Accuracy of 3D Printing in Orthodontics: A Systematic Review and Meta-analysis

Abstract

Objective

To assess and compare the accuracy of 3D-printed digital models, orthognathic surgical splints, retainers or aligners, and implant surgical guides printed using various 3D printing technologies.

Methods

The search comprised prospective and retrospective studies related to the accuracy of 3D-printed models, splints, implant surgical guides, and retainers/aligners in patients undergoing orthodontic and orthognathic surgical procedures. The outcomes were assessed in terms of linear measurements, degree of fit, and positional deviations in comparison to conventional plaster models and surgical acrylic splints. The methodologic quality of the articles and the level of evidence were assessed using the QUADAS-2 tool. A meta-analysis was performed to compare the accuracy of models printed by different technologies to plaster models using the random-effects model.

Results

Twenty-one retrospective studies were included. Quality assessment of all included articles showed moderate risk of bias and no article was excluded. The systematic review showed that there were no significant differences between printed surgical splints, retainers/aligners compared, and the control group. Meta-analysis of six eligible studies showed that printed models had a general trend toward overestimation of linear measurements (mean difference = 0.22 mm; 95% confidence interval = – 0.09 to 0.36 mm, I² = 85%). Measurements made on digital light processing-printed models were significantly different than those made on plaster models.

Conclusion

The mean difference of 0.22 mm in linear measurements made on 3D-printed models and plaster models was clinically acceptable. Among the various printing technologies, PolyJet and fused deposition modeling-printed models showed higher accuracy. The printed splints and retainers/aligners were as accurate and reliable as their respective gold standards. The findings should be interpreted with caution due to high level of heterogeneity.

PROSPERO Registration Number: CRD42020175511

Keywords

3D printing additive manufacturing dental models systematic review stereolithography computer-aided design

Introduction

The system of three-dimensional (3D) printers, functioning together with intraoral scanners and model scanners, is a landmark innovation in orthodontics. Synonymous to additive manufacturing, “3D printing is a technology whereby sequential layers of material are deposited on top of one another to eventually form an object”.¹ Chuck Hull in 1984² first developed the 3D printing concept and in 1986 established the “3D Systems Company” with the first machine for rapid prototyping, which was called “stereolithography (SLA)”. Fused deposition modeling (FDM) technology, developed by Scott Crump in 1988, was widely marketed by Stratasys in 1990.³ Objet Geometries in 1988 introduced “PolyJet photopolymer printing (PPP)”. A multitude of 3D printers adopting variants of SLA, FDM, digital light processing (DLP), and PolyJet technologies are now accessible from various companies all over the world.

Rapid prototyping generates a physical reproduction of a digital model and thus overcomes the latter’s limitation of being intangible and is much more convenient to be used for complicated treatment planning procedures.

Virtual surgical planning for bimaxillary orthognathic surgery is faster and cheaper than standard planning.⁴ 3D-printed surgical splints seem to be a viable alternative to traditional acrylic splints. Lauren and McIntyre commercialized digital splints by fabricating the first computer-designed occlusal splints in the year 2008.⁵

Geometric inaccuracies are common and are often caused during the impression collection and thermoforming processes for the fabrication of aligners and retainers; this could be minimized through the utilization of digital technologies, 3D modeling, and rapid prototyping techniques.

Miniscrew insertion using surgical guides is becoming increasingly popular as it improves the stability and accuracy of implant insertion. In 47.4% of maxillary and 48.3% of mandibular miniscrews inserted using a direct manual technique, contact with or damage to anatomic structures surrounding the roots of teeth was observed, according to Kuroda et al.⁶ Current literature search revealed many studies that have compared the accuracy of 3D-printed models, surgical splints, retainers/aligners, and implant surgical guides to the current “gold standards” adopted in clinical practice such as plaster models, acrylic splints, and thermoformed retainers/aligners.

Thus, the objectives of this systematic review were to (a) determine and compare the reliability and accuracy of 3D digital models printed using various 3D printing technologies with those obtained through conventional oral impressions, intraoral scanning, and model scanning; (b) compare the accuracy of conventional acrylic surgical splints and 3D-printed surgical splints; (c) compare the accuracy of thermoformed retainers/aligners and 3D-printed clear aligners/retainers; and (d) check the accuracy of miniscrew implant placement using 3D-printed implant surgical guides.

Materials and Methods

Registration of Protocol

“The Preferred Reporting Items for Systematic Reviews and Meta-Analyses” (PRISMA) 2020 guidelines⁷ were followed in the execution of this systematic review. The review was registered with the International Prospective Register of Systematic Reviews (PROSPERO) under the registration number (CRD42020175511).

Eligibility Criterion and Study Selection

Information and Sources

A literature search was conducted till April 2023 in the electronic databases, PubMed Central (NCBI), Cochrane Library, Embase, Scopus, and Web of Science as the primary sources. The electronic search was supplemented with manual search in the reference lists of relevant articles. Gray literature was identified by searching OpenGrey and Google Scholar (Table 1, Supplementary Table 1).

Table 1.

PICOS Criteria.

Category	Inclusion Criteria	Exclusion Criteria
Population	Patients reporting for orthodontic and ortho-surgical treatment	Patients with genetic syndromes and severe facial malformations Tooth anomalies and severe interproximal caries or attrition/typhodont models Studies not involving human subjects Studies without original English translations
Intervention	1) 3D-printed models 2) 3D-printed surgical splints 3) 3D-printed aligners/retainers 4) Miniscrew implant surgical guides irrespective of printing technology used	Studies involving 2D representation of study models
Comparison	1) Conventional plaster/digital models 2) Acrylic surgical splints 3) Thermoformed retainers/clear aligners 4) Planned position of implant placement using CBCT/CT/freehand method	Studies without a control group
Outcome	1) Accuracy of arch and tooth dimension measurements in printed models 2) Accuracy of degree of fit in surgical splints 3) Accuracy of degree of fit of printed aligners/retainers 4) Positional deviation in miniscrew implant placement
Study design	Prospective and retrospective studies (case-control, cohort studies)Quasi-randomized, randomized controlled trials	Reviews Expert opinions Case reports and case series
Others	Articles in English language	Other languages

Note: Research question: Are 3D digital models, splints, retainers/aligners, and implant surgical guides printed using various 3D printing technologies as accurate and reliable as conventional plaster models, acrylic splints, thermoformed retainers/aligners, and planned position of implants in cone beam computed tomography (CBCT)?

