Abstract
Objective
This study aimed to evaluate the accuracy of a three-dimensional (3D)-printed palatal auxiliary device (PAD) compared to computer-aided design (CAD) reference data.
Methods
Thirty patients who underwent orthodontic treatment using a PAD for maxillary protraction after palatal expansion were included in this study. Two groups of 15 patients were analyzed to compare the accuracy of the two PAD designs. Accuracy and adaptation were assessed through two sets of measurements 1) deviations between the printed PAD and its CAD reference to determine printing accuracy, and 2) deviations in PAD positioning when clinically applied, simulated on a printed model. Sixteen measurement points (anterior, posterior, left, and right) were evaluated across the x-, y-, and z-axes and t tests were performed for comparison.
Results
The PAD-only measurements showed errors greater than 0.1 mm on all axes but within the marginal limits of 0.05 mm to 0.25 mm (P < 0.05). Similar results were observed with the PAD-adapted measurements. Significant differences were found across all axes in the PAD-only group and between the two designs. Tukey’s post hoc analysis identified a specific ranking of errors in the combination-type PAD left, anterior, posterior, and right. However, for PAD-adapted group, majority of comparisons showed no significant differences and those that did lacked consistency in pattern.
Conclusions
This study demonstrates that 3D-printed PADs maintain clinically acceptable accuracy and can be effectively integrated into existing intraoral devices. Although errors varied in certain areas, they did not significantly impact the final adaptation.
Keywords: 3D printing, Palatal expansion, Orthodontic appliance, Computer-aided design
INTRODUCTION
Rapid advancements in digital technology have transformed modern dentistry significantly, particularly through the integration of digital charting, computer-aided design (CAD) computer-assisted manufacturing (CAM) technology, and intraoral scanning (IOS). Digitalization in dentistry enhances efficiency, improves patient outcomes, and enables in-house production of dental restorations and appliances.1 IOS has replaced traditional impression materials like alginate and silicone, offering greater accuracy, consistency, and comfort. Initially limited to single-tooth or short-span restorations in the 1990s, advancements in the early 2000s have led to the development of full-arch scanners. Studies have shown that modern IOS systems produce impressions of comparable quality to conventional methods while being less user-dependent and more comfortable for patients.2
After data acquisition through IOS, digital files are typically saved in stereolithography (STL) format and used in CAD software to design restorations or orthodontic appliances.1 Fabrication is performed via CAM using either subtractive manufacturing (milling) or additive manufacturing (three-dimensional [3D] printing).3-5 Common dental 3D printing technologies include STL, digital light processing (DLP), material jetting, and material extrusion, each with varying in accuracy, speed, and material compatibility.6-9
Digital technology has improved orthodontic treatment for maxillary transverse deficiency (MTD), a condition characterized by a narrow maxilla, crowded dentition, and potential airway issues. Maxillary expansion, a common treatment for MTD, has evolved from traditional to miniscrew-assisted expansion, which minimizes unwanted tooth movement. Among the methods, bone-borne expanders produce the least amount of unwanted tooth movement.10-12
In addition to maxillary expansion, patients with Class III malocclusion often require maxillary protraction to correct their bite.11,13-16 Conventionally, this is achieved using extraoral facemasks with elastics, but patient compliance is low owing to the bulky and unesthetic nature of the device. To address this issue, bone-anchored maxillary protraction, which utilizes miniscrews in the maxilla and mandible to apply intermaxillary elastics, was developed. This method has demonstrated superior maxillary advancement and improved patient compliance compared to traditional facemask therapy.14-16
A recent innovation in this field is the 3D-printed palatal auxiliary device (PAD), designed for attachment to pure bone-borne ATOZ expanders (MK Meditech Co., Sungnam, Korea) after maxillary expansion to facilitate protraction.17 The PAD is a removable appliance with two design variations: the wing type, featuring a housing compartment for retention and hooks for intermaxillary elastics, and the combination type, which includes wire attachments for traditional facemask therapy (Figures 1 and 2). When applied, the force is set to 250 g per side, 500 g in total. This study focused on PADs for bone-borne expanders owing to their unique attachment designs. Because retention depends on dimensional accuracy, this study aimed to assess the precision of printed PADs by comparing rescanned data to reference models and evaluating deviations upon adaptation.
Figure 1.
Types of palatal auxiliary device (PAD). A, Wing-type PAD, with the housing compartment divided into anterior and posterior, and the wings divided into left and right from the palatal view. B, Combination-type PAD, with the same components as the wing-type PAD with addition of wire attachment components extending from anterior housing.
Lt, left; Rt, right.
Figure 2.
A, Photography of palatal auxiliary device (PAD) applied to patient with facemask and B, intraoral photo of PAD applied to a patient. A-1, Frontal view of a patient with facemask; A-2, oblique view; A-3, lateral view; B-1, right lateral intraoral photo of a patient with PAD applied; B-2, occlusal intraoral photo; B-3, left lateral intraoral photo.
