From Digital Deviation to Clinical Fit: Do Inter-Scanner Discrepancies Predict Passive-Fit Differences in Full-Arch Implant Rehabilitation?

Contemporary intraoral scanners (IOSs) have expanded the possibilities of direct digital implant impressions, yet complete-arch implant scanning remains challenging.^3,4 Edentulous arches provide limited continuous anatomical landmarks for image registration, and small registration discrepancies may accumulate over long scanning spans, particularly toward terminal implant positions.^5,6 Accuracy may also be influenced by acquisition technology, scanning strategy, operator experience, scanbody geometry and seating, number and distribution of implants, mesh-to-library congruence, and reconstruction algorithms.^6,7-11

Alternative acquisition approaches, therefore, have been introduced. Extraoral photogrammetry is used to record implant coordinates independently of conventional sequential intraoral image stitching and has demonstrated high accuracy in complete-arch applications; however, it requires dedicated equipment and complementary intraoral scanning for soft-tissue acquisition.^12-14

More recently, intraoral photogrammetry has integrated photogrammetric implant-position capture within an intraoral scanning workflow.^15-17 In parallel, conventional IOS platforms have evolved through wider fields of view, multi-angle optical acquisition, and increasingly sophisticated reconstruction algorithms.^18,19

Much of the literature that has evaluated these technologies has done so through metrological outcomes such as trueness, precision, linear and angular deviation, and 3-dimensional surface comparison.^3,4 Although these measurements are essential for characterizing digital acquisition, their clinical significance is less straightforward. The question is, “What does a measurable digital deviation actually mean for the fit of the prosthesis?”

Measurable differences among intraoral scanning systems do not necessarily translate into clinically detectable differences in prosthetic passive fit. Although inter-scanner discrepancies may be clearly identifiable through 3-dimensional analysis, their magnitude may remain within a range that does not produce a detectable effect on prosthetic adaptation. This distinction between measurable digital difference and clinical relevance represents a central consideration when interpreting the accuracy of complete-arch digital implant workflows.

Clinical verification, therefore, remains an important translational step between the digital dataset and definitive restoration. Verification bars, combined with the rocking test, one-screw (Sheffield) test, passive screw insertion, and radiographic evaluation, provide a practical means of determining whether digitally identified positional discrepancies are expressed as detectable differences in prosthetic adaptation.²⁰

Against this background, this exploratory clinical investigation examined relative implant-position discrepancies among four contemporary intraoral scanning systems and related those findings to the clinical and radiographic passive fit of corresponding computer-aided design/computer-aided manufacturing (CAD/CAM) verification bars. Although differences in performance among contemporary scanning systems are clinically relevant, the absence of an external reference standard precluded assessment of absolute scanner trueness or definitive performance ranking. The investigation, therefore, addressed a clinically focused question: “Do measurable inter-scanner discrepancies predict clinically detectable passive-fit differences in complete-arch implant rehabilitation?”

Clinical Investigation

Study Design and Patient Selection

A prospective exploratory clinical investigation was conducted to examine whether relative differences in complete-arch implant position capture among contemporary intraoral scanning systems were associated with clinically detectable differences in passive fit. Four consecutive patients requiring implant-supported complete-arch rehabilitation were included: three maxillary cases and one mandibular case. Each patient had at least four implants and was at the definitive digital impression stage of treatment.

The investigation was conducted in accordance with the Declaration of Helsinki, and all participants provided written informed consent. Radiographs were obtained as part of routine clinical verification and not solely for research purposes.

Images depicting key aspects of the study are presented in Figure 1 through Figure 16.

Digital Implant Position Capture

For each patient, definitive digital impressions were acquired during the same clinical session using four intraoral scanning systems representing different acquisition approaches: iTero Lumina™ (Align Technology, Inc., itero.com), Straumann SIRIOS™ X3 (Straumann, straumann.com), TRIOS 5 Wireless (3Shape, 3shape.com), and Aoralscan Elite (Shining 3D Dental, shining3ddental.com) (Table 1).

All scans were acquired by the same clinician, who had more than 15 years’ experience in implant prosthodontics and digital workflows. Scanner sequence was randomized for each patient to minimize potential sequence-related and operator-fatigue bias.