Search Strategy

All the searches were independently conducted by two different investigators. The stepwise strategy involved screening of the articles received after electronic search by reading the title and abstract. After the removal of duplicates, the references of these articles were manually searched to include all related studies that have been missed in the preliminary search. Finally, selected articles were included after full text reading according to the inclusion and exclusion criteria. Lack of agreement at any stage was resolved by mutual discussion and consensus

Data Extraction

Each study was assessed separately for result derivation. The primary data collected from the included studies were (a) author, journal name, and year of publication; (b) sample size; (c) number of subjects in the intervention group and type of intervention; (d) number of subjects in the control group; (e) mean age if mentioned; (f) printer type; (g) outcome measures; (h) results; and (i) conclusions.

Wherever standard error was not originally reported in individual studies, it was computed using MD – 1.96SD = loa; MD + 1.96SD = LOA, SE = SD/√(n); MD: mean difference, SD: standard deviation, loa: lower limit of agreement, LOA: upper limit of agreement, SE: standard error, n: sample size.⁸

Data Analysis

“Review Manager (RevMan), Version 5.3, The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen, 2014” was used to statistically pool the data and conduct the meta-analysis. Using the inverse variance weighted approach, the MD and SE of the individual studies were combined.

The trials that had similar treatments, participants, and results were combined, and the results were expressed as an effect measure using the highest reported MDs in linear measurements and the 95% confidence intervals (CIs) that went along with them. A random-effects model was used to calculate the weighted mean difference (WMD) of the treatment effect. Sensitivity analysis was done wherever there was a lot of heterogeneity. Publication bias was inspected using the generated funnel plot.

Quality (Risk of Bias) Assessment of the Selected Studies

The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool⁹ (Table 2, Supplementary Table 2) was used to rate the methodologic quality of the articles and to assess the level of evidence. Quality assessment was done independently by two investigators. The quality assessment was based on patient selection, index test, reference standard, flow, and timing to measure the risk of bias and applicability concerns.

Table 2.

Results of the Methodological Quality Assessment According to QUADAS-2 Tool.

Sl. No.	Study	Risk of Bias				Result
Sl. No.	Study	Patient Selection	Index Test	Reference Standard	Flow and Timing	Result
1	Brown et al.¹⁰	+	?	+	+	Unclear
2	Camardella et al.¹¹	?	+	+	+	Unclear
3	Wan Hassan et al.¹²	?	+	–	+	High
4	Hazevald et al.¹³	–	?	+	+	High
5	Rebong et al.¹⁴	?	?	+	+	Unclear
6	Keating et al.¹⁵	?	?	+	+	Unclear
7	Kasparova et al.¹⁶	+	?	–	+	High
8	Zhang et al.¹⁹	–	?	?	+	High
9	Dietrich et al.¹⁸	?	?	?	+	Unclear
10	Camardella et al.¹⁷	?	?	–	+	High
11	Ho et al.²⁰	+	?	+	+	Unclear
12	Gateno et al.²¹	?	+	–	+	High
13	Shqaidef et al.²²	?	?	+	+	Unclear
14	Shaheen et al.²³	?	+	+	+	Unclear
15	Song and Baek²⁴	–	?	+	+	High
16	Nasef et al.²⁵	?	+	+	+	Unclear
17	Jindal et al.²⁶	?	–	?	+	High
18	Cole et al.²⁷	?	?	+	+	Unclear
19	Morea et al.²⁸	?	?	+	+	Unclear
20	Liu et al.²⁹	?	?	+	+	Unclear
21	Bae et al.³⁰	?	?	+	+	Unclear

Sl. No.	Study	Applicability Concerns			Result
Sl. No.	Study	Patient Selection	Index Test	Reference Standard	Result
1	Brown et al.¹⁰	+	+	+	Low
2	Camardella et al.¹¹	+	+	+	Low
3	Wan Hassan et al.¹²	+	+	+	Low
4	Hazevald et al.¹³	+	+	+	Low
5	Rebong et al.¹⁴	+	+	+	Low
6	Keating et al.¹⁵	+	+	+	Low
7	Kasparova et al.¹⁶	+	+	+	Low
8	Zhang et al.¹⁹	+	+	+	Low
9	Dietrich et al.¹⁸	+	+	+	Low
10	Camardella et al.¹⁷	+	+	+	Low
11	Ho et al.²⁰	+	+	+	Low
12	Gateno et al.²¹	+	+	+	Low
13	Shqaidef et al.²²	+	+	+	Low
14	Shaheen et al.²³	+	+	+	Low
15	Song and Baek²⁴	+	+	+	Low
16	Nasef et al.²⁵	+	+	+	Low
17	Jindal et al.²⁶	+	+	+	Low
18	Cole et al.²⁷	+	+	+	Low
19	Morea et al.²⁸	+	+	+	Low
20	Liu et al.²⁹	+	+	+	Low
21	Bae et al.³⁰	+	+	+	Low

Note: +, low risk of bias if all criteria were met; ?, unclear risk of bias if at least one criterion was assessed as having an unclear risk; –, high risk of bias (plausible bias that seriously weakens confidence in the results) if at least one criterion was assessed as having a high risk.

Results

Study Selection

An overview of the selection process is depicted in the PRISMA flow diagram (Figure 1). The search strategy yielded a total of 43 studies. 21 studies were finally included in this systematic analysis after two-stage selection criteria which included title and abstract reading followed by full text reading. 22 studies were excluded including case reports and noncomparative studies due to lack of consideration of linear parameters (Figure 1 and Supplementary Table 3). No randomized clinical trials were found pertaining to this topic.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Flow Diagram of Study Selection.

Risk of Bias Assessment in Included Studies (Table 2)

The most recurrent sources of bias were related to the patient selection processes and the lack of proper blinding procedures (Figure 2). No article was excluded after the QUADAS-2 tool application. A funnel plot, which is a scatterplot of each treatment effect (mean) by the standard deviation of the study’s precision, was used to analyze publication bias. Figure 3 shows all studies are of similar size and equally distributed on both sides. So, there is no need to repeat literature search. The symmetric funnel plot in this study shows the risk of bias by placing the most accurate studies at the top.

