INTRODUCTION

Recent advances in intraoral scanning and orthodontic treatment planning software, followed by the expansion of low-cost three-dimensional (3D) printers, have increased applications for 3D printing in dentistry. Several 3D printer types, such as fused deposition modeling, material jetting, selective laser doping, stereolithography (SLA), and digital light processing (DLP), are available in the market.¹ However, SLA and DLP are the two most used technologies for 3D printing of dental models and devices due to their rapid processing, relatively low cost, and high resolution.²

Various factors can directly influence the 3D printing process, leading to the need for more research to evaluate these new products and technologies. Initially, researchers were interested in the accuracy of 3D printed models and the conventional workflow fabrication of aligners, retainers, and splints on these printed models.^2–6 Different printing technologies have evolved, and a new generation of resins has emerged, allowing for the direct fabrication of clear devices using 3D printing without the need for intermediate models, acrylic manipulation, or thermoforming steps.^7,8 The growing popularity of 3D printed aligners, retainers, and occlusal splints has aroused the interest of dental clinicians to implement their manufacture in the office. With this new technology, the time-consuming and complicated conventional workflow manufacturing process has the potential to be overcome. Also, there is the ability to replicate the same device design in case of loss.⁹ Direct 3D printing can improve efficiency, enabling faster delivery of appliances or aligners to patients,^7,8,10 bypassing the dental lab.² Thus, the conventional process can be replaced by implementing a digital workflow for everyday use in dental practice.⁹

Precision and accuracy are critical in fabricating dental models or in-office appliances in orthodontics. Camardella et al. showed that PolyJet printing technology produced more accurate models than other printing technologies.⁴ It may be inferred that, since SLA and DLP printers require postcure processing with heat exposure, it could lead to additional shrinking of the resin³ and less dimensional accuracy. Brown et al. showed similar results when comparing DLP and PolyJet printed models with stone models for almost all recorded measurements.³

Researchers have investigated the effect of print orientation and ultraviolet (UV) light curing duration on the dimensional accuracy of clear appliances fabricated directly using an SLA 3D printer. It was shown that the print orientation and postprint curing duration had little effect on the overall accuracy of the 3D-printed aligner design.⁶ In another study, investigators concluded that UV light cure duration did not significantly impact the dimensional accuracy of 3D-printed clear aligners.⁷ Additionally, researchers evaluated the precision, trueness, and accuracy of clear orthodontic retainers fabricated using 3D printers with different technologies, concluding that they were clinically acceptable and accurate compared with the standard reference file.⁶ Marcel et al. investigated the accuracy of computer-aided design/computer-aided manufacturing (CAD/CAM)-fabricated bite retainers using milling and 3D printing methods, horizontal and vertical orientations, and various material options. In that study, it was concluded that milled and 3D-printed bite retainers produced clinically accurate results.¹¹ Williams et al. evaluated the accuracy and precision of 3D-printed retainers using an SLA printer (Form 2, Clear resin, Formlabs Inc, Somerville, Mass) at different printing angles (15°, 30°, 45°, 60°, and 90°). They assessed the printing time and amount of resin consumed. The results showed that all printing angles were accurate within 0.25 mm, with the 15° angle being the most time-efficient and the 45° angle being the most cost-effective printing setting.⁸ Meanwhile, more information is needed concerning the fabrication of in-office retainers or clear aligners directly.

The purpose of this study was to evaluate the impact of printer technology (SLA and DLP) and print orientation (0° and 90°) on the dimensional accuracy of directly printed retainers. The null hypothesis was that the printer technology and print orientation would not affect the dimensional accuracy of 3D-printed retainers.

MATERIALS AND METHODS

A maxillary splint standard tessellation language (STL)¹² file previously created using 3D software (Meshmixer, version 3.2; Autodesk, San Rafael, Calif) was used to fabricate the samples (Figure 1). A sample size of 10 printed retainers for each group was calculated, considering the limit of ±0.25 mm, 80% power and α value of 0.05. A total of 40 occlusal retainers were fabricated using two different 3D technologies: SLA (Form 3; Formlabs) and DLP (Sprintray Pro, SprintRay Inc; Los Angeles, Calif). The master file was prepared for printing in two orientations (0° and 90°) using each printer preparation software (SLA printer: PreForm Software, version 2.19.3, Formlabs; and DLP printer: Rayware, version 2.9.2, SprintRay; Figure 1B,C). The two printing orientations were 0° (occlusal plane parallel to the print platform) and 90° (occlusal plane perpendicular to the print platform). Each software application automatically generated the support structures. The supports were edited when needed to avoid their inclusion on the intaglio surface. Each print job included five retainers when printing in 0° and 10 retainers when printing in 90°. The retainers were fabricated using a resolution and layer thickness of 100 mm for both printer types. Forty copies of the master splint design file were printed in six different print jobs using Dental LT clear resin for the SLA printer and SPLINT clear resin for the DLP printer (Figure 2A).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