Scanning followed the respective manufacturers’ instructions. A standardized linear continuous (unidirectional) scanning strategy was used for conventional intraoral acquisitions, beginning at one end of the arch and proceeding continuously to the contralateral side along a vestibular–occlusal–palatal/lingual path. This provided a systematic acquisition path while limiting interruptions and realignment during complete-arch scanning.²⁰

Manufacturer-compatible non-calibrated scanbodies were used for conventional intraoral scanning, whereas the intraoral photogrammetry system used dedicated encoded scanbodies. All scanbodies were inspected before scanning to confirm complete seating and correct orientation.

For the intraoral photogrammetry workflow, the arch was initially scanned at the multi-unit abutment level without scanbodies. Dedicated encoded scanbodies were then connected and acquired in photogrammetry mode, after which the implant-position dataset was aligned with the soft-tissue scan.

Scanners were calibrated before acquisition using manufacturer-specific procedures when available. Scanbody scanning time was limited to a maximum of 35 seconds whenever possible. Environmental conditions were standardized at 21°C. All datasets were exported in STL format.

Digital Comparison and Inter-Scanner Agreement

Datasets were processed through a standardized CAD workflow and anonymized before quantitative analysis. Scanbody meshes were replaced with the corresponding library geometries to standardize implant representation across datasets.^9-11

Because no validated external in vivo reference dataset was available, the investigation was not designed to determine absolute trueness or precision. Instead, datasets obtained from the four systems within each case were compared to characterize relative inter-scanner agreement.

Three-dimensional superimposition and deviation analysis were performed using dedicated metrology software. Implant-position differences were evaluated spatially, with particular attention given to the vertical (z-axis) component because of its potential relevance to framework seating. Colorimetric maps and numerical values were used to characterize deviation magnitude and distribution.

Verification Bar Fabrication and Passive-Fit Assessment

One CAD/CAM verification bar was fabricated from each scanner-derived dataset, resulting in 16 bars across the four clinical cases. All bars were designed using identical CAD parameters and milled from 7075 aluminum using the same manufacturing protocol. All CAD and manufacturing procedures were performed by the same experienced technician, who was blinded to scanner identity.

Before clinical evaluation, the bars were anonymized and their assessment order randomized for each patient. Two blinded clinicians independently evaluated passive fit using four predefined criteria:

• Rocking test: absence of visible or tactile rocking.
• One-screw (Sheffield) test: absence of a visible gap at the contralateral terminal implant after tightening one prosthetic screw at the opposite end.
• Passive screw insertion: insertion and tightening of all prosthetic screws without forced seating.
• Radiographic verification: absence of a discernible radiolucent gap at the implant–bar interface on standardized periapical radiographs.

Disagreements between evaluators were resolved by consensus. Because of the exploratory design and limited number of cases, no inferential statistical comparisons or formal scanner rankings were performed. Findings were analyzed descriptively by relating the magnitude and spatial distribution of inter-scanner discrepancies to the corresponding clinical and radiographic passive-fit observations.

Results: From Measurable Deviation to Clinical Fit

Inter-Scanner Digital Discrepancies

Four complete-arch rehabilitations—three maxillary and one mandibular—were evaluated, generating four scanner-derived datasets and four corresponding verification bars per case.

Three-dimensional analysis demonstrated measurable relative inter-scanner discrepancies in all four cases. Maximum vertical deviations ranged from approximately 20 μm to 80 μm, with a consistent spatial pattern: central implant positions generally demonstrated greater agreement, whereas divergence increased toward the terminal regions. The greatest discrepancies were repeatedly identified at the posterior terminal implant located contralateral to the starting point of the scan, corresponding to the end of the scanning path, with the predominant component occurring along the vertical (z) axis (Table 2).

These measurements represent relative inter-scanner discrepancies rather than absolute trueness values because no external in vivo reference standard was used.

Clinical and Radiographic Passive Fit

Clinical assessment of the 16 verification bars demonstrated variable expression of the measured digital discrepancies (Table 3). Bars derived from the iTero Lumina datasets demonstrated complete seating and passive adaptation in all four cases, with no detectable rocking, negative one-screw (Sheffield) tests, passive insertion of all prosthetic screws, and no discernible radiographic gaps.

Bars generated from the other scanner-derived datasets demonstrated variable findings, including slight rocking, a positive Sheffield test at a terminal implant, and localized radiographic discrepancies. When detectable clinical discrepancies occurred, they were predominantly located in posterior regions corresponding to areas of greater digital divergence.

Relationship Between Digital Discrepancy and Passive Fit

Measurable digital discrepancy did not translate proportionally into clinically detectable misfit. Although inter-scanner differences were identified in every case, not every measurable discrepancy produced a detectable passive-fit difference. Discrepancies concentrated at terminal implant positions appeared more likely to be associated with changes in verification-bar behavior.