Figure 2.

Funnel Plot of Comparison of Overall Outcome between 3D-printed Models and Plaster Models.

Figure 3.

Risk of Bias for the Included Studies.

Table 3 shows descriptive data of the included studies. Out of 21 included studies, 7 assessed accuracy of 3D-printed models using digital caliper measurements on physical 3D-printed and plaster models.^10–16 Three studies evaluated accuracy through 3D superimposition of the digitized models.^17–19 Five studies assessed accuracy of printed surgical splints.^20–24 The accuracy of 3D-printed clear aligners or retainers^25–27 and accuracy of 3D-printed miniscrew implant insertion guides^28–30 were evaluated by three studies each.

Table 3.

Descriptive Data of Selected Studies.

Sl. No.	Author	Type of Study	Study Group	Control Group	Sample Size	Parameters Assessed and Methods of Assessment
1	Brown et al.¹⁰ (2018, Oklahoma)	Prospective	PolyJet-printed modelsDLP-printed models	Plaster models	Study: 120Control: 60Total: 180	Crown height, crown width, intercanine width, intermolar width, arch length measured on physical printed and plaster models
2	Camardella et al.¹¹ (2017, The Netherlands)	Prospective	SLA-printed models	Plaster models	28 patients(21–39 years)Study group: 56Control: Plaster models: 56Total: 112	Intercanine width, intermolar width measured on printed digital models and plaster models
3	Wan Hassan et al.¹² (2017, Kuala Lumpur, Malaysia)	Prospective	FDM-printed models	Stone models	Study group: 60Control: 60Total:120	Stone and printed models were measured using digital calipers for crown height, crown width, buccolingual thickness, and arch dimensions
4	Hazeveld et al.¹³ (2014, The Netherlands)	Prospective	DLP-printed models	Plaster models	Study group: 18Control:18Total: 36	Crown height and crown width were measured using digital calipers on plaster and printed models.
5	Rebong et al.¹⁴ (2018, Indianapolis)	Prospective	FDM-printed modelsSLA-printed modelsPolyJet-printed models	Plaster models	Study group: 24 each groupControl: Plaster models: 24Total: 96	Crown height of canine, intercanine width, intermolar width were measured using digital calipers on physical printed and plaster models
6	Keating et al.¹⁵ (2008, Cardiff)	Prospective	SLA-printed models	Plaster models	Study group: 2Control: 2Total: 4	X plane: intercanine widths, interpremolar widths, and first and second intermolar widths in both upper and lower archesZ plane: clinical crown height of teeth (5–5)Linear measurements were recorded on plaster models, printed models using digital caliper
7	Kasparova et al.¹⁶ (2013, Prague)	Prospective	FDM-printed modelsSLA-printed models	Plaster models	Study group: 20 (10 each)Control: 10Total: 30	X-plane: Intercanine distanceY-plane: The distance of tips of the canine and, mesiopalatal cusp of the first permanent molar in the jawZ-plane: The clinical crown height of canineMixed distance: Distance between mesial edge of the first incisor and cusp of canineLinear measurements made on models using digital caliper
8	Zhang et al.¹⁹ (2019, Shanghai, China)³⁴	Prospective	DLP-printed modelsSLA-printed models	Digital models	Study group: 24Control: Digital model: 1Total: 25	Difference in linear measurements between markers placed on digital models
9	Dietrich et al.¹⁸ (2017, Switzerland)	Prospective	PolyJet-printed modelsSLA-printed models	Digital models	Study group: 4 (2 each)Control: Digital models	Linear difference in position of landmarks placed on digital models following their superimposition
10	Camardella et al.¹⁷ (2017, Brazil and Netherlands)	Prospective	PolyJet-printed modelsSLA-printed models	Digital models	Study group: 40 (20 each group)Control: Digital models	Linear difference in position of landmarks following superimposition of digital models
11	Ho et al.²⁰ (2019, Taiwan)	Prospective	PolyJet-printed splints	Acrylic surgical splints	Study group: 30Control: 30Total: 60	Clinical degree of fit of splint checked in vivoFit: No interferences or the splint slides during fitting processUnfit: Interferences present during fitting procedure
12	Gateno et al.²¹ (2003, Houston, Texas)³⁷	Prospective	Digitized SLA-printed splints	Acrylic surgical splints	Study group: 7Control: 7Total: 14	Degree of fitness was measured in terms of air space between the splint and the tooth surface digitally
13	Shqaidef et al.²² (2014, Hong Kong)	Prospective	SLA-printed splints	Acrylic surgical splints	Study group: 10Control: 10Total: 20	Six landmarks were used for superimposition of mandibular model positions obtained using conventional and printed splints
14	Shaheen et al.²³ (2018, Belgium)	Prospective	PolyJet-printed splints	Acrylic surgical splints	Study group: 20Control: 20Total: 40	Change in position of mandibular casts were noted after superimposition and generating color coded maps
15	Song and Baek²⁴ (2009, South Korea)	Prospective	SLA-printed splints	Acrylic surgical splints	Study group: 1Control: 1Total: 2	“#11—X#11 Y#11 Z#11: Position of the midpoint of the incisor edge of the upper right central incisor in the transverse, vertical and anteroposterior coordinates respectively”“#16—X#16 Y#16 Z#16: Positions of the mesiobuccal cusp tip of the upper right first molar in the transverse, vertical and anteroposterior coordinates respectively”“#26—X#26 Y#26 Z#26: Positions of the mesiobuccal cusp tip of the upper left first molar in the transverse, vertical and anteroposterior coordinates respectively” were measured digitally using dial gauges
16	Nasef et al.²⁵ (2017, Cairo Egypt)	Prospective	SLA-printed retainers	Thermoformed retainers	Study group: 10Control: 10Total: 20	Intercanine width, interpremolar width, intermolar width, canine cusp tip to midline, canine cusp tip to MB cusp tip of 1st molar
17	Jindal et al.²⁶ (2019; Chandigarh, India, and Nottingham, England)	Prospective	SLA-printed aligners	Thermoformed retainers	Study group: 5Control: 5Total: 10	Tooth height measurement of all mandibular teeth (1–14)
18	Cole et al.²⁷ (2019, Virginia)	Prospective	Commercially available vacuum-formed retainers3D-printed aligners	Traditional vacuum-formed retainers	Study group: 18Control: 9Total: 27	The shortest available distance in millimeters between the surface of the reference and the surface of the comparison part at 18 reference points
19	Morea et al.²⁸ (2011, Brazil)	Prospective	SLA-printed surgical guides	Planned position on CBCT	Study group: 410 orthodontic mini implants	Distance between the following points on planned and executed implants:Point 1: Center of the coronal limit of the miniscrewPoint 2: Center of the miniscrewPoint 3: Center of the apical limit of the miniscrewPoint 4: Directional vector that passes through the long axis of the planned implant
20	Liu et al.²⁹ (2010, China)	Prospective	SLA-printed surgical guides	Planned position on CBCT	34 orthodontic mini implantsStudy group: 11	Linear and angular position deviations in the distomesial, vertical, buccopalatal, and apical areas were measured
21	Bae et al.³⁰ (2013, South Korea)	Prospective	SLA-printed surgical guides	Planned position on CBCT	Study group: 25	3D analysis program was used to determine the deviations in angle and location between the planned and actual miniscrew placements in the cadaver.Maxillae following superimposition:Angular deviation wrt the long axis of the miniscrews, distance deviation at the coronal and apical areas

