Using the depth deviation based on three‐dimensional images to evaluate the correction of nasolabial folds: A prospective and quantitative analysis

The deepening of the nasolabial fold (NLF) is a common occurrence during the aging process, necessitating the development of an objective method to evaluate changes in NLFs. The objective of our study was to introduce a method of depth deviation based on three‐dimensional (3D) image for evaluating the efficacy of hyaluronic acid (HA) injection in correcting NLFs.

techniques are employed to modify and restore facial shape, as well as to address the issue of facial wrinkles. 5 the nonsurgical correction of the NLF, various tissue fillers, including hyaluronic acid (HA), collagen, polycaprolactone, poly-L-lactic acid, and autologous fat, have been reported as treatment options. 4,6However, evaluating the effectiveness of cosmetic interventions for the NLF remains a challenge.Existing evaluation tools, such as the wrinkle severity rating scale (WSRS) and the global aesthetic improvement scale (GAIS), are subjective and rely on the assessments of physicians and patients. 7,8Furthermore, these traditional scales can be influenced by factors such as lighting conditions and positioning, potentially affecting the accuracy of the results.
Consequently, there is a need for a quantitative and objective tool to assess the NLF in a more reliable manner.
8][19][20][21] However, quantifying the changes in the NLF can be challenging due to its nature as a wrinkle without well-defined boundaries.The assessment process is complex, and the results may vary depending on the individual profiles of the NLF.
Therefore, this study aims to introduce a method for objectively evaluating the change of NLFs based on depth deviation measured by a 3D imaging system.This study was conducted in accordance with the guidelines of the 1964 Helsinki Declaration, and the 3D images of the nasolabial area were solely used for the purposes of this study.All participants provided written informed consent prior to their inclusion in the study.

| Injection
All subjects were injected with HA filler Restylane (Q-Med) by the treating investigator.Prior to administration, the injection areas were prepared using an antiseptic solution.Restylane was then injected into the dermis of the NLF area using a 27-gauge sharp needle.The needle was inserted into the target site and subsequently retracted while injecting the filler.To prevent HA leakage, the injection process was immediately halted upon needle withdrawal.The recommended amount of Restylane filler for one side of the NLF was advised to be less than 1.5 mL, with the actual volume determined by the treating investigator.Following the injection, a gentle massage was performed at the injection site to ensure uniform distribution of the filler.Subjects were instructed to apply ice to the proposed injection areas before and after treatment to alleviate any pain.In cases where no improvement was observed 1 month after the initial injection, a touch-up injection was administered.The data on adverse events was collected during each visit throughout the injection and follow-up period.

| Three-dimensional image scanning and analysis
The measurement of the maximum value of depth deviation (MVD) was performed by the 3D technique.In this study, the 3D images of subjects were taken by a handheld 3D scanner (EinScan-Pro 2X, SHINING 3D).Based on a structured light scanning system, the device could capture 3D images with an accuracy of less than 0.1 mm.
On the day of treatment, the preoperative 3D image was taken as the baseline image, while the postoperative 3D images were taken at 1-, 3-, 6-, and 12-month after the injection.The 3D scanning was performed in a room with the same lighting, and the subject was required to maintain a neutral expression.
The analysis of 3D images was conducted using the Geomagic Control software (3DSystems).Prior to performing the calculations, a meticulous optimal alignment process was carried out, as it was crucial for the subsequent analysis.Fixed facial structures were identified as landmarks, including the nasal tip, nasal sill, lateral canthi, medial canthi, oral commissures, and any nevi outside the treated area.The postoperative images were precisely superimposed onto the baseline image based on these landmarks.
Subsequently, an area containing the NLF was selected for analysis.Since alignment had already been performed, the selection process for the preoperative and postoperative images was carried out simultaneously.These areas were then exported for the MVD.
The MVD represents the maximum change in the depth of the NLF, and it visualizes the differences in the area through color representation.Each node was assigned a specific color based on the magnitude of the change.Nodes that remained stationary were colored green, while nodes that moved the furthest forward were colored red, and those that moved backward were colored blue.
Other colors were linearly scaled between these two extremes, indicating the distance between the two scans and illustrating the degree of change.

| Statistical analysis
In this descriptive study, quantitative variables were presented as the number of subjects and the mean ± standard deviation (SD),

| Subject population
A total of 56 subjects participated in the study, consisting of 55 females (98.2%) and 1 male (1.8%).However, two subjects dropped out during the follow-up period, leaving a final sample size of 54 females.A total of 108 NLFs were analyzed in the study.The mean age of the subjects was 41.44 years, with a range of 20 to 60 years.
At the baseline assessment, 38 subjects (70.4%) had a moderate degree of NLF (WSRS = 3), while 16 subjects (29.6%) had a severe degree of NLF (WSRS = 4).The average volume of HA gel injected for the left side of the NLF was 1.04 ± 0.10 mL, while for the right side, it was 0.99 ± 0.12 mL.The difference in volumes between the two sides was not statistically significant (p > 0.05).All subjects achieved optimal results during the first follow-up visit, and no touch-up injections were required after the initial treatment.Additionally, no serious adverse events were observed during the study.

