Impact of scanning range and image count on the precision of digitally recorded intermaxillary relationships in interocclusal record using intraoral scanner

Abstract

Purpose: The effect of scan range and the number of scanned images on the precision of in vivo intermaxillary relationship reproduction was evaluated using digital scans acquired with an intraoral scanner.

Methods: The study involved 15 participants with normal occlusion. Two different interocclusal recording settings were employed using the intraoral scanner (TRIOS 4): 'MIN,' focusing on the minimal scan range of the first molar region, and 'MAX,' including the scan range from the right first premolar to the right second molar. These settings were combined with three different image counts, resulting in six experimental conditions. Interocclusal recordings were performed four times for each condition. Dimensional discrepancies between datasets were analyzed using three-dimensional morphometric software and compared using two-way analysis of variance.

Results: Median dimensional discrepancies (interquartile range; IQR) of 39.2 (30.7-49.4), 42.2 (32.6-49.3), 30.3 (26.8-44.1), 20.1 (16.0-34.8), 21.8 (19.0-25.1), and 26.6 (19.9-34.5) µm were found for MIN/200, MIN/400, MIN/600, MAX/200, MAX/400, and MAX/600, respectively. Significant differences in dimensional discrepancies according to scan range were found. Wilcoxon signed-rank test showed significant differences between MAX and MIN (P < 0.01).

Conclusion: Scan range may affect the precision of intermaxillary relationship reproduction. Thus, scanning of the most extensive region practically achievable is recommended.

Introduction

The trueness and precision of three-dimensional images acquired through intraoral scanners (IOSs) have been the subject of multiple investigations [1,2,3,4,5,6,7,8,9,10,11,12,13]. Generally, previous studies have found that the trueness and precision of digital scans acquired using IOS are either clinically acceptable or superior to those obtained with conventional impression methods, particularly when the scanning range is restricted [1,4,6,8,11,12]. However, when the scan incorporates the entire dental arch, the trueness and precision of the digitally acquired data appear to be less favorable than those achieved via conventional impression methods [2,3,5,7,8,9,10,12,13].

In the context of interocclusal recording, digital recordings through IOSs involve capturing the buccal surfaces of the upper and lower dental arches while they are in maximal intercuspal contact. Unlike traditional methods, this digital approach obviates the need for using occlusal recording materials and mounting stone casts onto an articulator—factors that introduce dimensional inaccuracies. Consequently, this digital method for interocclusal recording is theoretically susceptible to fewer dimensional errors. Consistent with this hypothesis, multiple studies have demonstrated the superior reliability of digital scans acquired using IOS for reproduction of intermaxillary relationships. Zimmermann et al. [14] and Iwauchi et al. [15] reported that digital interocclusal recordings offered better precision than conventionally acquired data. In addition, Edher et al. [16] evaluated the accuracy of interocclusal recordings using a master model and reported that the accuracy of interocclusal scans of complete arches was worse than that of quadrant arch scans. This may be due to the scan range dependence of digital scanning with IOS, as described above [2,3,5,7,8,9,10,12,13].

Digital interocclusal recording requires three-dimensional alignment of digital scans of the mandibular and maxillary arches, utilizing lateral buccal scan data. In this context, inaccuracies may occur during this alignment process. Although IOS software can generally achieve alignment with minimal lateral buccal scan data, such as a single set of opposing teeth, the extent of this scan data can affect the trueness and precision of the subsequent interocclusal recording. Limited scan data may compromise registration quality due to a dearth of reference points. In clinical settings, several commercially available intraoral scanner systems prompt a notification suggesting completion of the registration when a minimal lateral scan record for as small as one tooth antagonistic pair is obtained. However, completing an interocclusal recording with such a minimal scan range might also affect its trueness and precision.

Conversely, extending the scan range—albeit still within the limits of the quadrant arch—increases the scan duration and increases the risk of positional shifts in the mandible during the procedure. In the present context, determination of an optimal scanning methodology, in terms of both range and duration, remains an unresolved issue.