Note: DLP, digital light processing; SLA, stereolithography; FDM, fused deposition modeling; CBCT, cone beam computed tomography.

Six studies^{10, 12–16} out of twenty-one studies with adequate recording of data and similar methodology were included for meta-analysis. The MD in linear measurements (crown height, crown width, intercanine width, and intermolar width) between printed models and plaster models was presented using forest plot. Subgroup analysis was done to assess the accuracy of different printing technologies.

Accuracy of Printed Models

Linear Measurements

The arch dimensions measured in the included studies involved intercanine width, intermolar width, and arch depth. Three studies^{12, 13, 16} assessed these parameters individually while other studies^{15, 17, 20} evaluated all parameters combined as arch dimensions. Arch depth was measured as a separate entity only by Brown et al.¹⁰ where the MD between plaster and printed models ranged from –0.37 to 0.63 mm for DLP and –0.32 to 0.55 mm for PolyJet printers.

Figure 4 shows forest plot of the combined results of the meta-analysis. The overall results showed significant difference between measurements made on printed models compared to plaster models (MD = 0.22 mm; 95% CI = 0.09–0.36 mm). However, this outcome showed high heterogeneity (I² = 85%).

Figure 4.

Forest Plot Depicting Comparison between Accuracy of Printed Models and Plaster Models.

Subgroup analysis showed an overestimation of measurements by PolyJet models (MD = –0.07 mm; 95% CI = –0.09 to 0.23, I² = 30%), SLA models (MD = 0.32 mm; 95% CI = 0.08–0.57, I²= 68%), DLP-printed models (MD = 0.15 mm; 95% CI = –0.13 to 0.44, I² = 92%), and FDM models (MD = 0.11 mm; 95% CI = –0.08 to 0.31, I² = 55%). Forest plot shows only significant results for SLA versus plaster models (p = 0.010), although with heterogeneity of 68%. All these findings show high heterogeneity and hence should be considered with caution (Figure 5).

Figure 5.

Subgroup Analysis.

Accuracy of 3D-printed Splints

All studies showed acceptable accuracy of 3D-printed splints compared to conventional splints.^20–24 The difference in only the vertical parameters showed clinical significance (Table 4).²³

Accuracy of Printed Aligners and Retainers

Cole et al.²⁷ reported that the traditional vacuum-formed (TVF) retainers had the smallest deviation from their reference models ranging from 0.10 to 0.20 mm as compared to 0.10 to 0.40 mm of deviation for the printed retainers. Jindal et al.²⁶ reported that the printed aligners showed a deviation of 2.55% (0.21 mm) and the thermoformed aligner, 4.41% (0.37 mm) from their parent STL file. Nasef et al.²⁵ found no clinically significant differences between thermoformed and printed retainers except for the distance of the left mandibular canine from the midline (Table 4).

Accuracy of Printed Miniscrew Insertion Guides

Planned and actual miniscrew positions were compared in three dimensions using cone beam computed tomography (CBCT). Morea et al.,²⁸ Bae et al.,³⁰ and Liu et al.²⁹ showed that the linear and angular deviations between planned and actual miniscrew positions were statistically significant (Table 4).

Discussion

In this systematic review, strict selection criteria were used to include the articles related to 3D-printed models, surgical splints, and retainers/aligners with the objective of evaluating the accuracy of 3D printing technologies. Funnel plot was used to analyze the publication bias. The upper portion of the plot shows the studies with lowest standard deviation and are the most accurate. The graph should have a shape of an inverted funnel which is symmetrical in the absence of bias and heterogeneity. In this study the funnel plot is symmetrical with the most accurate studies at the top of the plot.31 The six studies to assess the accuracy of printed models and conventional plaster models as control were included in the meta-analysis. The findings suggest that there was a significant difference in the linear measurements made on the printed 3D models compared to plaster models. Whereas subgroup analysis showed significant differences only for SLA compared to plaster models. The accuracy of all other printed models was statistically similar to plaster models. The accuracy of 3D-printed surgical splints, retainers, and implant surgical guides is also comparable to their respective gold standards. However, the reliability of the results is questionable due to the high degree of heterogeneity.

The quality assessment tool (QUADAS-2) showed that subject selection method was not random in most of the included studies. The methods used to perform the measurements were very clear and comprehensive enough to allow the reviewer to answer the review question in all studies. However, in only two studies, index tests were evaluated independent of the reference standard. Sample sizes in the studies included were small and the interpretations of results were also diverse. Thus, the lack of randomization and blinding may have introduced selection bias. Small samples and diverse results may have contributed to the high heterogeneity.