MATERIALS AND METHODS
Sample selection
Patients treated at the Orthodontic Department of Orthodontics, Kyung Hee University, Dental Hospital, between January 2018 and August 2023 were selected based on the following criteria: receiving maxillary expansion treatment using a pure bone-borne ATOZ expander and maxillary protraction using a pure 3D-printed PAD. Of 88 patients, 15 were selected for each PAD type (wing or combination). This study was approved by the Institutional Review Board of Kyung Hee University Dental Hospital (IRB No. KH-DT23033). The requirement to obtain informed consent was waived.
Palatal auxiliary device fabrication procedure
The general manufacturing procedure is illustrated in Figure 3. In this study, post-processing procedures, such as polishing and glazing, were omitted to prevent alterations in the surface layers that affected the results.
Figure 3.
Diagram of the process for manufacturing a palatal auxiliary device (PAD). A, A scan file is made into digital model with base; B, the PAD is designed using Denture Module of 3Shape Dental System; C, the PAD is printed on a building plate; D, the support is removed; E, the PAD receives an isopropyl alcohol wash in an ultrasonic bath; F, post-curing with Tera Herz Cure; G, adaptation of the PAD; H, removal of support residuals and surface polish; I, glaze application; J, finished PAD on the model.
Data processing and design
Intraoral data were processed into model data using the Model Builder of the 3Shape Dental System (3Shape, Copenhagen, Denmark). The model was constructed with a 4 mm base, without hollowing. The model file was used as the base for designing the PAD using a Denture Base Module in Dental System (3Shape). Wing and combination types were designed accordingly (Figure 4). Both the PAD and model designs were exported as STL files for fabrication.
Figure 4.
Computer-aided design process for (A) wing type and (B) combination type palatal auxiliary device (PAD). A-1, B-1, Indicating direction of insertion; A-2, B-2, digital blockouts; A-3, B-3, outline of PAD; A-4, B-4, digital wax-up; A-5, B-5, trimming out areas that directly touch palatal mucosa by 1–2 mm; B-6, adding wire attachment holes.
3D printing setup and printing
The STL files were imported to Uniz Dental (Uniz Technology LCC, San Diego, CA, USA) for orientation and support attachment. The models were oriented flat on a building plate without a support. The PADs were turned palatally, with housing compartments parallel to the building plate (Figure 5), lifted by 4 mm, and supports were added. Printing layers were set at 100 μm for both model and PADs. The models were first printed with SprintRay Pro 95 (SprintRay Inc., Los Angeles, CA, USA) with S-200 grey resin (Graphy Inc., Seoul, Korea) and a Grey Model Resin preset.
Figure 5.
A, Illustration of orientation of model and B, palatal auxiliary device (PAD). A-1, Side view of model orientation on building plate; A-2, posterior view of model orientation on building plate; B-1, side view of PAD orientation on building plate; B-2, posterior view of PAD orientation on building plate.
Postprocessing
After printing, the models were removed from the building plate and washed with 99% isopropyl alcohol (Echo & Chem, Seoul, Korea) for 5 minutes using ultrasonic cleaning. It was dried completely and then post-cured with a Tera Herz Cure (Graphy Inc.) at level 2 for 2 minutes, followed by cooling at room temperature. The PADs were printed with the same printer using TFDH resin (Graphy Inc.) and a Denture Base preset. After the printing was completed, the PAD was removed from the building plate. The supports were removed and the PAD was washed with 99% isopropyl alcohol for 1 minute under ultrasonic cleansing. It was dried completely and cured using the same curing machine under a level 2 setting, nitrogen gas for 15 minutes. Immediately after post-curing, the PAD was adapted to the model during cooling.
Final scanning
To evaluate clinical adaptation, two scans were performed: the PAD alone after printing (PAD-only) and the PAD seated on the printed model to simulate clinical placement (PAD-adapted). Both were scanned using an i900 intraoral scanner (Medit Inc., Seoul, Korea). This process aims to capture both the as-printed condition and positional accuracy of the device when placed onto the model.
Superimposition
The deviations were quantitatively analyzed through 3D digital superimposition of the reference and rescan data files of the printed PAD using Geomagic Control X (3D Systems, Rock Hill, CA, USA). The origin of coordination was set at the most anterior right corner of the housing component. The measurement axes were defined as follows (Figure 6): anteroposterior (x-axis), positive values indicated anterior deviation; negative values indicated posterior deviation; transverse (y-axis), positive values indicated right deviation; negative values indicated left deviation; vertical (z-axis), positive values indicated gingival deviation; and negative values indicated coronal deviation. The PAD-only and PAD-adapted superimpositions followed the same geometric alignment methods. PAD-only superimposition begins by importing reference data, and the surface is segmented using automation. The rescan data were then imported and quickly aligned. The best-fit alignment was additionally executed by restricting the alignment area of the flat surfaces of the housing compartment. Similar processes were followed in the PAD-adapted superimposition. The scanned model-PAD assembly was segmented into two parts: the model and the PAD. During the best-fit alignment, only the model surface was selected as the reference alignment area. This allowed an accurate comparison of the position of the PAD when adapted to the model, simulating clinical placement (Figure 6).
Figure 6.