| Qualitative measurement
The average WSRS score at baseline was 3.30 ± 0.46.After the injection of HA, the average WSRS score significantly decreased to 1.22 ± 0.42 at the 1-month follow-up (p < 0.0001).This meant that the treatment was effective in reducing the severity of NLFs.
Additionally, the 6-month GAIS score was significantly higher than the score at the 12-month visit (p < 0.0001).
In terms of the satisfaction questionnaire, at the 1-month visit, 15 subjects (27.8%) reported being very satisfied, 36 subjects (66.7%) were satisfied, and 3 subjects (5.6%) were dissatisfied (Table 1).The degree of satisfaction at the 3-month visit remained the same as the The WSRS score of subjects pre and posttreatment (***p < 0.0001).

| Quantitative measurement
Compared to the baseline, the MVD of the right NLF was 2.09 ± 0.53 mm at the 1-month follow-up.Subsequently, at the 3-, 6-, and 12-month follow-ups, the values decreased to 1.66 ± 0.54 mm, 1.39 ± 0.51 mm, and 0.87 ± 0.32 mm, respectively (Figure 3).On the left side, the MVDs were 2.11 ± 0.66 mm, 1.68 ± 0.63 mm, 1.3 ± 0.67 mm, and 0.89 ± 0.39 mm at 1-, 3-, 6-, and 12-month, respectively.The average MVD values of both sides were 2.10 ± 0.56 mm, 1.67 ± 0.57 mm, 1.34 ± 0.53 mm, and 0.88 ± 0.34 mm at 1-, 3-, 6-, and 12-month, respectively.There were no significant differences observed between the two sides (p > 0.05).Although the MVD reduced over time, the effect of HA filler was still maintained throughout the 12-month follow-up period.The 3D images and the deviation analysis of one representative subject are shown in Figure 4, illustrating the changes in the NLF after the HA injection.

| Relationship between the MVD and other clinical parameters
The analysis revealed significant correlations between the average MVD values of the two sides and the GAIS scores during the followup period.Specifically, the average MVDs showed significant correlations with the GAIS scores at the 1-month (r = 0.413, p = 0.002),

TA B L E 1
The satisfaction rate of subjects (n = 54).

F I G U R E 3
The change of MVD after treatment.

| DISCUSS ION
With the increasing interest in improving appearance, the development of new techniques for measuring the effect of facial rejuvenation is required.A variety of methods have been reported in recent years.For example, the WSRS was proposed to evaluate the severity of wrinkles, and the GAIS was used to evaluate the improvement of appearance after cosmetic procedures. 7,8The scores are widely used in clinical practice for the feature of convenience.However, as the scores are obtained from persons, the results are subjective and can be affected by the environment.
Applying 3D techniques in cosmetic procedures is an exciting development.The 3D images can be used to assess the effectiveness of treatment.By comparing pre and posttreatment scans, it is possible to measure the exact degree of improvement in the area.This can be used to determine if additional treatments are necessary or if the results are satisfactory.Meier et al. found that compared to 2D photography, the 3D imaging system showed a more reliable and precise small-volume filler enhancement in the midface region, which could improve patients' overall satisfaction level. 22And Donath et al.
reported that 3D technology was a powerful and important tool for evaluating the outcome of tear trough augmentation with HA. 11 3D images of the lower periorbital area were also used to evaluate the outcome of lower blepharoplasty with fat. 12 Additionally, for structured filler rhinoplasty, the 3D technology also showed a good ability in demonstrating early changes in the nose following treatment. 15cently, several studies have been reported to demonstrate the application of 3D technology in NLFs.Turlier et al. assessed   the efficacy of correcting the NLF by measuring the volume of NLF with the 3D imaging method. 17The volume from the bottom of wrinkles with a 20 × 30 mm area was analyzed, which decreased significantly after the injection of HA.In another study, Lowe et al. used 3D digital surface imaging to calculate the mean change in nasolabial volume. 18Although the volumetric change has been used as a method for evaluating the improvement of NLFs in previous studies, we find that it is difficult to get an accurate value of volumetric change for the NLF.Because the NLF is a wrinkle that goes from the nose to the corner of the mouth without a definite boundary, the area selected for calculating the volume remains controversial.Moreover, as the diffusion area of the filler is variable, the result of the volumetric change is not credible in a constant area.Therefore, this study introduced the MVD as a tool to assess the efficacy of the treatment of NLFs.
A deviation analysis is usually used in metrology by comparing two 3D models with each other, which is able to document the differences with a false color comparison.The analysis is mainly used to document reverse engineering deviations from the original 3D scan for quality assurance purpose.In this study, the MVD was calculated directly by comparing two scans before and after treatment, which was easier to get than the volumetric change.
Besides, the output false color mapping of depth deviation based on 3D images is visible, and that is more intuitional compared with the volumetric change.
The MVD is an index based on the change of depth, and using the MVD to assess the efficacy of NLF correction is creditable and consistent with previous studies.Fang et al. reported that there was a significant linear relationship between the maximum depth of the NLF and age-related facial change. 19Additionally, the depth of NLF was considered as the best single indicator for the assessment of NLF correction by Wang et al., because they found that the most significant change after NLF correction was the depth. 20 this study, both the GAIS score and MVD were used to illustrate the change in NLFs after HA injection.The results showed that the MVD was positively correlated with the GAIS score, which verified that the MVD could be used as a tool for evaluating the efficacy of correcting NLFs in clinical practice.With the absorption of HA over time, the improvement of treatment gradually decreased.As a quantitative parameter, the MVD could give us detailed and objective results.On the other hand, we found there was no significant correlation between the MVD and the degree of satisfaction at any time point.It was possible that the sample in this study was small and the majority of subjects felt satisfied with the treatment during the follow-up.Additionally, the MVD did not exhibit a correlation with the amount of HA injected, it might be the reason that the filler diffused unevenly after treatment.