Accuracy of Printed Models

In terms of the printing process, the accuracy of the constructed models may be affected by the accuracy of the machine, the materials used, and the subsequent model shrinkage during building and postcuring.³² The combined MD of linear measurements between printed models and plaster models according to the present meta-analysis is 0.22 mm (0.09–0.36 mm) which is in the clinically acceptable range of 0.20–0.50 mm.^{9, 12, 33, 34}

The results of the present meta-analysis revealed that the MD between DLP, SLA, PolyJet and FDM-printed models, and plaster models were 0.15, 0.32, 0.07, and 0.11 mm, respectively. This shows that PolyJet printers were slightly more accurate than the rest of the printing technologies, though all of them were accurate enough for clinical use.^{9, 12, 33, 34}

Dental models printed with the PolyJet technology are fully cured during the building process and do not need a postcuring process, which leads to a faster and more efficient printing process. The only disadvantage of this type of printer is its greater cost of printing and wastage of the printing resin. Our meta-analysis showed a significant overestimation of linear measurements in SLA-printed models than the other printing technologies as the SLA build technology requires postcuring the printed parts to improve the stability of the printed object; this residual polymerization affects the overall accuracy of the model as supported by almost every study which evaluated this type of printer.^{14–16, 22} DLP printers were reported to be faster than SLA printers by Zhang et al.¹⁹ and this might be due to the use of light projectors in DLP printers that can cure a much larger area while in SLA printing technology, the laser source cures only a small area of the resin covering the build tray.

Accuracy of Printed Splints

Irrespective of the printing technology used, virtual planning of orthognathic surgery and printed splints generated using the STL files of the patients’ occlusion are accurate enough for surgical use and can replace acrylic splints. The difference between conventional model occlusion and virtual occlusion was found to be in the range of 0.45–0.6 mm,^{20, 35} which is within 1.5 mm that is considered to be tolerable positional variation between the virtual intermaxillary position and intraoperative intermediate intermaxillary relationship according to Hernández-Alfaro and Guijarro-Martínez.³⁶ Three out of five comparative studies included in this review reported that there exists no significant difference between printed and acrylic splints while the rest reported clinically significant differences. The difference in maxillary position registered using conventional and 3D digital splints was between 0.18 and 1.73 mm.^{22, 24, 37}

Accuracy of Printed Aligners and Retainers

The direct 3D printing of clear orthodontic retainers has not been investigated much in the literature. Zilberman et al.³² did not find any statistically significant differences between the selective laser sintering (SLS) printed and thermoformed retainers. This could be because the study’s measurements were small. In the study by Cole et al., the TVF retainers were used as a benchmark to assess and contrast the fit of the 3D-printed retainers.²⁷ The printed retainers deviated from the reference model the most, up to 0.40 mm, with the majority of the deviations occurring at the reference points that were on the teeth’s smooth surfaces. As opposed to this study, Jindal et al.²⁶ concluded that cured aligners that were 3D-printed exhibited superior dimensional accuracy and compressive mechanical strength in comparison with uncured 3D-printed aligners and thermoformed aligners.

Accuracy of Printed Miniscrew Insertion Guides

All three studies^28–30 that evaluated the accuracy of miniscrew placement using printed surgical guides unanimously concluded that it was a superior and more accurate method of implant insertion. All the included studies suggest that root contact was rare and that the miniscrews could be stably placed with a surgical guide.

Limitations

The methodological quality of most of the included studies in this systematic review was moderate due to the lack of randomization of samples and lack of blinding procedures. Randomized clinical trials with proper blinding to compare the accuracy of printed models/splints/miniscrew surgical guides/retainers/aligners should be conducted.

Conclusion

Though the differences in linear measurements made to determine the accuracy of 3D-printed models and plaster models (mean: 0.22 mm) were significant but these differences were clinically acceptable for all the printers. However, the reliability of the results is questionable due to high heterogeneity. 3D-printed surgical splints were equal to acrylic splints in terms of accuracy and resultant positional variation of skeletal bases during orthognathic surgery. The accuracy of 3D-printed aligners/retainers when compared to thermoformed retainers had conflicting results. There is a need for further studies on a larger sample size. 3D-printed implant surgical guides were found to be accurate and reliable for precise miniscrew insertion.

Table 4.

Results of Qualitative Synthesis of Selected Studies.