Orientation of axis in palatal auxillary device (PAD)-only (A, C) and PAD-adapted (B, D). Anteroposterior measurement, linear measurements (mm) of deviation along the x-axis. An anterior deviation was regarded as a positive value, and posterior deviation as negative value; Transverse measurement, linear measurements (mm) of deviation along the y-axis. A deviation to the right was regarded as a positive value and deviation to the left was regarded as a negative value; Vertical measurement, linear measurements (mm) of deviation along the z-axis. A deviation in gingival direction is regarded as a positive value and a deviation in the coronal direction is regarded as a negative value.
Measurements
Two sets of measurements were conducted to assess the accuracy and adaptation of the printed PAD: 1) deviations between the printed PAD and its reference data to determine the printing accuracy, and 2) deviations in the PAD’s position when clinically applied by simulating its adaptation on a printed model.
To evaluate printing accuracy, 16 measurement points were placed on the adaptation side of the PAD. These included eight points on the housing component, further divided into anterior (points 1–4) and posterior (points 5–8), and four points on each wing component (points 9–12 for the left wing and points 13–16 for the right wing) (Figure 7). All measurement points were located at the corners of three concurrent lines outlining the outer borders of the PAD contact area with the expander body and base to ensure retention.
Figure 7.
Specification of palatal auxiliary device (PAD)-only measurement points. A, Measurement points on wing type; B, combination type; C, detailed measurement points on wing type; D, detailed measurement points on combination type.
1, The innermost anterior right corner in the anterior housing; 2, the innermost posterior right corner in the anterior housing; 3, the innermost posterior left corner in the anterior housing; 4, the innermost anterior left corner in the anterior housing; 5, the innermost anterior right corner in the posterior housing; 6, the innermost posterior right corner in the posterior housing; 7, the innermost posterior left corner in the posterior housing; 8, the innermost anterior left corner in the posterior housing; 9, the innermost corner of anterior right corner where 3 axes meet in the right wing; 10, the innermost corner of posterior right corner where 3 axes meet in the right wing; 11, the innermost corner of a posterior left corner where 3 axes meet in the right wing; 12, the innermost corner of an anterior left corner where 3 axes meet in the right wing; 13, the innermost corner of an anterior right corner where 3 axes meet in the left wing; 14, the innermost corner of a posterior right corner where 3 axes meet in the left wing; 15, the innermost corner of a posterior left corner where 3 axes meet in the left wing; 16, the innermost corner of an anterior left corner where 3 axes meet in the left wing.
To assess the adaptation accuracy, 16 measurement points were selected on the marginal surface of the PAD. Owing to design variations, the anterior measurement points differed slightly between the wing-type and combination-type designs. In the wing-type design, the anterior points (1–4) were the contact points of the PAD margin and the lateral lines of the anterior expander arms. In the combination-type design, the wire attachments extended from the anterior portion of the PAD, requiring new measurement positions that better represented deviations. The middle anterior points (1’ and 3’) corresponded to the intersecting trimming lines, whereas the lateral points (2’ and 4’) bisected the angle formed by the tangents to the anterior and lateral borders of the wire attachments. For the posterior and wing measurements, points 5–8 represented the contact between the PAD margin and the lateral lines of the posterior expander arms. In the wings, points 9 and 13 marked the anterior contact points between the wings and lateral lines of the expander base. Points 10 and 16 were the most lateral points on the wing margins, whereas points 11 and 15 were at the tips of the lateral hook margins. Finally, points 12 and 14 marked the posterior contact points of the wings with the lateral lines of the expander base (Figure 8).
Figure 8.
Individual point deviation of palatal auxiliary device-only in different marginal limits. Specification of palatal auxiliary device (PAD)-adapted measurement points. A, Wing-type PAD-adapted: 16 measurement points, 4 points each in 4 different components. 1, most anterior contact point of anterior housing margin and right expander arm; 2, second most anterior contact point of anterior housing margin and right expander arm; 3, most anterior contact point of anterior housing margin and left expander arm; 4, second most anterior contact point of anterior housing margin and left expander arm; 5, second most posterior contact point of posterior housing margin and right expander arm; 6, most posterior contact point of posterior housing margin and right expander arm; 7, most posterior contact point of posterior housing margin and left expander arm; 8, second most posterior contact point of posterior housing margin and left expander arm; 9, lateral anterior contact point of right wing margin and right expander base; 10, most lateral marginal point in right wing; 11, sharpest marginal point of lateral hook in right wing; 12, lateral posterior contact point of right wing margin and right expander base; 13, lateral anterior contact point of left wing margin and left expander base; 14, lateral posterior contact point of left wing margin and left expander base; 15, sharpest marginal point of lateral hook in left wing; 16, most lateral marginal point in left wing. B, Combination-type PAD-adapted: 16 measurement points, 4 points each in 4 different components. 1’, most anterior right point of intersecting lines of anterior border; 2’, marginal point that bisects the angle formed by tangents to the anterior and lateral borders of right wire attachment; 3’, most anterior left point of intersecting lines of anterior border; 4’, marginal point that bisects the angle formed by tangents to the anterior and lateral borders of left wire attachment. The definitions for points 5–16 are the same as described for the wing-type PAD. C, Detailed measurement points of PAD-adapted for wing type; D, detailed measurement points of PAD-adapted for combination type.