Across the four cases, datasets with the lowest observed maximum vertical deviations were also associated with the most consistent passive-fit outcomes. However, deviation magnitude alone did not function as a simple predictor of clinical behavior. The direction and spatial distribution of the discrepancy, particularly when vertically concentrated at a terminal implant, appeared relevant to its clinical expression.

Discussion

From Digital Discrepancy to Clinical Relevance

Complete-arch implant scanning remains a demanding application of digital impression technology, particularly across long acquisition spans with limited anatomical landmarks. Although digital performance can be characterized through multiple metrological outcomes, the clinical significance of these measurements is not always straightforward.^3,4,21,22

The present investigation addressed this distinction by examining whether relative inter-scanner discrepancies were reflected in the passive fit of corresponding verification bars, rather than attempting to determine absolute scanner trueness. Measurable discrepancies were identified in all four cases, but they did not invariably result in detectable loss of passive fit. The relationship between digital discrepancy and clinical behavior, therefore, appeared more complex than a direct numerical correlation.

Magnitude, Direction, and Location

A consistent pattern was the concentration of greater discrepancies toward posterior terminal implant positions, whereas central regions generally demonstrated closer inter-scanner agreement. This finding is consistent with the potential for cumulative registration discrepancies during long-span sequential intraoral acquisition, particularly in edentulous regions with limited anatomical landmarks.^22,23

The predominant differences occurred along the vertical (z) axis, which may directly influence seating of a rigid framework across multiple implants. The findings suggest that where and in which direction a discrepancy occurs may be as relevant as its magnitude.

Rather than supporting a single numerical threshold separating clinically acceptable from unacceptable scans, the observations suggest that magnitude, direction, spatial distribution, and implant location should be considered together when interpreting digital discrepancies.

Acquisition Technology and Interpretation

Contemporary complete-arch acquisition systems use different strategies, including sequential optical scanning, expanded or multi-angle optical capture, and photogrammetric workflows, to limit positional error.^24,25 In this investigation, verification bars generated from iTero Lumina datasets demonstrated the most consistent clinical and radiographic passive adaptation across the four cases, and the corresponding datasets also demonstrated the lowest maximum vertical discrepancies.^18,19 Aoralscan Elite demonstrated the next-lowest observed vertical deviations, followed by TRIOS 5 Wireless and Straumann SIRIOS X3.

These observations should not be interpreted as evidence of absolute scanner accuracy or technological superiority. The limited sample and absence of an external reference standard preclude formal scanner ranking. Furthermore, complete-arch acquisition is influenced by multiple variables, including scanning strategy, operator experience, scanbody design and seating, implant distribution and angulation, acquisition conditions, library congruence, and reconstruction algorithms.^26,27

The individual contribution of these factors cannot be isolated within the present clinical design. The findings therefore demonstrate that different acquisition systems may generate measurable relative differences in implant-position capture and that some of these differences may become clinically detectable when translated into a rigid prosthetic structure; they do not establish that one technology is universally superior.

Verification and Clinical Implications

Translation of each digital dataset into a physical CAD/CAM verification bar provided a clinically relevant link between digital measurement and prosthetic behavior. Complementary clinical and radiographic assessments, together with blinding of the technician and clinical evaluators, reduced the potential influence of expectation-related bias.

Digital deviation values should thus be interpreted in context, with particular attention given to discrepancies vertically concentrated at terminal implant positions. Careful control of scanning strategy, scanbody seating, calibration, acquisition conditions, and operator-related variables remains important for minimizing variability.^8,9,20,27

Clinical verification should not be considered redundant simply because a digital workflow appears numerically precise. The objective of implant-position capture is ultimately not the production of an accurate-looking dataset, but the fabrication of a restoration that seats predictably and passively. The verification step, therefore, remains an important bridge between digital measurement and clinical reality.

Future investigations should determine which combinations of deviation magnitude, direction, and location are most strongly associated with clinically meaningful changes in passive fit and, ultimately, long-term prosthetic outcomes.

Limitations

The findings of the present investigation should be interpreted within the limitations of its exploratory design. First, the number of included full-arch cases was limited. Although all patients underwent identical clinical and digital protocols and generated a large number of implant-level comparisons, the relatively small sample size does not allow broad generalization of the findings. Future studies involving larger patient cohorts are needed to confirm the observed trends and improve the external validity of the results.