After printing, retainers were processed according to each manufacturer’s instructions. The retainers were removed from the build platform and rinsed in isopropyl alcohol using Form Wash for the Formlabs, and Pro Wash/Dry for the Sprintray. Compressed air was used to dry the retainer surfaces. Finally, the retainers were placed in each heated curing unit. The retainers were scanned using cone-beam computed tomography (CBCT)⁸ to create corresponding digital images (Figure 2B), with the supports still attached. The settings were chosen according to previous research using CBCT.^13–16 The obtained images of the printed retainers were saved as DICOM files and were then converted to STL files using Dolphin Imaging software (version 11.95, Patterson Dental, Chatsworth, Calif). Removal of the supports was carried out digitally to allow adequate comparison of the printed retainers with the master file. This was done by superimposing the digital file of each printed retainer on the master file, using the best-fit tool on Geomagic software (Geomagic Control, version 2015.1.1; 3D Systems, Rock Hill, SC), considering the region from first molar to first molar. A ±0.25 mm tolerance was set to detect differences between the superimposed files (Figure 3). Trueness was determined by comparing the digital data of the printed splint with the reference STL (master) file.⁸

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

Average Deviation (mm) from the Initial STL Retainer

The median and the interquartile range (IQR) of the average deviation of the printed retainers compared with the master STL model per group of different printers/orientations is presented in Table 1. The lowest median average deviation was observed for the DLP horizontally printed models, followed by the SLA horizontally printed retainers, while the greatest was found for the DLP vertically printed retainers. The Kruskal-Wallis test (P < .001) indicated that there was at least one group with an average deviation distribution different from the others. Pairwise group comparisons are presented in Table 2.

Table 1. Average Deviation From the Initial Standard Tessellation Language (STL) Retainers. IQR Indicates Interquartile Range; DLP Indicates Digital Light Processing; SLA Indicates Stereolithography

View Fullscreen

Table 2. Average Deviation of the Three-Dimensional Printed Models Compared With the Initial Standard Tessellation Language (STL) Retainer. DLP Indicates Digital Light Processing; SLA Indicates Stereolithography

View Fullscreen

Inside the Tolerance Levels Ratio Percentage

The median and the IQR of the inside the tolerance levels ratio of the 3D printed retainers compared with the master STL model are presented in Table 3. The Kruskal-Wallis test (P < .001) indicated that there was at least one group with inside the tolerance levels ratio distribution different from the others. The results of the pairwise group comparisons are presented in Table 4.

Table 3. Inside the Tolerance Levels (±0.25 μm) Compared With the Initial Standard Tessellation Language (STL) Retainers. IQR Indicates Interquartile Range; DLP Indicates Digital Light Processing; SLA Indicates Stereolithography

View Fullscreen

Table 4. Inside the Tolerance Levels (±0.025 μm) Ratio of the Three-Dimensional Printed Models Compared With the Initial Standard Tessellation Language (STL) Model. DLP Indicates Digital Light Processing; SLA Indicates Stereolithography

View Fullscreen

DISCUSSION

In this study, we aimed to evaluate the impact of printing technology (SLA or DLP) and print orientation (0° or 90°) on the dimensional accuracy of directly printed occlusal retainers. The null hypothesis was that the printer technology and print orientation of the splint during printing would not affect the dimensional accuracy of the 3D-printed splint. The hypothesis was rejected.

The fundamental difference between SLA and DLP is the light source. In SLA printers, the UV laser beam moves from point to point, tracing the geometry, while DLP uses a stationary UV light projector. Both processes work by selectively exposing the liquid resin to a light source; SLA uses UV light to cure the resin on the platform, building layer by layer from the bottom until the final product is obtained, and DLP uses a projector instead of a thin UV laser which allows for curing the entire layer each time.¹⁹ The printing angulation is another critical point to consider since it can influence accuracy and may alter the number of appliances to be fabricated each time, affecting the efficiency of the manufacturing process. In an orthodontic workflow, increasing the number of parts that can fit on the platform while still generating clinically acceptable results can maximize time and reduce operational costs.⁷

In the present research, the postprinting UV was performed according to each resin manufacturer’s recommendation, and, therefore, does not seem to have been the reason for the differences between the groups. The retainers printed with the two printing technologies (SLA and DLP) showed mean discrepancies of <0.25 mm, indicating accuracy within the clinical acceptance tolerance (Table 1).