Statistical analysis
The reliability and repeatability of the measurements were evaluated by comparing rescan data from the printed PAD with the reference file. Measurements were taken at 2-week intervals and reliability was quantified using the intraclass correlation coefficient, with a threshold of 85%, indicating reliable repeatability.
To compare the deviations at individual measurement points across the different PAD designs, the mean absolute differences were calculated and analyzed within three marginal limits: 0.1 mm, 0.25 mm, and 0.5 mm. A t test was performed to assess statistical significance, with a P value of < 0.05 considered significant.
Deviations between the two types of PAD were analyzed by axis using a paired t test. One-way analysis of variance (ANOVA) was conducted to assess the interactions among different components, followed by Tukey’s post hoc test for significance ranking.
All statistical analyses were performed using SPSS software (version 22.0; IBM, Armonk, NY, USA).
RESULTS
Comparisons within different marginal limits for palatal auxiliary device
Across both PAD designs, deviations remained within the 0.5 mm limit, with most within 0.25 mm and 0.1 mm. At the 0.1 mm threshold, points 1–8 (anterior/posterior) had P values > 0.05, suggesting that they exceeded this stricter limit, a trend consistent across both wing and combination types (Supplementary Table 1). Grouped analysis confirmed higher deviations in anterior and posterior regions at the 0.1 mm margin. Notably, combination-type PADs showed better precision (P < 0.05) at this level than wing types, although their mean deviations remained low (Supplementary Table 2).
Comparisons within different marginal limits for palatal auxiliary device adaptation on the printed model
The PAD-adapted measurements showed trends similar to those of the PAD-only analysis. P values for 0.25 mm and 0.5 mm margins were below 0.05 across all axes, indicating a good overall fit. At the stricter 0.1 mm margin, more points, particularly in the x- and y-axes, had P values > 0.05, whereas the z-axis showed statistical significance for points 1–9, 13, and 16. The anterior points (1–4) may be less reliable owing to the design variability. For combination-type PAD, most points were significant at the 0.25 mm and 0.5 mm margins, especially on the y- and z-axes. At the 0.1 mm level, 5 (x), 6 (y), and 3 (z) points were significant. The wing type showed strong statistical significance in the z-axis at 0.25 mm and 0.5 mm, with fewer significant points at 0.1 mm across all axes (Supplementary Table 3). The component-level analysis confirmed that all designs were statistically significant at 0.25 mm the 0.5 mm. At 0.1 mm, significance was observed mostly on the z-axis, except for the anterior (combination type) and lateral regions (both types). On the y-axis, only total and anterior combination types were statistically significant, whereas on the x-axis, total and anterior wing types and posterior regions showed statistical significance (Supplementary Table 4).
Comparison between wing and combination types
A comparison between the two PAD designs revealed significant differences on the x- and y-axes in the PAD-adapted group (P < 0.05). In the anterior region, the z-axis for PAD-only, and all three axes for PAD-adapted showed significant differences. In the posterior region, only the y- and z-axes for PAD-only were significantly different between designs. In the left region, the z-axis of the PAD-adapted was significantly different (Table 1).
Table 1.
Deviation difference between types of palatal auxiliary device (PAD) according to adaptation and component
| Axis | Combination type | Wing type | P | |||||
|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | |||||
| Axis | Combination type | Wing type | P | |||||
| Mean | SD | Mean | SD | |||||
| PAD-only | X | 0.090 | 0.090 | 0.080 | 0.090 | 0.623 | ||
| Y | 0.100 | 0.090 | 0.090 | 0.090 | 0.076 | |||
| Z | 0.100 | 0.090 | 0.110 | 0.110 | 0.058 | |||
| PAD-adapted | X | 0.110 | 0.120 | 0.070 | 0.100 | < 0.001*** | ||
| Y | 0.090 | 0.090 | 0.120 | 0.140 | < 0.001*** | |||
| Z | 0.090 | 0.090 | 0.070 | 0.100 | 0.094 | |||
| Anterior | PAD-only | X | 0.120 | 0.070 | 0.140 | 0.100 | 0.156 | |
| Y | 0.120 | 0.070 | 0.130 | 0.110 | 0.521 | |||
| Z | 0.130 | 0.090 | 0.220 | 0.110 | < 0.001*** | |||
| PAD-adapted | X | 0.180 | 0.130 | 0.020 | 0.030 | < 0.001*** | ||
| Y | 0.060 | 0.060 | 0.110 | 0.150 | 0.005** | |||
| Z | 0.090 | 0.080 | 0.030 | 0.050 | < 0.001*** | |||
| Posterior | PAD-only | X | 0.040 | 0.040 | 0.040 | 0.050 | 0.490 | |
| Y | 0.060 | 0.070 | 0.030 | 0.030 | 0.007** | |||
| Z | 0.060 | 0.060 | 0.040 | 0.030 | 0.008** | |||
| PAD-adapted | X | 0.110 | 0.090 | 0.120 | 0.120 | 0.636 | ||
| Y | 0.090 | 0.100 | 0.110 | 0.110 | 0.240 | |||
| Z | 0.080 | 0.090 | 0.110 | 0.110 | 0.068 | |||
| Left | PAD-only | X | 0.150 | 0.100 | 0.130 | 0.080 | 0.137 | |
| Y | 0.180 | 0.110 | 0.150 | 0.090 | 0.068 | |||
| Z | 0.160 | 0.110 | 0.150 | 0.100 | 0.504 | |||
| PAD-adapted | X | 0.040 | 0.060 | 0.030 | 0.040 | 0.387 | ||
| Y | 0.100 | 0.070 | 0.130 | 0.140 | 0.157 | |||
| Z | 0.070 | 0.080 | 0.040 | 0.060 | 0.011* | |||
| Right | PAD-only | X | 0.030 | 0.030 | 0.020 | 0.030 | 0.342 | |
| Y | 0.040 | 0.050 | 0.040 | 0.050 | 0.994 | |||
| Z | 0.040 | 0.040 | 0.040 | 0.050 | 0.645 | |||
| PAD-adapted | X | 0.120 | 0.130 | 0.130 | 0.120 | 0.765 | ||
| Y | 0.090 | 0.110 | 0.140 | 0.170 | 0.107 | |||
| Z | 0.100 | 0.100 | 0.110 | 0.110 | 0.682 | |||
P value is calculated using paired t test.
