Comparative evaluation of dental measurements using 3D virtual models and 3D printed models

INTRODUCTION

Dental study models are fundamental tools in orthodontics for diagnosing malocclusions and formulating treatment plans.^[1] Conventional plaster casts, though widely used, are fragile, require extensive storage space, and are difficult to duplicate or share.^[2,3] To address these limitations, digital models were introduced in the late 1990s.^[1,4] OrthoCAD™ was among the first platforms to provide electronic models, reducing many of the problems associated with plaster casts.^[5] Subsequent studies confirmed that digital models can achieve accuracy comparable to conventional casts.^[6,7]

Intraoral scanners, such as the iTero® system, further advanced digital workflows by generating precise, powder-free scans. These virtual files can also be converted into tangible replicas using 3D printing technology, thereby combining the advantages of digital storage and communication with the familiarity of physical models.^[8-12] Previous investigations have demonstrated the reliability of intraoral scans and their applicability in analyses such as Bolton’s, supporting their clinical use.^[5-7,13]

Despite these advances, limited research has directly compared the accuracy of dental measurements across different digital platforms and 3D printed models. Considering the growing adoption of digital workflows in orthodontics, it is clinically important to establish whether these approaches provide interchangeable results with plaster casts.

Accordingly, the present study evaluates intercanine and intermolar widths, overjet, overbite, and Bolton’s ratios across three approaches: (i) digital models analyzed with GOM Inspect, (ii) digital models analyzed with OrthoCAD, and (iii) 3D printed models measured with calipers under magnification. The aim is to determine agreement and reproducibility among these methods, thereby providing evidence to guide the transition from plaster to digital and printed model workflows in orthodontic practice.

MATERIAL AND METHODS

Study groups

• Group A: 17 digital scans evaluated through GOM Inspect software [Figure 1]

  

  Close

  Figure 1:

  iTero scan.

  Export to PPT

• Group B: 17 digital scans evaluated through OrthoCAD software [Figure 2]

  

  Close

  Figure 2:

  OrthoCAD software.

  Export to PPT

• Group C: 17 corresponding 3D printed models evaluated using dental loupes and Vernier calipers [Figures 3 and 4].

Close

Figure 3:

3D printing process (a) ArchForm software, (b) CHITUBOX Slicer, (c) Phrozen Sonic Mighty 4k Printer and ANYCUBIC Wish and Cure Plus machine, and (d) 3D Printed models with supports attached on the platform.

Export to PPT

Close

Figure 4:

3D printed models.

Export to PPT

Measurement protocol

Dental parameters assessed

• Maxillary and mandibular inter-canine widths (cuspal tips of right and left canines) [Figure 5]

• Maxillary and mandibular intermolar widths (mesiobuccal cusp tips of right and left first molars) [Figure 6]

  

  Close

  Figure 5:

  2 point measurements of intercanine distance through different techniques (a) GOM software, (b) OrthoCAD software, and (c) Vernier caliper.

  Export to PPT

  

  Close

  Figure 6:

  2 point measurements through different techniques (a) GOM software, (b) OrthoCAD software, and (c) Vernier caliper.

  Export to PPT

• Overjet (horizontal distance between maxillary and mandibular central incisor edges) [Figure 7]

  

  Close

  Figure 7:

  2 point measurements of overjet through different techniques (a) GOM software, (b) OrthoCAD software, and (c) Vernier caliper.

  Export to PPT

• Overbite (vertical overlap of maxillary and mandibular central incisors perpendicular to the occlusal plane) [Figure 8]

  

  Close

  Figure 8:

  Two-point measurements of overbite through different techniques – (a) GOM inspect software, (b) OrthoCAD software, and (c) Manual.

  Export to PPT

• Bolton’s anterior and overall ratios (sum of mesiodistal widths of designated teeth) [Figure 9].

Close

Figure 9:

2 point measurements of mesiodistal width through different techniques (a) GOM software, (b) OrthoCAD software, and (c) Vernier caliper.

Export to PPT

Procedure

• In Groups A and B, measurements were taken directly on the virtual models using the respective software’s two-point linear measurement tool [Figures 6a-8a for GOM; Figures 6b-8b for OrthoCAD].

• In Group C, measurements were taken on printed models using a stainless-steel Vernier caliper (accuracy ± 0.01 mm) under 3.5× magnification dental loupes [Figures 6c-8c].

• Each measurement was performed 3 times by the same examiner on separate days, and the mean value was recorded.

Reliability assessment

• Intra-examiner reliability: Assessed by repeating measurements on 20% of samples after a 2-week interval (ICC = 0.98)

• Inter-examiner reliability: Assessed by a second calibrated examiner on the same 20% of samples (ICC = 0.97).