	Author	Parameters Assessed and Results							Conclusion
1	Brown et al.¹⁰ (2018)	Comparison of groups	Mean difference in parameters between control and study group						All recorded measurements on PolyJet-printed models were closer to the stone modelsWhile both DLP and PolyJet-printed models showed greater crown width and arch dimensions, only PolyJet showed greater crown height measurementsThe accuracy of PolyJet and DLP-printed models was found to be clinically acceptable
			Crown height (mm)	Crown width (mm)	Intercanine width (mm)	Intermolar width (mm)	Arch length/depth (mm)
		Plaster models vs. DLP-printed models	0.29 (–1.04 to 0.46)	–0.06 (–0.2 to 0.43)	0.06 (–0.61 to 0.49)	0.07 (–0.61 to 0.47)	–0.13 (–0.37 to 0.63)
		Plaster models vs. PolyJet-printed models	–0.08 (–0.57 to 0.73)	–0.05 (–0.33 to 0.43)	0.04 (–0.48 to 0.41)	0.01 (–0.51 to 0.49)	–0.11 (–0.32 to 0.55)
2	Camardella et al.¹¹	Mean difference in intercanine width (mm)			Mean difference in intermolar width (mm)				Posterior transverse dimensions measured on SLA-printed digital models were significantly lower (<0.50) than plaster models
		0.321 ± 0.4347			0.682 ± 0.572
3	Wan Hassan et al.¹² (2017)	Clinical crown height plane	Mesiodistal (crown width) plane	Buccolingual plane	Arch dimensions plane				FDM-printed models showed larger measurements in the mesiodistal plane and smaller in the buccolingual dimensions irrespective of the degree of crowdingCrown height measurements were significantly smaller and arch dimension measurements were significantly greater than the measurements on plaster models
		0.00 ± 0.28 (–0.55 to 0.56)	–0.11 ± 0.29 (–0.68 to 0.45)	0.15 ± 0.28 (–0.41 to 0.71)	–0.02 ± 0.39 (–0.79 to 0.75)
4	Hazeveld et al.¹³ (2014)	Mean difference in clinical crown height (6-6)				Mean difference in crown width (6-6)			DLP and jetted photopolymer-printed models showed measurements similar to plaster models. Crown height measurements made on powder-based 3D-printed models showed statistically significant differences compared to plaster models but these differences were clinically acceptable
		0.04 (–0.26 to 0.34)				–0.05 (–0.45 to 0.34)
5	Rebong et al.¹⁴ (2018)	Mean difference of clinical crown height of canine		Intercanine width (mm)		Intermolar width (mm)			FDM models had the closest accuracy to plaster models while SLA and PolyJet models had greater transverse dimensions and lower vertical dimensions than plaster models with values within 0.35 mm
		0.07 ± 0.21 mm	0.11 ± 0.19 mm	0.14 ± 0.42 mm	0.22 ± 0.41 mm	–0.11 ± 0.42 mm	–0.30 ± 0.36 mm
		0.07 ± 0.18 mm	0.16 ± 0.24 mm	–0.29 ± 0.48 mm	–0.34 ± 0.40 mm	0.02 ± 0.37 mm	0.12 ± 0.47 mm
6	Keating et al.¹⁵ (2008)	Mean difference in linear measurements between FDM/SLA/PolyJet-printed models and plaster casts <0.35 mm							All the measurements of SLA-printed models were generally smaller compared to conventional plaster models. Thus, the accuracy of SLA-printed models was considered inferior for clinical use
		Z plane: 0.42 ± 0.23 mm			X axis: 0.15 ± 0.16 mm
		Mean difference in plaster and printed model measurements in X, Y, Z planes: 0.26 ± 0.22 mm
7	Kasparova et al.¹⁶ (2013)	Mean difference in crown height: 0.03 mm (–0.16 to 0.09)			Mean difference in intercanine width: –0.17 mm (–0.35 to 0.01)				The distance and the precision of the measurements on the FDM-printed models were similar to those on plaster models
8	Zhang et al.¹⁹ (2019)	Nine markers were selected for superimposition with uniformity in both the arches							DLP printers were found to be faster and more accurate than SLA printerEvoDent 50 µm (DLP) had the highest printing accuracy, while Form 2 (SLA) 100 µm showed the lowest printing accuracy
		Mean difference between reference points in maxilla			Mean difference between reference points in mandible
		DLP: 0.0354 ± 0.0502 mm			0.0486 ± 0.0662 mm
		SLA: 0.0511 ± 0.0709 mm			0.0570 ± 0.0826 mm
9	Dietrich et al.¹⁸ (2017)	Printed replicas were scanned digitally and the files obtained were superimposed on the source files for measurement of trueness. Precision was measured by superimposing the STL files of PolyJet and SLA replicas on themselves (multiple replicas were made)							PolyJet-printed models were more true while the SLA models were more precise with both the printing technologies suitable for clinical use
		Trueness			Precision
		109 ± 4 µm			92 ± 23 µm
		66 ± 4 µm			62 ± 8 µm
10	Camardella et al.¹⁷(2017, Brazil and Netherlands)	Average differences in measurements upon superimposition of digitized models <0.01 mm for both SLA and PolyJet-printed models							Printed dental models using the PolyJet and SLA printing techniques are accurate, regardless of the model base design
			Intercanine width	Interfirst premolar width	Intersecond premolar width	Interfirst molar widths	Intersecond molar width
		SLA	32.78 ± 2.33	40.77 ± 2.99	46.10 ± 2.60	50.49 ± 3.74	56.84 ± 2.84
		PolyJet	33.01 ± 2.24	40.71 ± 2.80	46.18 ± 2.58	50.69 ± 3.72	56.86 ± 4.11
11	Ho et al.²⁰ (2019, Taiwan)	Splint was judged to be fit or unfit2/30 conventional splints were unfit while 0/30 printed splints were unfit (fitness was indicated by no interference or sliding when trying the splint; otherwise, it was regarded as unfit)							No significant differences existed between printed and acrylic splints and both were found to be suitable for surgical use
12	Gateno et al.²¹ (2003, Houston, Texas)³⁷	1 mm thick slices were made through the long axis of the teethDifference in airspace between teeth and splint in two methods: 0.2 ± 0.2 mmsq (–0.3 to 1.2 mmsq)							SLA-printed splints made using the authors’ CAD technique were almost similar to the conventional splints in terms of accuracy
13	Shqaidef A et al.²²(2014, Hong Kong)	Absolute difference between two lower model landmarks when either wafers were used to articulate the upper and lower models: 0.94 (0.04–1.73) mm							3D-printed wafers with an absolute error within the range of 0.04–1.73 mm were considered not suitable for surgical use as an error of 1.73 mm will affect mandibular movement
14	Shaheen et al.²³ (2018, Belgium)	19/20 splints accurate fit and clinically acceptable1 splint posterior dimension difference of 1 mm							Printed splints using PolyJet technology were found to be accurate enough for clinical use
15	Song and Baek²⁴ (2009, South Korea)	Conventional (mm)			3D printing (mm)				Accuracy of SLA-printed splints needs to be improved
		X#11	79.05 ± 2.27		79.10 ± 2.36
		Y#11	81.35 ± 3.48		81.14 ± 3.72
		Z#11	92.25 ± 5.93		92.04 ± 6.05
		X#16	102.57 ± 1.74		102.48 ± 2.31
		Y#16	76.45 ± 2.80		75.91 ± 3.19
		Z#16	66.89 ± 5.07		67.23 ± 5.76
		X#11	5.31 ± 3.48		5.57 ± 3.77
		Y#16	6.56 ± 3.51		76.64 ± 4.10
		Z#26	69.77 ± 6.97		69.39 ± 7.20
16	Nasef et al.²⁵ (2017, Cairo, Egypt)	Mean Difference in mm							3D-printed retainers are as accurate and reliable as thermoform retainers
		Intercanine width	Interpremolar width	Intermolar width	Canine cusp tip to midline	Canine cusp tip to MB cusp tip of 1st molar
		0.046 ± 0.011	First PM: 0.0181 ± 0.14Second PM: 0.096 ± 0.22	0.08 ± 0.24	Left: 0.06 ± 0.07Right: 0.19 ± 0.07	Left: 0.27 ± 0.12Right: 0.18 ± 0.11
17	Jindal et al.²⁶ (2019, Chandigarh, India and Nottingham, England)	Compression testing was conducted utilizing a universal testing machine and the behavior of aligner deformation with load application was obtained using the compatible data acquisition softwareAverage deviation from reference models for 3D-printed retainer was 0.21 mm and for thermoformed retainer was 0.37 mm							3D-printed aligners were high yielding, higher load resisting, and lower deforming compared to thermoformed counterparts
18	Cole et al.²⁷ (2019, Virginia)	The comparison tool of the Netfabb software foundAverage deviation from reference models for thermoformed retainers was 0.10–0.20 mm; for commercially vacuum-formed retainers was 0.10–0.30 mm and for SLA-printed retainers was 0.40 mm							TVF retainers showed the least amount of deviation from the original reference modelsThe 3D-printed retainers showed the greatest deviation at the reference points located on the smooth surfaces of the teeth but showed close adaptation at the incisal edges and cusp tips and seemed to be similar to the TVF in fit
19	Morea et al.²⁸ (2011, Brazil)	Differences in the linear and angular measurements between planned and actual position of the miniscrews							SLA surgical guides built using CBCT data for orthodontic miniscrew insertion represent a precise technique that allows for safe insertion, nearly eliminating possible placement complications and errors and matching planning with orthodontic requirements
		Coronal distance	Apical distance	Central distance	Angle
		0.86 mm	0.87 mm	0.71 mm	1.76°
20	Liu et al.²⁹ (2010, China)	Distomesial at tip		Vertical at tip		Bucco-palatal at tip		Placement sitelinear	The average measured deviations of the miniscrews were in the safe range and each deviation was smaller than the allowed one. These results verified the template’s accuracy and security
		Linear (mm)	Angular	Linear (mm)	Angular	Linear (mm)	Angular
		0.42 ± 0.13	1.2 ± 0.43	0.47 ± 0.12	1.3 ± 0.41	0.59 ± 0.26	1.6 ± 0.79	0.10 ± 0.012 mm
21	Bae et al.³⁰ (2013, South Korea)	Coronal (mm)		Apical (mm)		Angular			Miniscrews were placed more accurately when surgical guides were used than when a direct method was used.Miniscrews placed with surgical guides did not cause damage to surrounding anatomic structures
		Surgical guide group	Control group	Surgical guide group	Control group	Surgical guide group	Control group
		0.73 mm (0.26–1.12 mm)	1.56 mm (0.59–2.95 mm)	0.73 mm (0.26–1.12 mm)	1.28mm (0.26–3.81 mm)	3.14 (1.02°–10.9°)	9.57° (3.15°–35.60°)

