In this study, we assessed the precision and trueness of dental models printed with 3-dimensional (3D) printers via different printing techniques.
Digital reference models were printed 5 times using stereolithography apparatus (SLA), digital light processing (DLP), fused filament fabrication (FFF), and the PolyJet technique. The 3D printed models were scanned and evaluated for tooth, arch, and occlusion measurements. Precision and trueness were analyzed with root mean squares (RMS) for the differences in each measurement. Differences in measurement variables among the 3D printing techniques were analyzed by 1-way analysis of variance (α = 0.05).
Except in trueness of occlusion measurements, there were significant differences in all measurements among the 4 techniques ( P <0.001). For overall tooth measurements, the DLP (76 ± 14 μm) and PolyJet (68 ± 9 μm) techniques exhibited significantly different mean RMS values of precision than the SLA (88 ± 14 μm) and FFF (99 ± 14 μm) techniques ( P <0.05). For overall arch measurements, the SLA (176 ± 73 μm) had significantly different RMS values than the DLP (74 ± 34 μm), FFF (89 ± 34 μm), and PolyJet (69 ± 18 μm) techniques ( P <0.05). For overall occlusion measurements, the FFF (170 ± 55 μm) exhibited significantly different RMS values than the SLA (94 ± 33 μm), DLP (120 ± 28 μm), and PolyJet (96 ± 33 μm) techniques ( P <0.05). There were significant differences in mean RMS values of trueness of overall tooth measurements among all 4 techniques: SLA (107 ± 11 μm), DLP (143 ± 8 μm), FFF (188 ± 14 μm), and PolyJet (78 ± 9 μm) ( P <0.05). For overall arch measurements, the SLA (141 ± 35 μm) and PolyJet (86 ± 17 μm) techniques exhibited significantly different mean RMS values of trueness than DLP (469 ± 49 μm) and FFF (409 ± 36 μm) ( P <0.05).
The 3D printing techniques showed significant differences in precision of all measurements and in trueness of tooth and arch measurements. The PolyJet and DLP techniques were more precise than the FFF and SLA techniques, with the PolyJet technique having the highest accuracy.
The PolyJet technique provides the greatest trueness and precision.
Accuracy of 3-dimensional printers differs in the sizes of print outcomes.
Three-dimensional printing techniques may be used for orthodontic purposes.
Advances in digital technology and manufacturing have rapidly changed dentistry. Three-dimensional (3D) printing is the most advanced technology in the manufacturing industry because it shortens manufacturing lead time, reduces required costs, and allows printing of items with complex structures. Thus, it has been implemented in dentistry to manufacture clear orthodontic aligners, implant surgical templates, orthognathic surgical wafers, and provisional crowns. Three-dimensional printers produce 3D structures, based on a 3D design file. Three-dimensional printing is an additive manufacturing process in which materials are added layer on layer to produce an object, as opposed to reductive manufacturing in which material is subtracted to produce the object. Rapid prototyping is another term for the additive manufacturing process.
Scanning technology could potentially be used to convert plaster models or impressions into 3D digital models. However, physical models are required to fabricate orthodontic appliances. If 3D digital models could be printed with a 3D printer to fabricate a physical model, several steps of the traditional model-manufacturing process could be omitted, thereby shortening the lead time and facilitating the production of multiple copies without distortions of shape. Thermoplastic orthodontic appliances are widely fabricated on the basis of physical models printed by 3D printers.
Three-dimensional printed models could also be used to fabricate orthodontic appliances directly. As early as 2006, Ciuffolo et al fabricated a tray for indirect bracket bonding via rapid prototyping for clinical usage. A retainer was recently manufactured using a selective laser sintering (SLS) 3D printer. A virtual wafer has been produced using a stereolithography apparatus (SLA) 3D printer.
To use 3D printed dental models for clinical purposes, accuracy of the printed outcome must be ensured. To date, the accuracy of 3D printed outcomes falls short of that produced via computer numeric control processing as a reductive manufacturing process. In some instances, 3D printed products require postprocessing to ensure smooth surfaces. Therefore, the types and features of 3D printers should be considered for appropriate applications in orthodontics.
Few studies have validated the accuracy of 3D printed models. To assess the accuracy of the models, Hazeveld et al fabricated dental models using 3 types of rapid prototyping and measured the size of teeth with digital calipers. They measured only the mesiodistal width and height, but not the buccolingual width. However, the buccolingual width of teeth is also influenced by the method of polymerization and printing, which may affect the fit of individualized trays or orthodontic appliances. Murugesan et al also manufactured dental models using 3 types of rapid prototyping and measured the teeth using digital calipers to compare the accuracy of the models.
The authors of both aforementioned studies used digital calipers to measure the teeth; this might have resulted in measurement errors because it is difficult to find a reference point on the tooth surface. Furthermore, they printed each model from each printer only once and compared the accuracy between models printed by different printers. To address these shortcomings, we established reference points on the teeth and the gingival areas, and applied 3D software to validate the precision and trueness of dental models fabricated by 3D printers.