The aim of this study was to critically assess the effect of variations in scan range and scan duration, measured by the number of captured images, on the fidelity of digitally replicated intermaxillary relationships in a living individual. Using IOS digital scan methodology, the study sought to either confirm or refute the null hypothesis that the scan range and the number of scanned images would have no discernible effect on the precision of reproduced intermaxillary relationships.

Materials and Methods

Participants and settings

The study sample comprised 15 participants (seven males and eight females; mean age 26.8 ± 2.2 years) exhibiting complete natural dentition and classified as skeletal Angle Class 1. All individuals were in good general health, as indicated by an American Society of Anesthesiologists physical status of 1, and had not undergone any dental rehabilitation. Any individuals with periodontitis, temporomandibular joint dysfunction, orofacial pain, or any form of acute oral disease were excluded from the study.

Digital scan and interocclusal recording using IOS

The experimental design is depicted in Fig. 1. A single operator with four years of clinical experience performed the digital scans using IOS. All scanning activities were conducted in a laboratory environment with regulated conditions: room temperature 26.0ºC and humidity 35%. Participants underwent preliminary training to maintain a maximal intercuspal position using consistent light occlusal force, guided by visual electromyographic feedback from the masseter muscle (PowerLab, ADInstruments Pty Ltd., Bella Vista, Australia). Subsequently, a digital scan of the right mandibular and maxillary molar regions was captured for each participant using an IOS (TRIOS 4, 3Shape A/S, Copenhagen, Denmark). Figure 2 illustrates two distinct scanning settings: the minimal scan range focused on the right first molar region (MIN) and an extended range from the first premolar to the second molar on the right side (MAX). Scanning durations were approximately 20, 40, and 60 s, and these were the durations required to scan image counts of 200, 400, and 600 images. Figure 3 provides an overview of these six configurations.

During the scanning process, participants were directed to maintain the maximal intercuspal position. The buccal region was scanned in accordance with the manufacturer’s guidelines. Each configuration was scanned four times, employing identical maxillary and mandibular arch data for each participant. All scanning procedures were conducted in a single, uninterrupted session. A total of 24 sets of intermaxillary relationship data, amalgamated with maxillary and mandibular arch data, were acquired for each participant. These data sets were subsequently converted and exported in the standard tessellation language (STL) format.

Fig. 1 Schematic diagram illustrating the study design

Fig. 2 Designing of the scanning scope for assessment of intermaxillary relationships, focused around the right first molar.

Two distinct bite scan ranges (MIN and MAX) were employed.

Fig. 3 Classification into six groups based on scan range (MIN and MAX) and number of scanned images (200, 400, and 600 images).

A total of 24 sets of intermaxillary relationship data combining maxillary and mandibular arch data, 4 sets per group, were duplicated for each participant.

Evaluation of precision

The STL datasets corresponding to the maxillary and mandibular dentitions were imported into a specialized three-dimensional analysis software package (PolyWorks, InnovMetric Software Inc., Québec, Canada). In these datasets, the region of interest was demarcated as the right premolars and molars. Soft tissue STL data specific to these dental regions were extracted via a trimming process. As the maxillary and mandibular dentition data remained invariant across the six distinct scanning settings, the mandibular position served as the reference for evaluating the precision of the spatially located maxillary dental images obtained from interocclusal records.

Subsequently, pairs of STL datasets for each test condition were analyzed using the best-fit algorithm method, also known as the least-squares method. This algorithm calculated the nearest distance and directional orientation for each vertex of a given polygon in one dataset relative to the surface of the corresponding polygon in the paired dataset. These calculations were conducted automatically for all polygonal vertices. This method was selected because it can quantify discrepancies between repeatedly measured three-dimensional morphological data. Furthermore, it has been used as a standard method in previous studies to investigate the precision of the IOS [17,18]. The mean absolute value of these distances was employed to assess the level of discrepancy between repeated measurements within each group, thereby serving as an index of the group’s precision. Color mapping was also utilized to visually represent the degree of discrepancy at each corresponding measurement point.