Note: DLP, digital light processing; SLA, stereolithography; FDM, fused deposition modeling; CAD, computer-aided design; TVF, traditional vacuum formed; CBCT, cone beam computed tomography.

Footnotes

Availability of Data and Materials

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Supplementary Material

ORCID iDs

Anjana Rajagopalan

Vinay Kumar

References

Groth

, Kravitz

, Jones

, Graham

, Redmond

WR.

Three-dimensional printing technology. J Clin Orthod . 2014;48(8):475–485.

Nguyen

, Lewis

, Valley

, Tarnoff

HL.

Systems D. View . Van Nuys 1991:114.

Crump

, Muir

AEPD.

Creating Three-dimensional Objects . 1992:15.

Resnick

, Inverso

, Wrzosek

, Padwa

, Kaban

, Peacock

ZS.

Is there a difference in cost between standard and virtual surgical planning for orthognathic surgery?

J Oral Maxillofac Surg . 2016;74(9):1827–1833. doi:10.1016/j.joms.2016.03.03

Lauren

, McIntyre

A new computer-assisted method for design and fabrication of occlusal splints. Am J Orthod Dentofacial Orthop . 2008;133(4):S130–S135. doi:10.1016/j.ajodo.2007.11.01

Kuroda

, Sugawara

, Deguchi

, Kyung

, Takano-Yamamoto

Clinical use of miniscrew implants as orthodontic anchorage: Success rates and postoperative discomfort. Am J Orthod Dentofac Orthop . 2007;131(1):9–15. doi:10.1016/j.ajodo.2005.02.03

Page

, McKenzie

, Bossuyt

, . The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ . 2021;372:n71. doi:10.1136/bmj.n7

Giavarina

Understanding Bland Altman analysis. Biochem Medica . 2015;25(2):141–151. doi:10.11613/BM.2015.01

Whiting

, Rutjes

AWS

, Westwood

, . QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med . 2011;155(8):529–536. doi:10.7326/0003-4819-155-8-201110180-0000

10.

Brown

, Currier

, Kadioglu

, Kierl

JP.

Accuracy of 3-dimensional printed dental models reconstructed from digital intraoral impressions. Am J Orthod Dentofacial Orthop . 2018;154(5):733–739. doi:10.1016/j.ajodo.2018.06.00

11.

Camardella

, Vilella

, van Hezel

, Breuning

KH.

Accuracy of stereolithographically printed digital models compared to plaster models. J Orofac Orthop Fortschritte Kieferorthopädie . 2017;78(5):394–402. doi:10.1007/s00056-017-0093

12.

Wan Hassan

, Yusoff

, Mardi

NA.

Comparison of reconstructed rapid prototyping models produced by 3-dimensional printing and conventional stone models with different degrees of crowding. Am J Orthod Dentofacial Orthop . 2017;151(1):209–218. doi:10.1016/j.ajodo.2016.08.01

13.

Hazeveld

, Huddleston Slater

JJR

, Ren

Accuracy and reproducibility of dental replica models reconstructed by different rapid prototyping techniques. Am J Orthod Dentofacial Orthop . 2014;145(1):108–115. doi:10.1016/j.ajodo.2013.05.01

14.

Rebong

, Stewart

, Utreja

, Ghoneima

AA.

Accuracy of three-dimensional dental resin models created by fused deposition modeling, stereolithography, and Polyjet prototype technologies: A comparative study. Angle Orthod . 2018;88(3):363–369. doi:10.2319/071117-460

15.

Keating

, Knox

, Bibb

, Zhurov

AI.