In this study, we also investigated the effect of print angulation on the accuracy and precision of 3D-printed dental models. Orientation has been claimed to affect the accuracy of the printing process, as vertical printing requires more print layers than horizontal printing. One possible explanation is that an increase in print layers leads to an increase in printing time, which could increase errors during printing.^3,6 In a previous study,⁷ in which authors evaluated the impact of print orientation on the accuracy of clear aligner design using SLA technology, it was found that the devices printed at a 45° angle had the least deviation from the original file. The authors observed deviations occurring at different parts of the aligners depending on the orientation used, but consistently at the same locations within each group. They concluded that the printing orientation could affect the clinical usefulness of the printed aligner. However, they noted that the study had limitations, such as using a contrast spray or powder to facilitate scanning. They also found that print orientation and duration of postprint curing had minimal impact on the overall accuracy of the 3D-printed aligner design.⁷ Boyer et al.²⁰ investigated the effect of print orientation on the dimensional accuracy of orthodontic aligners printed directly with an SLA 3D printer. A digitally designed aligner was printed at eight angulations using a stereolithography 3D printer. Printing at a 90° angulation was recommended since that was the group with the most accurate prints compared with the seven other orientations investigated, although not all differences were statistically significant.²⁰

Previous studies compared different 3D printing angulations for retainer fabrication and reported accuracy,^8,21 precision,^8,21 time of printing, and amount of resin used with SLA technology.⁸ Naeem et al.²¹ evaluated differences in the precision, trueness, and accuracy of 3D-printed clear orthodontic retainers fabricated using printer systems with different printing technologies. Retainers fabricated by different technologies were shown to be clinically acceptable and accurate compared with the standard reference file. Based on both high precision and trueness, SLA printers yielded the most accurate retainers.

Williams et al.⁸ compared the accuracy and precision of 3D-printed retainers at various angulations and evaluated the effect of angulation on printing time and the amount of resin used. Three-dimensional retainers using an SLA printer were found to be accurate within 0.25 mm at all print angulations at the cusp tips and incisal edges when compared with the digital reference file. However, the smooth facial surfaces did not meet the required clinical acceptability standards. Print angulation affected the cost and amount of resin used in 3D printing. It was found that 3D printing at 15° was the most time efficient, while 3D printing at 45° was the most cost effective. However, errors were observed in the anterior surface (incisors region) of the 3D-printed retainers. Therefore, authors of future studies should evaluate the fit of 3D-printed retainers and assess their clinical performance.⁸ In the present study, most of the differences were found at the inner surfaces of the molar area.

In a recent systematic review,² authors examined the accuracy of various types of 3D printers and factors that may impact 3D printing of dental models in orthodontics. They searched Medline, Scopus, and Cochrane Library databases and identified 11 articles that met their inclusion criteria, consisting of in vitro prospective studies with a low risk of bias. The findings indicated that the accuracy of a printed dental cast can be influenced by the type of 3D technology used, the design of the dental model base, and the printing materials employed. However, the layer height and position of the model on the building platform did not appear to affect the printing accuracy.²

In contrast with prior research, a CBCT scan was used in the present study rather than an intraoral scanner to obtain the sample,^7,20 thereby addressing a previous limitation. Additionally, in the present research, we investigated the influence of printing orientation on the accuracy of the printed models with two different technologies (SLA and DLP). Like previous studies,^6,20 in the current research, differences in deviations from the original (master) file were found among the groups, especially when considering the printing orientation (Table 1, Figures 4 and 5).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Regardless of the printing orientation, the SLA printer produced more accurate prints than the DLP printer. The SLA printer achieved a higher median percentage of accurate prints for both printing orientations compared with the DLP printer. Additionally, the average deviation from the master file was smaller for the SLA printer than the DLP printer, indicating a higher level of accuracy in the prints produced by the SLA printer. The difference in accuracy between the printers was more pronounced when the prints were produced at a 90° orientation. These findings suggest that the SLA printer may be more suitable for applications where high accuracy is a critical factor (Figures 4 and 5).

One of the limitations of the present study was the use of two different resins and handling processes for each printer type, following the respective manufacturer recommendations. Therefore, some of the observed differences in the results may have beer attributable to the material characteristics rather than solely due to the 3D printing technology. Additional research is warranted concerning these new resins on the market, their handling processes, and other possible variations derived from the direct fabrication process of 3D-printed devices, such as postprocessing procedures, location on the platform, and differences from the scanning process.

The directly printed clear appliances from DLP and SLA printers with both print orientations (0° and 90°) were suitable for orthodontic applications, considering the limit of ±0.25 mm set as the reference.^6,8.Both printers are suitable for orthodontic applications, although print orientation was more prone to affect the accuracy and quality of the printed appliance. Caution should be taken when interpreting these results, as technologies vary in several factors.

CONCLUSIONS

• Both technologies (DLP and SLA) produced 3D-printed results that are compatible with orthodontic clinical needs.

• Within the limitations of this study, the printing orientation was more important than the printer type regarding its accuracy.

• Further studies are needed to assess the accuracy of the printed appliances clinically.