SD, standard deviation.
*P < 0.05, **P < 0.01, ***P < 0.001.
Interaction among different compartments
For the PAD-only, both designs exhibited significant differences among the components. In the combination type, the deviation ranking was consistent across all axes, following the order of left, anterior, posterior, and right, with decreasing magnitude. In contrast, wing type did not exhibit a specific deviation ranking pattern (Table 2).
Table 2.
Comparison among different components of palatal auxiliary device (PAD)-only and PAD-adapted by design types
| Type | Axis | n | Mean | SD | F | P | Post hoc | |
|---|---|---|---|---|---|---|---|---|
| PAD-only | ||||||||
| Combination | X | Anterior | 60 | 0.122 | 0.075 | 45.183 | < 0.001*** | l > a > p > r |
| Left | 60 | 0.152 | 0.104 | |||||
| Right | 60 | 0.027 | 0.033 | |||||
| Posterior | 60 | 0.044 | 0.042 | |||||
| Y | Anterior | 60 | 0.120 | 0.074 | 40.381 | < 0.001*** | l > a > p > r | |
| Left | 60 | 0.177 | 0.106 | |||||
| Right | 60 | 0.041 | 0.048 | |||||
| Posterior | 60 | 0.055 | 0.066 | |||||
| Z | Anterior | 60 | 0.126 | 0.092 | 29.447 | < 0.001*** | l > a > p > r | |
| Left | 60 | 0.160 | 0.108 | |||||
| Right | 60 | 0.040 | 0.042 | |||||
| Posterior | 60 | 0.060 | 0.061 | |||||
| Wing | X | Anterior | 60 | 0.144 | 0.105 | 44.281 | < 0.001*** | a > l > p > r |
| Left | 60 | 0.130 | 0.084 | |||||
| Right | 60 | 0.021 | 0.027 | |||||
| Posterior | 60 | 0.038 | 0.051 | |||||
| Y | Anterior | 60 | 0.130 | 0.108 | 40.871 | < 0.001*** | l > a > r > p | |
| Left | 60 | 0.150 | 0.087 | |||||
| Right | 60 | 0.040 | 0.045 | |||||
| Posterior | 60 | 0.029 | 0.028 | |||||
| Z | Anterior | 60 | 0.220 | 0.114 | 73.058 | < 0.001*** | a > r > p > l | |
| Left | 60 | 0.147 | 0.099 | |||||
| Right | 60 | 0.044 | 0.047 | |||||
| Posterior | 60 | 0.035 | 0.027 | |||||
| PAD-adapted | ||||||||
| Combination | X | Anterior | 60 | 0.178 | 0.128 | 17.489 | < 0.001*** | a > r > p > l |
| Left | 60 | 0.040 | 0.057 | |||||
| Right | 60 | 0.122 | 0.127 | |||||
| Posterior | 60 | 0.106 | 0.092 | |||||
| Y | Anterior | 60 | 0.056 | 0.063 | 3.500 | 0.016* | a > l | |
| Left | 60 | 0.103 | 0.071 | |||||
| Right | 60 | 0.094 | 0.105 | |||||
| Posterior | 60 | 0.090 | 0.097 | |||||
| Z | Anterior | 60 | 0.086 | 0.084 | 0.987 | 0.400 | ||
| Left | 60 | 0.073 | 0.084 | |||||
| Right | 60 | 0.101 | 0.103 | |||||
| Posterior | 60 | 0.082 | 0.087 | |||||
| Wing | X | Anterior | 60 | 0.021 | 0.029 | 22.837 | < 0.001*** | l > r > p > a |
| Left | 60 | 0.033 | 0.044 | |||||
| Right | 60 | 0.128 | 0.121 | |||||
| Posterior | 60 | 0.116 | 0.120 | |||||
| Y | Anterior | 60 | 0.115 | 0.146 | 0.541 | 0.654 | ||
| Left | 60 | 0.128 | 0.135 | |||||
| Right | 60 | 0.139 | 0.173 | |||||
| Posterior | 60 | 0.108 | 0.113 | |||||
| Z | Anterior | 60 | 0.028 | 0.048 | 15.606 | < 0.001*** | r > p > l > a | |
| Left | 60 | 0.038 | 0.059 | |||||
| Right | 60 | 0.109 | 0.112 | |||||
| Posterior | 60 | 0.112 | 0.113 |
P value is calculated using one-way analysis of variance.