A comparison of plaster, digital and reconstructed study model accuracy. J Orthod . 2008;35(3):191–201. doi:10.1179/14653120722502262

16.

Kasparova

, Grafova

, Dvorak

, . Possibility of reconstruction of dental plaster cast from 3D digital study models. Biomed Eng Online . 2013;12(1):49. doi:10.1186/1475-925X-12-4

17.

Camardella

, de Vasconcellos Vilella

, Breuning

Accuracy of printed dental models made with 2 prototype technologies and different designs of model bases. Am J Orthod Dentofacial Orthop . 2017;151(6):1178–1187. doi:10.1016/j.ajodo.2017.03.01

18.

Dietrich

, Ender

, Baumgartner

, Mehl

A validation study of reconstructed rapid prototyping models produced by two technologies. Angle Orthod . 2017;87(5):782–787. doi:10.2319/01091-727

19.

Zhang

, Li

, Chu

, Shen

Influence of the three-dimensional printing technique and printing layer thickness on model accuracy. J Orofac Orthop Fortschritte Kieferorthopädie . 2019;80(4):194–204. doi:10.1007/s00056-019-00180

20.

, Lin

, Lo

LJ.

Intraoral scanning and setting up the digital final occlusion in three-dimensional planning of orthognathic surgery: Its comparison with the dental model approach. Plast Reconstr Surg . 2019;143(5):1027e–1036e. doi:10.1097/PRS.000000000000555

21.

Gateno

, Xia

, Teichgraeber

, Rosen

, Hultgren

, Vadnais

The precision of computer-generated surgical splints. J Oral Maxillofac Surg . 2003;61(7):814–817. doi:10.1016/S0278-2391(03)00240

22.

Shqaidef

, Ayoub

, Khambay

BS.

How accurate are rapid prototyped (RP) final orthognathic surgical wafers? A pilot study. Br J Oral Maxillofac Surg . 2014;52(7):609–614. doi:10.1016/j.bjoms.2014.04.01

23.

Shaheen

, Coopman

, Jacobs

, Politis

Optimized 3D virtually planned intermediate splints for bimaxillary orthognathic surgery: A clinical validation study in 20 patients. J Cranio-Maxillofac Surg . 2018;46(9):1441–1447. doi:10.1016/j.jcms.2018.05.05

24.

Song

, Baek

SH.

Comparison of the accuracy of the three-dimensional virtual method and the conventional manual method for model surgery and intermediate wafer fabrication. Oral Surg Oral Med Oral Pathol Oral Radiol Endodontology . 2009;107(1):13–21. doi:10.1016/j.tripleo.2008.06.00

25.

Nasef

, El-Beialy

, Eid

FHK

, Mostafa

YA.

Accuracy of orthodontic 3D printed retainers versus thermoformed retainers. Open J Med Imaging . 2017;7(4):169–179. doi:10.4236/ojmi.2017.7401

26.

Jindal

, Juneja

, Siena

, Bajaj

, Breedon

Mechanical and geometric properties of thermoformed and 3D printed clear dental aligners. Am J Orthod Dentofac Orthop . 2019;156(5):694–701. doi:10.1016/j.ajodo.2019.05.01

27.

Cole

, Bencharit

, Carrico

, Arias

, Tüfekçi

Evaluation of fit for 3D-printed retainers compared with thermoform retainers. Am J Orthod Dentofacial Orthop . 2019;155(4):592–599. doi:10.1016/j.ajodo.2018.09.01

28.

Morea

, Hayek

, Oleskovicz

, Dominguez

, Chilvarquer

Precise insertion of orthodontic miniscrews with a stereolithographic surgical guide based on cone beam computed tomography data: A pilot study. Int J Oral Maxillofac Implants . 2011;26(4):860–865.

29.

Liu

, Liu

, Wang

, Zhao

Accuracy of surgical positioning of orthodontic miniscrews with a computer-aided design and manufacturing template. Am J Orthod Dentofacial Orthop . 2010;137(6):728. e1–728.e10. doi:10.1016/j.ajodo.2009.12.02

30.

Bae

, Kim

, Park

, . Accuracy of miniscrew surgical guides assessed from cone-beam computed tomography and digital models. Am J Orthod Dentofacial Orthop . 2013;143(6):893–901. doi:10.1016/j.ajodo.2013.02.01

31.

Susin

, Haas

, Rösing

Understanding and interpreting systematic review and meta-analysis results. Published 2010. Accessed January 4, 2022. https://www.semanticscholar.org/paper/Understanding-and-Interpreting-Systematic-Review-Susin-Haas/5260a2ad3f10541d1350a496bf522751cf8de0f7

32.

Zilberman

, Huggare

JAV

, Parikakis

KA.

Evaluation of the validity of tooth size and arch width measurements using conventional and three-dimensional virtual orthodontic models. Angle Orthod . 2003;73(3):301–306. doi:10.1043/0003-3219(2003)073<0301:EOTVOT>2.0.CO

33.

Halazonetis

DJ.

Acquisition of 3-dimensional shapes from images. Am J Orthod Dentofac Orthop . 2001;119(5):556–560. doi:10.1067/mod.2001.11545

34.

Bell

, Ayoub

, Siebert

Assessment of the accuracy of a three-dimensional imaging system for archiving dental study models. J Orthod . 2003;30(3):219–223. doi:10.1093/ortho/30.3.21

35.

Nadjmi

, Mollemans

, Daelemans

, Van Hemelen

, Schutyser

, Bergé

Virtual occlusion in planning orthognathic surgical procedures. Int J Oral Maxillofac Surg . 2010;39(5):457–462. doi:10.1016/j.ijom.2010.02.00

36.

Hernández-Alfaro

, Guijarro-Martínez

New protocol for three-dimensional surgical planning and CAD/CAM splint generation in orthognathic surgery: An in vitro and in vivo study. Int J Oral Maxillofac Surg . 2013;42(12):1547–1556. doi:10.1016/j.ijom.2013.03.02

37.

Kim

, Oh

, Cevidanes

LHS

, . Analysis of 3D soft tissue changes after 1- and 2-jaw orthognathic surgery in mandibular prognathism patients. J Oral Maxillofac Surg . 2013;71(1):151–161. doi:10.1016/j.joms.2012.02.00

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.25 MB