To ascertain the requisite sample size, a power analysis was conducted with a pre-specified effect size of 0.3, using statistical software (G*power 3.1.9.6, Heinrich-Heine-Universität, Düsseldorf, Germany). Based on a significance level (α) of 0.05, a power of 95%, and a standard deviation of ±100 µm across all groups for each participant, the required sample size was determined to be 15 participants per group. Because normality was rejected by Shapiro-Wilk test, aligned rank transform (ART) [Jacob O et al., In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '11): 143-146, 2011] was performed first, and then two-way analysis of variance (ANOVA) was conducted to evaluate the effect of the independent variables “scan range” and “scanned image counts” on dimensional precision, as assessed by the averaged dimensional discrepancies for each group, followed by Wilcoxon signed-rank test to examine the effect of a significant variable. All statistical analyses were performed with R version 4.3.2 (The R Foundation for Statistical Computing, Vienna, Austria), and the significance level was set at 0.05.

Results

The mean discrepancy, and 95% confidence interval (CI) for each participant, calculated using the best-fit algorithm method, are listed in Table 1, expressed in micrometers. Figure 4 shows representative color maps that display discrepancies between the measurements obtained from repeated scans for each impression method, as evaluated by the best-fit algorithm. Figure 5 depicts a sample that exemplifies the distribution of dimensional discrepancies across various settings. The median dimensional discrepancy (interquartile range; IQR) for each condition was as follows: 39.2 (30.7-49.4) µm for MIN/200, 42.2 (32.6-49.3) µm for MIN/400, 30.3 (26.8-44.1) µm for MIN/600, 20.1 (16.0-34.8) µm for MAX/200, 21.8 (19.0-25.1) µm for MAX/400, and 26.6 (19.9-34.5) µm for MAX/600.

Two-way ANOVA after ART found that the effect of “scan range” (F [1, 84] = 25.61, P < 0.01) was significant, while those of “scanned image counts” (F [2, 84] = 0.10, P = 0.90) and the interaction between two main effects (F [2, 84] = 1.12, P = 0.33) were not significant. Thereafter, Wilcoxon signed-rank test was performed to evaluate the effect of “scan range”, and this showed that the median score for the MAX (23.5 [17.5-33.4] µm) condition was significantly lower than that for the MIN condition (40.2 [29.5-48.6], P < 0.01; Fig. 6).

Table 1 The mean discrepancy and 95% confidence interval (CI) for each participant, calculated using the best-fit algorithm method