RESULTS

DISCUSSION

In this study, intercanine width, intermolar width, overjet, overbite, and Bolton’s overall and anterior ratios did not differ significantly across measurements from GOM, OrthoCAD, and 3D-printed models (all P > 0.05). Reproducibility was excellent (ICC = 0.98), indicating that both digital and printed workflows yield equivalent linear measurements suitable for orthodontic diagnosis and assessment.

Our findings are consistent with previous reports showing close agreement between digital and conventional methods. Bootvong et al.^[6] and Keating et al.^[5,14] demonstrated good to excellent concordance between virtual and plaster models, while Camardella et al.^[15] and Reuschl et al.^[16] found no significant differences in transverse or overjet/overbite measurements. Kasparova et al.^[17] and Wiranto et al.^[18] confirmed that digital and printed models can be used interchangeably for orthodontic evaluation. In contrast, some studies have highlighted potential variability: Bootvong et al.^[6] observed a tendency of intraoral scanning to slightly overestimate arch dimensions, and Al Jabri et al.^[19] reported differences in Bolton’s analysis depending on the model type. Leifert et al.^[20] also noted that methodological differences between digital and manual approaches can influence results.

These nuances align with our observation that OrthoCAD showed weaker correlations for sum-of-teeth and Bolton’s indices, likely reflecting differences in segmentation algorithms that affect mesiodistal tooth width detection.^[5,7,15]

From a clinical standpoint, these results support the use of digital and printed models for orthodontic diagnosis, space analysis, treatment planning, progress monitoring, and outcome assessment. Beyond measurement accuracy, digital workflows provide practical benefits. They are potentially more cost-effective, reducing the need for impression materials, plaster, and storage space, while also allowing long-term electronic archiving.^[8-11] They enhance time efficiency by streamlining record acquisition, enabling rapid transmission to laboratories, and reducing turnaround for appliance fabrication.^[12] Moreover, intraoral scanning improves patient comfort compared with conventional impressions, often leading to greater acceptance and cooperation. 3D-printed replicas, meanwhile, remain useful when a tangible reference or appliance fabrication model is required. Together, these advantages strengthen the case for incorporating digital workflows into everyday practice.

This investigation has certain limitations. It was conducted at a single center with a modest sample size, which may restrict generalizability. Although inclusion criteria controlled crowding to minimal–moderate levels, other malocclusion characteristics were not standardized. Inter-operator variability was assessed only in a subset of samples, and potential sources of error, such as scanning artifacts, STL trimming, printer calibration, resin shrinkage, and post-curing, were acknowledged but not quantified. Ensuring strict quality control in 3D printing and regular calibration of printers/scanners is essential to maintain accuracy and reproducibility in clinical settings. Furthermore, only linear two-point measurements were evaluated, and no patient-reported outcomes were assessed.

Nevertheless, the study has notable strengths, including examiner blinding, repeated measurements with excellent intra- and inter-examiner ICC values, and the use of a standardized 3D printing protocol with calibration and controlled post-processing. Evaluating both digital software and physical models under magnification also provides a comprehensive comparison of clinically relevant approaches.

Future research should include larger, multicenter studies encompassing severe crowding and complex malocclusions, as well as direct comparisons across scanners, printers, resins, and post-processing protocols. Algorithm-level evaluations are needed to clarify how segmentation affects Bolton’s analysis and tooth width detection. Importantly, prospective studies should incorporate economic assessments, workflow efficiency metrics, and patient-reported outcomes, since these factors strongly influence clinical adoption. The development of standardized quality-control guidelines for 3D printing (including calibration and shrinkage checks) would further enhance reliability.

Overall, digital and 3D-printed models demonstrated accuracy and reproducibility comparable to plaster casts for linear dental measurements. Clinicians may confidently integrate these workflows into practice while remaining cautious of software-specific variability, such as the weaker correlations observed with OrthoCAD. Until segmentation algorithms are optimized, verification with an alternative method is advisable when clinical decisions are borderline.

CONCLUSION

This study found no statistically significant differences in intercanine width, intermolar width, overjet, overbite, or Bolton’s overall and anterior ratios when comparing 3D virtual models (GOM, OrthoCAD) with 3D printed models, with all methods demonstrating excellent reproducibility (ICC = 0.98). These findings indicate that both digital and printed models can be used interchangeably in orthodontic diagnosis, space analysis, treatment planning, and progress assessment, offering a reliable alternative to conventional plaster casts. Given the comparatively weaker correlations observed for certain parameters in OrthoCAD, future work should focus on optimizing its segmentation algorithms to improve consistency with other digital workflows. Clinicians can confidently use either digital or printed models for routine orthodontic assessment, with awareness of OrthoCAD’s current limitations.