Tukey’s test was used for post hoc test.
SD, standard deviation; a, anterior; l, left; r, right; p, posterior.
*P < 0.05, ***P < 0.001.
DISCUSSION
3D printing is a versatile technology that enables the rapid in-house production of appliances and auxiliaries, thereby reducing treatment time and office visits. This study represents a novel application of 3D-printed appliances, as there have been no previous attempts to design an auxiliary on top of an orthodontic appliance already in place.
Most deviations in this study fell within the 0.25 mm and 0.1 mm marginal limits (Supplementary Tables 2–4). However, deviations inevitably occur at various stages of fabrication, and the accumulation of these errors may lead to unexpected treatment outcomes. Therefore, users should be aware of the potential steps in which errors may arise, and aim to minimize them as much as possible. One common type of printing deviation in vat polymerization manufacturing is the staircase effect. Because 3D printing constructs objects layer by layer, the result is an uneven surface rather than a smooth one.18-20 Although, in theory, this issue could be mitigated by printing in increasingly thinner layers, studies have shown that a smaller layer thickness can lead to an accumulation of errors, resulting in greater distortion than expected.18,21,22
Another factor contributing to printing deviations is the printer parameters, including the exposure time, lifting height, lifting speed, down speed, and delay times.22,23 These parameters typically can be adjusted using the software provided by the printer manufacturer, however, only a few companies offer such programs. Unless the resin provided by the printer manufacturer is used, optimizing the 3D printer settings is necessary to reduce distortion, which can otherwise lead to printing failure. Various resin formulations available in the market contain different chemical compositions that alter light absorption, refraction, and other factors. Consequently, if a third-party resin is used, optimization is crucial to minimize unwanted distortions.22,23 The resins used in this study were approved by the manufacturer for use with a specific printer; therefore, minimal deviation was expected owing to the optimization.
Previous studies have explored the most effective angulation for achieving high accuracy. While some studies claim that there are no significant differences between different printing angles, the most accurate results were achieved when the objects were angled between 30° and 45°.18,19,22,23 However, these studies primarily focused on models or retainers that follow the natural curvature of the dental arch. By contrast, in this study, the ATOZ expander body and housing were rectangular with sharp edges. Therefore, during printing, the objects were oriented with flat surfaces parallel to the base. The base of the model was placed directly on the build plate and the housing of the PAD was aligned as horizontally as possible. Supporting structures were carefully added at locations that did not interfere with the measurements.
Before washing, excess resin was manually removed from the housing compartment using a small brush to ensure that the narrow spaces were adequately cleaned. Despite this effort, the corners of the housing exhibited uneven and less-defined edges. Although the surface quality was not part of the study’s measurement criteria, some areas where supports were attached displayed relatively large deviations, with the highest deviations measuring between 0.3 and 0.4 mm. These deviations were caused by the residual support material even after removal. Typically, the supporting structures are connected to an object via a ball-shaped end to facilitate its removal. Smaller ball sizes improve the ease of removal and leave fewer residues; however, they also reduce the retention force, thereby increasing the likelihood of distortion or printing failure. While these deviations may seem large, they are generally addressed during the polishing stage of production, and thus do not impact the final functionality of the appliance.
According to the International Organization for Standardization (ISO) standards, accuracy consists of trueness and precision. Trueness refers to the comparison between the printed product and the reference, whereas precision refers to the consistency of repeated prints.4,9,24,25 In this study, only trueness was defined as accuracy because although precision is important for production quality, usually when printing in-house settings, it is unlikely that the same material is printed in large quantities. The accuracy of 3D printers continues to be studied in various dental fields. However, there is no universally accepted value or range for defining accuracy. Some studies consider deviations over 0.5 mm as clinically unacceptable, while others suggest that 0.25 mm is the maximum acceptable limit.9,18,23 For this study, three different marginal limits were used to assess the accuracy of PAD printing: 0.5 mm, 0.25 mm, and 0.1 mm. The results demonstrated that all values were within the 0.25 mm range, with some even falling within the 0.1 mm limit, indicating very high accuracy (Supplementary Tables 1 and 3).
This study used only one type of 3D printer (a DLP), which limits its applicability to other technologies. Additionally, the 3D printer used in this study did not allow control over any parameters, and a resin from a different provider was used, preventing optimization for the resin. This may have led to deviations in the overall results. Further studies should investigate different resin types to determine the optimal material to achieve the highest printing accuracy. By refining printing techniques and optimizing materials, 3D printing can continue to improve treatment outcomes in dentistry.
CONCLUSIONS
We successfully fabricated and evaluated a novel PAD using 3D printing and digital superimposition analysis. The results demonstrated that the internal accuracy of the 3D-printed PADs remained within clinically acceptable limits (0.25 mm). Although the housing component exhibited more deviation than the wing component in the PAD-only group, the final adaptation results were within acceptable margins, with no patterns of distortion.