MIN/200					MAX/200
participant	mean (µm)	95% CI (µm)			participant	mean (µm)	95% CI (µm)
		lower	upper				lower	upper
A	32.1	10.9	53.3		A	14.8	8.5	21.1
B	39.2	21.5	56.8		B	46.2	33.4	59.1
C	40.7	17.7	63.6		C	45.2	29.8	60.7
D	48.1	27.9	68.2		D	23.5	17.0	29.9
E	37.1	21.8	52.5		E	13.0	7.8	18.2
F	38.5	22.4	54.7		F	16.6	9.4	23.8
G	50.1	34.2	65.9		G	23.9	17.8	30.1
H	30.8	17.5	44.0		H	17.4	10.0	24.9
I	39.5	21.0	58.0		I	20.1	14.5	25.7
J	13.7	7.8	19.5		J	11.3	7.7	15.0
K	25.0	15.9	34.2		K	19.8	15.6	23.9
L	49.4	27.6	71.3		L	32.3	16.3	48.3
M	86.6	52.5	120.7		M	42.7	23.2	62.2
N	54.8	21.0	88.7		N	34.9	8.3	61.4
O	26.8	18.3	35.3		O	16.1	8.2	23.9
MIN/400					MAX/400
participant	mean (µm)	95% CI (µm)			participant	mean (µm)	95% CI (µm)
		lower	upper				lower	upper
A	35.6	12.1	59.1		A	19.9	13.7	26.2
B	56.7	35.3	78.1		B	24.3	18.9	29.7
C	42.5	7.7	77.3		C	23.9	18.6	29.2
D	41.4	24.0	58.8		D	19.8	13.0	26.6
E	47.7	22.2	73.3		E	11.7	9.1	14.3
F	24.9	17.0	32.9		F	16.9	10.3	23.5
G	49.2	17.2	81.3		G	34.7	14.2	55.3
H	43.3	18.3	68.3		H	21.8	12.0	31.6
I	32.6	20.4	44.7		I	19.1	11.4	26.8
J	20.1	10.2	30.0		J	17.5	11.6	23.5
K	42.2	28.8	55.5		K	25.1	17.3	32.9
L	55.9	32.8	78.9		L	24.2	8.8	39.5
M	52.8	19.0	86.5		M	34.5	21.2	47.8
N	40.2	18.9	61.5		N	36.3	23.2	49.4
O	15.0	11.0	18.9		O	19.3	12.0	26.5
MIN/600					MAX/600
participant	mean (µm)	95% CI (µm)			participant	mean (µm)	95% CI (µm)
		lower	upper				lower	upper
A	19.9	9.6	30.3		A	16.4	7.4	25.4
B	44.1	30.0	58.2		B	26.7	19.2	34.2
C	40.6	17.9	63.4		C	30.2	12.0	48.5
D	30.3	20.6	40.0		D	30.5	12.0	49.0
E	19.4	8.9	29.8		E	34.5	15.7	53.3
F	26.8	13.7	39.8		F	22.8	15.9	29.7
G	41.3	16.6	66.0		G	20.5	15.6	25.4
H	29.5	17.3	41.7		H	26.7	13.8	39.5
I	52.3	27.4	77.3		I	40.7	28.6	52.8
J	13.9	8.5	19.3		J	8.3	4.7	11.9
K	29.6	19.4	39.8		K	19.9	7.6	32.3
L	40.7	25.4	56.0		L	26.5	18.9	34.1
M	62.7	22.2	103.2		M	43.6	21.8	65.4
N	74.3	11.9	136.8		N	16.1	9.3	23.0
O	30.3	17.6	43.0		O	62.4	25.6	99.3

Fig. 4 Illustrative example of qualitative evaluation through color-coded morphological mapping of superimposed datasets for each group (Participant A).

Visual inspection of the color mapping data demonstrated a smaller discrepancy (green) for the MAX group than for the MIN group (yellow and blue).

Fig. 5 Representative sample detailing the distribution of dimensional discrepancies across different groups (Participant A).

Distribution for the MAX group was clustered more around zero than that for the MIN group.

Fig. 6 Box plot of dimensional discrepancies for MAX and MIN conditions (Wilcoxon signed-rank test, *P < 0.01)

Discussion

This study aimed to evaluate the effects of scan range and scanning time, quantified by the number of scanned images, on the precision of interocclusal recordings acquired via an IOS. The initial null hypothesis that neither scan range nor scanned image counts would affect the precision of intermaxillary relationship reproduction was partially refuted. Specifically, the scan range significantly affected the precision of intermaxillary relationship reproduction, whereas the scanned image counts did not. Thus, the data consistently indicated that larger scan ranges contributed to more precise reproductions of intermaxillary relationships across all participants.

IOS software typically includes a feature that notifies users when interocclusal recording is ostensibly complete. Often, this notification arises after scanning only a minimal range, such as a single pair of occluding teeth. The findings of the present study challenge this approach, suggesting that scanning should continue to encompass the entire quadrant of the dental arch for enhanced precision, even after the notification is displayed.

Regarding the metric of the scanning time, the number of scanned images was quantified in this study. While these terms are nearly synonymous, it is crucial to recognize that they may not be entirely interchangeable. The image count captured per unit time may vary due to several factors, including computational delays caused by processing limitations when constructing 3D images in real time. In this study, this variable was mitigated by employing a computer with the highest available processing power.