| Limitations of the study
The small sample is one limitation of this study.Because only 54 subjects (108 cases of NLFs) were enrolled, it was difficult to represent small changes at a statistically significant level.Although objective assessment based on 3D images is a good choice for evaluating the efficacy of cosmetic procedures, more studies with larger samples are needed for confirmation.Additionally, 12 months of follow-up after injection is limited.The HA still exists in the area of NLFs to some extent at the 12-month visit, long-term follow-up studies are anticipated to identify the late effects and the duration of filler following HA injection.Besides, the cost of a 3D imaging system is high, and how to apply 3D technology in clinical practice with an acceptable price should be considered.

| CON CLUS IONS
In summary, this study demonstrates the changing trend of moderate to severe NLFs after HA injection for up to 12 months, which may improve the understanding of the absorption of HA in facial rejuvenation.Moreover, we focus on using 3D technology in evaluating the efficacy of the correction of NLFs, and the quantitative data MVD is proved to be an objective method for the assessment.

2. 1 |
SubjectsBetween May 2019 and November 2019, a total of 56 healthy volunteers were recruited for this study.Inclusion criteria were as follows: (a) Chinese males or females aged 18-60 years old; (b) presence of NLFs with a score of 3-4 according to the WSRS scale and a desire for treatment; (c) approximately symmetrical NLFs; (d) no prior aesthetic treatments related to the study; (e) good compliance with the study protocol; (f) voluntary participation and signing of informed consent.Exclusion criteria were as follows: (a) pregnancy or lactation; (b) childbearing-age women unwilling to adopt contraceptive measures during the study; (c) abnormal laboratory results; (d) presence of wounds, infections, or any skin diseases in the NLF area that could affect the study outcomes; (e) history of autoimmune disease or ongoing immunosuppressive treatment; (f) previous surgical or nonsurgical cosmetic procedures in the NLF area, such as laser treatment, botulinum toxin injections, skin fillings, chemical peels, or other treatments.
while qualitative variables were expressed as the number of subjects and percentage.To assess the significance of changes at each visit compared to the baseline, paired t-tests and Wilcoxon's signedrank tests were conducted.The difference in MVD values between two groups was analyzed using an independent t-test.Additionally, at each follow-up visit, Pearson correlation tests and Spearman's rank correlation tests were employed to examine the relationship between multiple MVD values and other clinical parameters.A pvalue less than 0.05 was considered statistically significant.
3-month (r = 0.456, p = 0.001), 6-month (r = 0.383, p = 0.04), and 12-month (r = 0.440, p = 0.01) follow-up visits.This suggests that the extent of deviation in the NLF, as measured by MVD, is associated with the perceived improvement in aesthetics according to the GAIS scores.Additionally, the average MVD values showed significant correlations with the WSRS scores at the 1-month (r = −0.862,p < 0.0001), 3-month (r = −0.905,p < 0.0001), 6-month (r = −0.975,p < 0.0001), and 12-month (r = −0.767,p < 0.0001) visits.This indicates that greater reduction in the MVD is associated with lower WSRS scores, reflecting a lower severity of NLFs.However, no significant correlations were found between the MVD and the volume of HA injected or the degree of satisfaction at any time point (p > 0.05).

F I G U R E 4
False color mapping of nasolabial fold wrinkle measured with depth deviation based on 3D images.