When evaluating different marginal limits, most deviations remained within 0.5 mm, with a majority falling within 0.25 mm and 0.1 mm. However, at the strictest 0.1 mm limit, deviations in the anterior and posterior regions exceeded the significance thresholds. Differences were also observed between the wing- and combination-type PADs, with the combination type showing more consistent deviation patterns across all axes.
Despite minor variations, the overall adaptation accuracy of the 3D-printed PADs remained clinically acceptable. PADs can be safely used in clinical applications without compromising the treatment outcomes. Future research should optimize the printing parameters and assess the clinical impact of deviations.
ACKNOWLEDGEMENTS
The authors want to show special thanks to Mr. Un Seob Shim, CEO of Graphy Company, Seoul for supporting the article preparation.
SUPPLEMENTARY MATERIAL
Supplementary data is available at https://doi.org/10.4041/kjod25.065
Footnotes
AUTHOR CONTRIBUTIONS
Conceptualization: JHK, JJP, SHK. Data curation: JHK, FPC, SYL. Formal analysis: JHK. Investigation: JHK, JJP, HJ. Methodology: JJP, FPC. Project administration: SHK. Resources: SYL. Software: JJP. Supervision: SHK. Validation: JJP, FPC, SYL. Visualization: JHK, HJ. Writing–original draft: JJP, HJ. Writing–review & editing: HJ, SHK.
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING
None to declare.
REFERENCES
- 1.Zhivago P. Three-dimensional printing in prosthodontics, restorative, and surgical dentistry. Dent Clin North Am. 2025;69:275–98. doi: 10.1016/j.cden.2024.11.008. https://doi.org/10.1016/j.cden.2024.11.008. [DOI] [PubMed] [Google Scholar]
- 2.Christopoulou I, Kaklamanos EG, Makrygiannakis MA, Bitsanis I, Perlea P, Tsolakis AI. Intraoral scanners in orthodontics: a critical review. Int J Environ Res Public Health. 2022;19:1407. doi: 10.3390/ijerph19031407. https://doi.org/10.3390/ijerph19031407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sulaiman TA. Materials in digital dentistry-a review. J Esthet Restor Dent. 2020;32:171–81. doi: 10.1111/jerd.12566. https://doi.org/10.1111/jerd.12566. [DOI] [PubMed] [Google Scholar]
- 4.Lee CKJ, Yong CW, Tan SL, Seah JA, Chew MT, Ren Y. Accuracy and clinical fit of milled versus rapid prototyped orthognathic surgical splints. J Stomatol Oral Maxillofac Surg. 2025;126:102069. doi: 10.1016/j.jormas.2024.102069. https://doi.org/10.1016/j.jormas.2024.102069. [DOI] [PubMed] [Google Scholar]
- 5.Lee J, Ju S, Kim J, Hwang S, Ahn J. The comparison of the accuracy of temporary crowns fabricated with several 3D printers and a milling machine. J Adv Prosthodont. 2023;15:72–9. doi: 10.4047/jap.2023.15.2.72. https://doi.org/10.4047/jap.2023.15.2.72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dawood A, Marti Marti B, Sauret-Jackson V, Darwood A. 3D printing in dentistry. Br Dent J. 2015;219:521–9. doi: 10.1038/sj.bdj.2015.914. https://doi.org/10.1038/sj.bdj.2015.914. [DOI] [PubMed] [Google Scholar]
- 7.Srivastava M, Rathee S, Patel V, Kumar A, Koppad PG. A review of various materials for additive manufacturing: recent trends and processing issues. J Mater Res Technol. 2022;21:2612–41. doi: 10.1016/j.jmrt.2022.10.015. https://doi.org/10.1016/j.jmrt.2022.10.015. [DOI] [Google Scholar]
- 8.Tigmeanu CV, Ardelean LC, Rusu LC, Negrutiu ML. Additive manufactured polymers in dentistry, current state-of-the-art and future perspectives-a review. Polymers (Basel) 2022;14:3658. doi: 10.3390/polym14173658. https://doi.org/10.3390/polym14173658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nestler N, Wesemann C, Spies BC, Beuer F, Bumann A. Dimensional accuracy of extrusion- and photopolymerization-based 3D printers: in vitro study comparing printed casts. J Prosthet Dent. 2021;125:103–10. doi: 10.1016/j.prosdent.2019.11.011. https://doi.org/10.1016/j.prosdent.2019.11.011. [DOI] [PubMed] [Google Scholar]
- 10.Xu B, Park JJ, Bai J, Kim SH. Effect of bone-borne maxillary skeletal expanders on cranial and circummaxillary sutures: a cone-beam computed tomography study. Korean J Orthod. 2024;54:346–58. doi: 10.4041/kjod24.180. https://doi.org/10.4041/kjod24.180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Si M, Hao Z, Fan H, Zhang H, Yuan R, Feng Z. Maxillary protraction: a bibliometric analysis. Int Dent J. 2023;73:873–80. doi: 10.1016/j.identj.2023.06.001. https://doi.org/10.1016/j.identj.2023.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bai J, Choi JY, Kim SH. The design of bone-borne maxillary expander affects the different dentoalveolar inclination and expansion pattern: a CBCT study. Semin Orthod. 2025;31:258–70. doi: 10.1053/j.sodo.2024.05.002. https://doi.org/10.1053/j.sodo.2024.05.002. [DOI] [Google Scholar]