In the literature, Zimmermann et al. [14] reported that the precision of digital interocclusal recording had a range of 61-84 µm, depending on the IOS model. Jaschouz et al. [19] and Iwauchi et al. [15] reported a precision of 42 µm and 31 µm, respectively. While these results cannot be compared directly between the studies because of methodological differences, the precision of the digital method was consistently reported to be superior to that of the conventional method (99 µm by Zimmermann et al. [14], 135 µm by Jaschouz et al. [19], and 128 µm by Iwauchi et al. [15]), suggesting that the precision of digital interocclusal recording is clinically acceptable.

Furthermore, when compared to the results of the previous study [15], where the study protocol was identical and thus direct comparisons are possible, they are inferior to the results obtained under the MAX conditions and superior to the results obtained under the MIN conditions in this study. These findings are not unexpected because, in the previous study, the lateral interocclusal recording range was not specified, and it was suggested that the precision of the interocclusal recording would be improved by scanning the full range of the impression area.

Regarding the study’s limitations, the digital scans using IOS for both maxillary and mandibular dentitions were acquired only once, serving as the reference for evaluating intermaxillary relationships. This protocol does not fully emulate actual clinical scenarios, where both the impression-taking process and the recording of intermaxillary relationships are sequential activities. In such settings, the trueness and precision of the impressions could affect those of the intermaxillary relationship reproductions. A further limitation lies in that this study did not assess the trueness of the method for recording digital intermaxillary relationships. To accurately determine trueness in an in vivo setting, a gold standard would need to be established; however, this is feasible only in model-based experiments. Accordingly, the present study focused solely on evaluating the precision of the employed methodology.

In conclusion, this study has shown that the extent of the lateral scanning range is crucial for achieving better precision of digital interocclusal recordings. Within the above-mentioned limitations, the present results suggest that clinicians should expand the scanning range within the quadrant arch to achieve the highest level of precision of the interocclusal records.

Abbreviations

200, 400, and 600: the number of images scanned; ANOVA: analysis of variance; ART: aligned rank transform; CI: confidence interval; IOS: intraoral scanner; IQR: interquartile range; MAX: extended scan range from the first premolar to the second molar on the right side; MIN: minimal scan range focused on the right first molar region; STL: standard tessellation language.

Ethical Statements

Written informed consent was obtained from each participant, and the study’s experimental procedures received ethical approval from the Ethics Committee of Showa University (approval No. 22-315-A).

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding

This work was supported by the Japan Society for the Promotion of Science KAKENHI (Grant Numbers 18K17127 and 20K10056).

Author Contributions

YK and KB: contributed ideas. YK and YI: collected the data. YK, ST, and YI: analyzed the data. YK, ST, YI, and KB led the writing

ORCID iD

YK: [e-mail], https://orcid.org/0009-0006-9950-8200

ST: [e-mail], https://orcid.org/0000-0001-6907-2748

YI: [e-mail], https://orcid.org/0009-0006-2638-6193

KB*: [e-mail], https://orcid.org/0000-0001-8025-2168

Data Availability Statements

Raw data were generated at the Department of Prosthodontics, Showa University. Derived data supporting the findings of this study are available from the corresponding author on request.