The purpose of this study was to analyze the precision and trueness of dental models manufactured by 4 types of 3D printers. The null hypothesis was that there would be no significant differences between the printed models fabricated by different 3D printing techniques in precision and trueness.
Material and methods
The process for design and 3D printing of digital reference models is shown in Figure 1 . A pair of typodont models (D13PP-TR.1; Nissin, Kyoto, Japan), which included 14 maxillary and 14 mandibular permanent teeth, was chosen. The dental models were scanned using a 3D model scanner (Identica Hybrid; MEDIT, Seoul, Korea) with a precision of ±7 μm. All scanned files were converted into the stereolithography (.stl) format. An .stl file is a format used by stereolithography software to generate information needed to produce 3D models on stereolithography machines by rapid prototyping processes. To design reference markers on the scanned dental models, the .stl files were converted to CAD files using a 3D modeling software (Rapidform 2006; INUS Technology, Seoul, Korea). To standardize the measurement, 102 half-ball markers (diameter, 1.0 mm) were placed on the 3D CAD models as reference points ( Fig 2 ). Moreover, to clarify the relative positions of the models for occlusion, semicircular cylindrical notches (diameter, 2.0 mm) were placed at the model base of each maxillary and mandibular central incisor and left and right second molar ( Fig 2 , B ).
Four types of 3D printers were selected, based on the printing technique: stereolithography apparatus (SLA) technique (ZENITH; Dentis, Daegu, Korea), digital light processing (DLP) technique (M-One; MAKEX Technology, Zhejiang, China), fused filament fabrication (FFF) technique (Cubicon 3DP-110F; HyVISION System, Sungnam City, Korea), and PolyJet technique (Objet Eden 260VS; Stratasys, Eden Prairie, Minn) ( Table I ). The designed reference models were printed using these 4 printers and printed 5 times per printer.
|Category||Additive manufacturingprocess||Techniques||3D printers||x-y resolution||Layer thickness||Print time|
|Vat photopolymerization||An additive manufacturing process in which liquid photopolymer in a vat is selectively cured by light-activated polymerization.||Stereolithography apparatus||ZENITH||50 μm||50 μm||4 h and 5 min|
|Digital light processing||M-One||70 μm||75 μm||1 h and 35 min|
|Material extrusion||An additive manufacturing process in which material is selectively dispensed through a nozzle or orifice.||Fused filament fabrication||Cubicon 3DP-110F||100 μm||100 μm||2 h and 30 min|
|Material jetting||An additive manufacturing process in which droplets of build material are selectively deposited.||PolyJet||Objet Eden 260VS||1600 dpi||16 μm||1 h and 47 min|
The 3D printed models were scanned using the 3D model scanner. The scanned models were saved as .stl files. The dental models printed via the DLP technique were scanned after applying a scan spray to prevent reflection of light onto the material ( Fig 3 ).
Tooth, arch, and occlusion were measured using 3D inspection software (Geomagic Control; 3D Systems, Rock Hill, SC). If a half-ball marker was a measurement point, then a measurement was made with reference to the center of the half ball.
Two horizontal reference planes were created to reduce measurement errors. The middle horizontal plane was formed by connecting the 3 half-ball markers at the halfway points of the clinical crown heights of the first molars and right central incisors. The cervical horizontal plane was created by connecting the 3 half-ball markers in the cervixes of the first molars and the central incisors ( Fig 4 ).
For tooth measurements, the mesiodistal width was measured by the distance between the points perpendicular from the mesial and distal contact points to the middle horizontal plane ( Fig 5 , A ). The buccolingual width of the reference points was determined by measuring the distance between the reference points at the buccal cervix and lingual cervix. The buccolingual width was measured by using the section through the object function of the Geomagic Control software to obtain a cross-section of the dental model ( Fig 5 , B and C ), using the middle horizontal plane. The section through the object function shows a 2-dimensional cross-section of a 3D object cut through into a reference plane. The vertical crown height was the distance from the incisal edge or cusp tip of a tooth to the cervical horizontal plane.
For arch measurements, the intercanine width was the distance between the cervical half-ball markers of the canines. The intermolar width was the distance between the cervical half-ball markers of the first molars.
Occlusion was measured on the basis of interarch distances. The maxillary and mandibular models were occluded by the union function of the Geomagic Control software, and the distances between the half-ball markers of the left and right central incisors, canines, and first molar cervixes on the maxillary and mandibular models were measured.
Dimensional differences in tooth, arch, and occlusion measurements among the 3D printed and digital reference models were computed for precision and trueness ( Fig 1 ). Precision is the closeness of the results of repeatedly printed dental models, and trueness is the closeness of a dental model to a true value (ISO 5725-1). The greater the precision, the more predictable the measurement. A high trueness value is close to or equal to the actual dimensions of the measured object. To determine precision, the 3D printed models were combined to make 10 pairs to compare with each other by technique type. The differences between the 5 pairs of 3D printed models and the reference model were analyzed to determine trueness.
For each variable of tooth and arch measurements, root mean square (RMS) values were calculated with respect to precision and trueness. The overall RMS values of tooth, arch, and occlusion variables were also calculated using the following formula: