm), and E4D (46.2 μm). In general, the digital scans of more complicated cavity design showed more deviation. Color-coded maps showed positive vertical discrepancy with the E4D and negative vertical discrepancy with the Zfx IntraScan, especially on the cavity floor. Regarding precision, the highest value was observed in the E4D (37.7 μm), while the lowest value was observed with the TRIOS (7.0 μm). However, no significant difference was found between teeth with different inlay preparations. Scanning errors were more frequently seen in the cervical area.
Different IOSs and types of cavity design influenced the accuracy of the digital scans. Scans of more complex cavity geometry generally showed higher deviation. The E4D exhibited the most deviation in both trueness and precision, followed by the Zfx IntraScan. The E4D and Zfx IntraScan appeared to have less depth of field than the others to obtain digital scans for inlay preparation with different heights.
Inaccurate digital scans may contribute to clinically relevant discrepancies between the tooth and definitive restoration. Clinicians should be aware that the accuracy of digital scans varies, depending on the tooth preparation and type of IOS used.
Computer-aided design and computer-aided manufacturing (CAD-CAM) technology has revolutionized workflow in dentistry because of its speed, convenience, and accuracy. Accuracy of definitive restoration fabricated in a fully digital workflow is a crucial factor determining clinical success. When the restoration fails to seat properly in the cavity, more time needs to be spent for the necessary adjustments. If the spacer parameters are set at a high value during the CAD procedure to allow ease of seating, large internal and marginal gaps occur between the tooth and restoration. Poor adaptation jeopardizes longevity in indirect restorations, as large discrepancies may lead to clinical complications such as secondary caries or periodontal inflammation due to plaque accumulation.
The accuracy of intraoral scanners (IOSs) has been described, according to the definition 5725-1 of the International Organization for Standardization by using the following terms: trueness, referring to the closeness of agreement between the arithmetic mean of a large number of test results and the true or accepted reference value and precision, referring to agreement between test results. During the fabrication of CAD-CAM restorations, an error may be introduced in the digital scanning procedure with an IOS. Some studies have suggested that digital scans should not be used to digitize complete or edentulous arches because of distortion. However, others have reported that digital scans acquired using IOSs have better accuracy than conventional silicone impressions for both single and multiple-tooth scans. Nonetheless, some clinicians remain skeptical about intraoral digital scanning, and also, the purchasing and managing cost of such digital scanning systems remains high. In the conventional indirect restoration procedure, any minor undercuts in the tooth preparation can be addressed by the dental laboratory technician. Such undercuts are processed by software in CAD-CAM restorations. Studies have shown that more accurate scanner readings can be obtained if the scanned surface is smooth and regular.
Deviations in the digital scan are smallest when the camera of the IOS is positioned perpendicular to the surface being scanned, and the magnitude of deviation increases with the degree of tilt of the camera away from the perpendicular plane. However, clinicians must manipulate IOSs at various angles and positions relative to the teeth to adequately acquire a digital scan of an area. The IOS, being an optical device, has a limited depth of field that may vary depending on the system. In addition, soft and hard tissues may cause scanning interference.
The performance of IOSs has been compared by evaluating the adaptation of the fabricated restoration, and studies have addressed the influence of tooth geometry on the accuracy of the IOSs. However, information regarding the effects of cavity design on digital scan acquisition by IOSs is sparse. The purpose of this in vitro study was to investigate the reproducibility of digital scans of various complex inlay preparations using 6 IOSs. The null hypothesis was that no difference would be found in trueness or precision among the different cavity types or among the IOSs.
Material and methods
A typodont with various inlay preparations served as the master model. Artificial teeth (A50H-Set; J. Morita Europe GmbH) that had been prepared by the manufacturer using precision processing were used to eliminate procedural errors in cavity preparation. The type and position of the cavity preparations were as follows: occlusal preparation on the maxillary first premolar, mesio-occlusal preparation on the maxillary first molar (MxFM) and maxillary second molars (MxSM), mesio-occluso-distal cavity with a lingual cusp reduction preparation on the mandibular second premolar (MnSP), and bucco-occlusal preparation on the mandibular first molar (MnFM) ( Fig. 1 ). After the ideal initial tooth arrangement was established, none of the teeth were removed from or added to the master model, and no external forces were applied during the experiment. All experiments were carried out at 23 ±1 °C and 50 ±5% relative humidity.
A desktop scanner (Sensable S3; MEDIT) was used to obtain the reference standard tessellation language (STL) data set of the master model. This scanner system had a camera resolution of 1.4 megapixels with an accuracy of ±0.01 mm. The high capacity (140×140×100 mm) of the scanner system allows scanning of complete-arch casts. In the present study, 6 IOSs were used: the CEREC Omnicam (Dentsply Sirona), E4D (D4D Technologies), FastScan (IOS Technologies), iTero (Cadent), TRIOS (3Shape), and Zfx IntraScan (Zfx GmbH) ( Table 1 ). The accuracy of the IOSs was compared in 2 categories: data capture mode, which compared individual images (E4D, FastScan, and iTero) with video sequences (CEREC Omnicam, TRIOS, and Zfx IntraScan) and data capture principle, which compared accuracy among active triangulation (CEREC Omnicam, FastScan), confocal microscopy (iTero, TRIOS, and Zfx IntraScan), and optical coherence tomography (E4D). The scanning procedure was repeated 5 times with each scanner, according to the manufacturers’ instructions.