References

• 1.  Lee SJ, Betensky RA, Gianneschi GE, Gallucci GO (2015) Accuracy of digital versus conventional implant impressions. Clin Oral Implants Res 26, 715-719. doi: 10.1111/clr.12375
• 2.  Ajioka H, Kihara H, Odaira C, Kobayashi T, Kondo H (2016) Examination of the position accuracy of implant abutments reproduced by intra-oral optical impression. PLoS One 11, e0164048. doi: 10.1371/journal.pone.0164048
• 3.  Ender A, Attin T, Mehl A (2016) In vivo precision of conventional and digital methods of obtaining complete-arch dental impressions. J Prosthet Dent 115, 313-320. doi: 10.1016/j.prosdent.2015.09.011
• 4.  Ender A, Zimmermann M, Attin T, Mehl A (2016) In vivo precision of conventional and digital methods for obtaining quadrant dental impressions. Clin Oral Investig 20, 1495-1504. doi: 10.1007/s00784-015-1641-y
• 5.  Kuhr F, Schmidt A, Rehmann P, Wöstmann B (2016) A new method for assessing the accuracy of full arch impressions in patients. J Dent 55, 68-74. doi: 10.1016/j.jdent.2016.10.002
• 6.  Kamimura E, Tanaka S, Takaba M, Tachi K, Baba K (2017) In vivo evaluation of inter-operator reproducibility of digital dental and conventional impression techniques. PLoS One 12, e0179188. doi: 10.1371/journal.pone.0179188
• 7.  Renne W, Ludlow M, Fryml J, Schurch Z, Mennito A, Kessler R et al. (2017) Evaluation of the accuracy of 7 digital scanners: an in vitro analysis based on 3-dimensional comparisons. J Prosthet Dent 118, 36-42. doi: 10.1016/j.prosdent.2016.09.024
• 8.  Ahlholm P, Sipilä K, Vallittu P, Jakonen M, Kotiranta U (2018) Digital versus conventional impressions in fixed prosthodontics: a review. J Prosthodont 27, 35-41. doi: 10.1111/jopr.12527
• 9.  Treesh JC, Liacouras PC, Taft RM, Brooks DI, Raiciulescu S, Ellert DO et al. (2018) Complete-arch accuracy of intraoral scanners. J Prosthet Dent 120, 382-388. doi: 10.1016/j.prosdent.2018.01.005
• 10.  Ender A, Zimmermann M, Mehl A (2019) Accuracy of complete- and partial-arch impressions of actual intraoral scanning systems in vitro. Int J Comput Dent 22, 11-19.
• 11.  Miyoshi K, Tanaka S, Yokoyama S, Sanda M, Baba K (2020) Effects of different types of intraoral scanners and scanning ranges on the precision of digital implant impressions in edentulous maxilla: an in vitro study. Clin Oral Implants Res 31, 74-83. doi: 10.1111/clr.13548
• 12.  Bi C, Wang X, Tian F, Qu Z, Zhao J (2022) Comparison of accuracy between digital and conventional implant impressions: two and three dimensional evaluations. J Adv Prosthodont 14, 236-249. doi: 10.4047/jap.2022.14.4.236
• 13.  Lyu M, Di P, Lin Y, Jiang X (2022) Accuracy of impressions for multiple implants: a comparative study of digital and conventional techniques. J Prosthet Dent 128, 1017-1023. doi: 10.1016/j.prosdent.2021.01.016
• 14.  Zimmermann M, Ender A, Attin T, Mehl A (2018) Accuracy of buccal scan procedures for the registration of habitual intercuspation. Oper Dent 43, 573-580. doi: 10.2341/17-272-C
• 15.  Iwauchi Y, Tanaka S, Kamimura-Sugimura E, Baba K (2022) Clinical evaluation of the precision of interocclusal registration by using digital and conventional techniques. J Prosthet Dent 128, 611-617. doi: 10.1016/j.prosdent.2021.01.021
• 16.  Edher F, Hannam AG, Tobias DL, Wyatt CCL (2018) The accuracy of virtual interocclusal registration during intraoral scanning. J Prosthet Dent 120, 904-912. doi: 10.1016/j.prosdent.2018.01.024
• 17.  Cho YK, Kwon SW (2003) Rapid geometric modeling for construction automation. KSCE J. Civ. Eng. 7, 225-231.
• 18.  Valero E, Adan A, Cerrada C (2012) Automatic construction of 3D basic-semantic models of inhabited interiors using laser scanners and RFID sensors. Sensors 12, 5705-5724. doi: 10.3390/s120505705
• 19.  Jaschouz S, Mehl A (2014) Reproducibility of habitual intercuspation in vivo. J Dent 42, 210-218. doi: 10.1016/j.jdent.2013.09.010