- 13.Cifter M, Ekmen O, Gumru Celikel AD, Tagrikulu B, Erbay E. Does the face mask increase the impact of rapid maxillary expansion on sagittal airway dimensions? Eur Oral Res. 2023;57:28–35. doi: 10.26650/eor.20231133640. https://doi.org/10.26650/eor.20231133640. a1e7f5ca597742ea9333956cc46aa880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cordasco G, Matarese G, Rustico L, Fastuca S, Caprioglio A, Lindauer SJ, et al. Efficacy of orthopedic treatment with protraction facemask on skeletal Class III malocclusion: a systematic review and meta-analysis. Orthod Craniofac Res. 2014;17:133–43. doi: 10.1111/ocr.12040. https://doi.org/10.1111/ocr.12040. [DOI] [PubMed] [Google Scholar]
- 15.de Souza RA, Rino Neto J, de Paiva JB. Maxillary protraction with rapid maxillary expansion and facemask versus skeletal anchorage with mini-implants in class III patients: a non-randomized clinical trial. Prog Orthod. 2019;20:35. doi: 10.1186/s40510-019-0288-7. https://doi.org/10.1186/s40510-019-0288-7. 525b0122c08b4cb2970c808ad20a4bd3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cevidanes L, Baccetti T, Franchi L, McNamara JA, Jr, De Clerck H. Comparison of two protocols for maxillary protraction: bone anchors versus face mask with rapid maxillary expansion. Angle Orthod. 2010;80:799–806. doi: 10.2319/111709-651.1. https://doi.org/10.2319/111709-651.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bai J, Kim JH, Park JJ, Kim SH. Novel method for maxillary protraction following the purely bone-borne mini-screw assisted maxillary skeletal expansion. Semin Orthod. 2025;31:310–20. doi: 10.1053/j.sodo.2024.05.008. https://doi.org/10.1053/j.sodo.2024.05.008. [DOI] [Google Scholar]
- 18.Favero CS, English JD, Cozad BE, Wirthlin JO, Short MM, Kasper FK. Effect of print layer height and printer type on the accuracy of 3-dimensional printed orthodontic models. Am J Orthod Dentofacial Orthop. 2017;152:557–65. doi: 10.1016/j.ajodo.2017.06.012. https://doi.org/10.1016/j.ajodo.2017.06.012. [DOI] [PubMed] [Google Scholar]
- 19.Ko J, Bloomstein RD, Briss D, Holland JN, Morsy HM, Kasper FK, et al. Effect of build angle and layer height on the accuracy of 3-dimensional printed dental models. Am J Orthod Dentofacial Orthop. 2021;160:451–8.e2. doi: 10.1016/j.ajodo.2020.11.039. https://doi.org/10.1016/j.ajodo.2020.11.039. [DOI] [PubMed] [Google Scholar]
- 20.Young Kim RJ, Cho SM, Jung WS, Park JM. Trueness and surface characteristics of 3-dimensional printed casts made with different technologies. J Prosthet Dent. 2024;132:1324.e1–11. doi: 10.1016/j.prosdent.2022.12.002. https://doi.org/10.1016/j.prosdent.2022.12.002. [DOI] [PubMed] [Google Scholar]
- 21.Williams A, Bencharit S, Yang IH, Stilianoudakis SC, Carrico CK, Tüfekçi E. Effect of print angulation on the accuracy and precision of 3D-printed orthodontic retainers. Am J Orthod Dentofacial Orthop. 2022;161:133–9. doi: 10.1016/j.ajodo.2021.01.020. https://doi.org/10.1016/j.ajodo.2021.01.020. [DOI] [PubMed] [Google Scholar]
- 22.Piedra-Cascón W, Krishnamurthy VR, Att W, Revilla-León M. 3D printing parameters, supporting structures, slicing, and post-processing procedures of vat-polymerization additive manufacturing technologies: a narrative review. J Dent. 2021;109:103630. doi: 10.1016/j.jdent.2021.103630. https://doi.org/10.1016/j.jdent.2021.103630. [DOI] [PubMed] [Google Scholar]
- 23.Lohfeld S, Belnap B, Retrouvey JM, Walker MP. Effect of model body type and print angle on the accuracy of 3D-printed orthodontic models. Biomimetics (Basel) 2024;9:217. doi: 10.3390/biomimetics9040217. https://doi.org/10.3390/biomimetics9040217. af09b0a8d1994d15bed9337fc255d276 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nulty A. A comparison of trueness and precision of 12 3D printers used in dentistry. BDJ Open. 2022;8:14. doi: 10.1038/s41405-022-00108-6. https://doi.org/10.1038/s41405-022-00108-6. c227239d3c0c479bb3a05610db2785a7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kim SY, Shin YS, Jung HD, Hwang CJ, Baik HS, Cha JY. Precision and trueness of dental models manufactured with different 3-dimensional printing techniques. Am J Orthod Dentofacial Orthop. 2018;153:144–53. doi: 10.1016/j.ajodo.2017.05.025. https://doi.org/10.1016/j.ajodo.2017.05.025. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.