|System||Manufacturer||Scanner Technology||Light Source||Acquisition Method||Necessity of Coating|
|CEREC Omnicam||Dentsply Sirona||Active triangulation with strip light projection||Light||Video sequence||None|
|E4D dentist (initial version)||E4D Technologies||Optical coherence tomography||Laser||Individual image||Occasional|
|FastScan||IOS Technologies, Inc||Active triangulation and Scheimpflug principle||Laser||Individual image||Yes|
|iTero (1st generation)||Align Technology Inc||Parallel confocal microscopy||Red laser||Individual image||None|
|TRIOS (2nd generation)||3Shape A/S||Confocal microscopy||Light||Video sequence||None|
|Zfx IntraScan||Zfx GmbH||Confocal microscopy and Moiree effect detection||Laser||Video sequence||None|
Inspection software (Geomagic Verify v126.96.36.199.; 3D Systems) was used to obtain the trueness and precision values of the IOSs. The digital models were modified to remove parts not relevant for measurement, leaving only the teeth with cavity preparations: maxillary first premolar, MxFM, MxSM, MnSP, and MnFM. The STL data set of each tooth from each IOS was superimposed onto the reference data set of the corresponding tooth, which was itself obtained using the reference scanner and a best-fit algorithm to evaluate the trueness of the IOS. Color-coded maps were used to observe the magnitude and pattern of deviation between the reference scanner and the digital models acquired by each IOS. To evaluate the precision of the IOSs, the STL data sets from the same scanner were superimposed.
Statistical analyses were conducted using statistical software (IBM SPSS Statistics, v20.0; IBM Corp) (α=.05). As the data were not normally distributed, as indicated by the Kolmogorov-Smirnov test, the median trueness and precision values of the IOSs, as well as the data capture principle values, were analyzed using a Kruskal-Wallis test. For the post hoc test, the dependent variable was converted into the rank variable, and 1-way ANOVA was conducted, followed by multiple comparisons of the rank variable by the Tukey honestly significant difference test.
The median trueness values were in descending order: E4D (46.2 μm), Zfx IntraScan (36.4 μm), iTero (26.8 μm), CEREC Omnicam (23.2 μm), TRIOS (22.7 μm), and FastScan (22.1 μm). The trueness values of the CEREC Omnicam, FastScan, and TRIOS did not differ significantly, but these scanners did show significantly lower trueness values than the iTero, E4D, and Zfx IntraScan, except between the CEREC Omnicam and iTero; the E4D exhibited significantly higher trueness value than the Zfx IntraScan, which showed significantly higher trueness value than the iTero ( P <.05) ( Table 2 ; Fig. 2 ). Among the cavity types, the smallest deviations were found in the MxSM with the mesio-occlusal inlay preparation (22.3 μm) and in the MnFM with the bucco-occlusal inlay preparation (25.2 μm). These values were significantly smaller than those of the MnSP with an onlay preparation (31.2 μm) ( Table 2 ; Fig. 3 ). In the comparison between data capture principles, optical coherence tomography showed significantly higher trueness value and active triangulation, significantly lower trueness value, than systems that used the confocal microscopy principle ( P <.05). Video sequence data capture showed greater trueness value than individual image data capture ( P <.05) ( Table 3 ).
|CEREC Omnicam||5||21.9 (19.2-29.0)||28.7 (28.2-29.2)||18.6 (18.4-19.9)||27.1 (26.4-32.3)||22.6 (22.1-24.0)||23.2 CD (20.5-28.7)||4||16.541||.002|
|E4D||5||50.1 (46.0-61.7)||44.1 (38.2-53.9)||50.6 (37.7-56.3)||40.2 (38.0-44.3)||46.2 (43.7-48.7)||46.2 A (42.5-51.0)||4||6.628||.157|
|FastScan||5||22.1 (20.8-27.3)||21.6 (19.9-24.5)||18.7 (16.4-23.4)||24.5 (22.1-25.4)||24.8 (21.2-25.2)||22.1 D (20.5-25.1)||4||5.271||.261|
|iTero||5||25.3 (21.5-28.6)||24.4 (22.8-30.9)||23.6 (20.1-29.8)||34.9 (34.6-35.4)||30.6 (26.9-32.9)||26.8 C (23.5-34.3)||4||12.374||.015|
|TRIOS||5||20.3 (18.5-21.0)||24.4 (23.5-25.0)||14.3 (13.7-16.3)||30.7 (29.1-30.9)||23.2 (22.1-24.9)||22.7 D (18.5-25.4)||4||22.116||<.001|
|Zfx IntraScan||5||40.3 (32.8-41.1)||38.9 (36.4-50.2)||35.8 (30.2-36.3)||34.3 (30.9-45.5)||33.5 (25.4-42.8)||36.4 B (91.0-41.1)||4||4.416||.353|
|Total||26.0 ab (20.8-40.4)||28.2 ab (24.0-37.3)||22.3 b (17.4-34.4)||31.2 a (27.1-36.0)||25.2 ab (22.6-36.1)||27.0 (22.3-36.0)|