Second, no validated in vivo reference standard was available. Consequently, the present study was not designed to evaluate the absolute trueness of the investigated systems. Instead, the analysis focused on relative inter-device agreement and the potential clinical implications of the observed deviations. Future investigations incorporating extraoral photogrammetry systems, coordinate-measuring technologies, or other validated reference methods may help further clarify the relationship between relative agreement and true positional accuracy.

Another limitation is that only one verification bar per scanner and per patient was fabricated. Although this approach allowed direct translation of each digital dataset into a clinical outcome, repeated manufacturing and repeated measurements could provide additional information regarding reproducibility and manufacturing variability.

Furthermore, the clinical assessment of passive fit was intentionally designed to reproduce real-world clinical decision-making through the use of the rocking test, Sheffield test, screw seating evaluation, and radiographic analysis. While these methods remain widely accepted in implant prosthodontics, they inevitably include a degree of clinical interpretation that cannot be entirely eliminated. Future studies may benefit from combining clinical verification with quantitative fit assessment methods and inter-examiner agreement analysis.

Nevertheless, the authors believe that the main strength of the present study lies precisely in its clinical nature. While most available investigations focus primarily on metrological measurements obtained in vitro, the current work attempted to evaluate how digital discrepancies may translate into clinically observable outcomes during full-arch implant rehabilitation.

Future research should, therefore, focus not only on measuring digital deviations, but also on understanding which deviations are truly clinically relevant. Establishing a stronger correlation between digital agreement metrics, passive fit assessment, prosthetic complications, and long-term clinical outcomes may ultimately provide more meaningful information for clinicians than isolated numerical accuracy values alone.

Ultimately, further prospective clinical studies with larger cohorts, validated reference standards, and long-term prosthetic follow-up are required to better define the clinical significance of digital implant impression discrepancies in full-arch implant dentistry.

Conclusions

Within the limitations of this exploratory clinical investigation, the following conclusions can be drawn:

• Relative differences in implant position capture were observed among intraoral scanners based on different acquisition technologies when used for full-arch implant-supported rehabilitations.
• The observed discrepancies were predominantly localized at terminal implant sites and were mainly expressed along the vertical axis.
• Measurable digital deviations did not invariably translate into clinically detectable misfit, suggesting that the magnitude of a digital discrepancy alone may not fully predict its clinical relevance.
• Verification bars fabricated from different scan datasets demonstrated varying levels of clinical and radiographic adaptation, indicating that relative differences in implant position capture may influence passive fit outcomes.
• Under the conditions of the present study, verification bars derived from the iTero Lumina datasets demonstrated the most consistent clinical and radiographic adaptation across the evaluated cases.
• The findings support the concept that clinical verification remains an essential step in full-arch implant rehabilitation, regardless of the digital acquisition technology employed.

Ultimately, the present study suggests that the relationship between digital deviations and clinical fit is more complex than a simple numerical accuracy value. Future research should focus not only on measuring digital discrepancies, but also on determining which deviations are truly clinically relevant when translated into definitive prosthetic outcomes. In other words, digital accuracy should not be evaluated exclusively on the basis of numerical deviation values, but also in relation to the predictability of the resulting clinical outcomes.

ACKNOWLEDGMENT

The authors thank Fabio Ferri of New Ancorvis for his valuable technical assistance in performing the digital measurements and deviation analyses used in this investigation.

DISCLOSURE

The authors had no disclosures to report.

ABOUT THE AUTHORS

Alessandro Agnini, DDS 

Private Practice, Modena and Sassuolo, Italy, specializing in fixed prosthodontics, periodontology, and implant dentistry; Founding Member, Italian Society of Prosthodontics and Oral Rehabilitation

Andrea Agnini, DDS

Private Practice, Modena and Sassuolo, Italy, specializing in fixed prosthodontics, periodontology, and implant dentistry

Tommaso Weinstein, DDS, PhD

Private Practice, Milan, Italy, specializing in restorative and digital dentistry

Tommaso Matteucci, DDS 

Private Practice, Modena and Sassuolo, Italy, specializing in fixed prosthodontics, periodontology, and implant dentistry

Giulia Benedetti, DDS 

Private Practice, Modena and Sassuolo, Italy, specializing in digital and conventional orthodontics, patient digitization, and digital smile design

Mario Imburgia, DDS, PhD 

Private Practice, Palermo, Italy, specializing in prosthodontics, implant dentistry, and digital dentistry