Three-Dimensional Change of Lip after Two-Jaw Surgery in Facial Asymmetry Using Facial Scanner

Round 1

Reviewer 1 Report

Thank you for your manuscript.

A thorough work was performed and the issue is well presented, although I have a few thoughts about the design and validity of the findings.

As for sample calculation - didn't understand how your calculation matches what you did. What are the comparison groups? If you need 15, doesn't it mean that any of your groups in inadequate in size?

Also, a month for soft tissue evaluation is a very short time in orthognathic surgery recovery and it can't really be used s a measurement of soft tissue movement.

Finally, is there a clinical value for your findings? There is no choice but o a surgery in asymmetry cases. Soft tissue alignment is side effect, albeit desired one. So how could a clinician change his practice to match your findings?

Author Response

<Reviewer 1>

1. Reviewer: A thorough work was performed and the issue is well presented, although I have a few thoughts about the design and validity of the findings.
   As for sample calculation - didn't understand how your calculation matches what you did. What are the comparison groups? If you need 15, doesn't it mean that any of your groups in inadequate in size?

Response: This study did not involve comparison of two groups. We retrospectively analyzed the changes before and after surgery for 22 patients. We used G*Power 3 to calculate the adequate sample size for paired t-test.

We have added the following text to the Materials and Methods part (lines 172-175, 177-180):

• …, paired t-test was used to evaluate the changes in landmarks before and after surgery for each patient. Pearson’s correlation analysis was also performed to determine the factors related to the vertical change of the cheilions.
• Based on a preliminary study [21], we obtained a minimum sample size of 15 at a significance level of p < 0.05, power of 80%, and effect size of 0.8 to detect differences in hard and soft tissue landmark changes before and after surgery using paired t-test.

2. Reviewer: Also, a month for soft tissue evaluation is a very short time in orthognathic surgery recovery and it can't really be used s a measurement of soft tissue movement.

Response: In this study, we intended to evaluate the change in the hard tissue landmarks by orthognathic surgery rather than the change caused by orthodontic treatment. For this reason, the amount of change in hard tissue was evaluated using CT images at least one month (mean 2.0 ± 0.9 months) after surgery, and the amount of change in soft tissue was evaluated using FS images approximately three months (mean 3.1 ± 0.9 months) after surgery.

To use the same 3D coordinate system for the CT image, “only” the pre-operative (T1) FS image was registered on the CT image. Then, for soft tissue superimposition, the T1 and T2 FS images of each patient were registered with five reference points and the wide surface of the forehead and nasal bridge. However, there are limitations associated with superimposition. There could be a registration error between T1 and T2 FS images, as well as between FS and CT images.

 

3. Reviewer: Finally, is there a clinical value for your findings? There is no choice but o a surgery in asymmetry cases. Soft tissue alignment is side effect, albeit desired one. So how could a clinician change his practice to match your findings?

Response: In this study, correlation analysis was performed to determine the factors related to the vertical change in cheilion after surgery. As a result, a significant correlation was observed between the mandibular differential setback and the downward movement of the cheilions; and there was a significant correlation between the impaction of the non-deviated side of the maxilla and the upward movement of the non-deviated side of the cheilion.

Pre-operative dentoalveolar decompensation is often necessary for relatively stable post-operative occlusion and for sufficient surgical movement of jaws in treating patients with facial asymmetry. When the temporary anchorage devices (TADs) are used during the pre-operative decompensation process, sometimes canting correction of the occlusal plane could unintentionally be performed.

 The findings of the present study reveal that it would be better to plan more impaction of the non-deviated side of the maxilla to improve relatively severe lip asymmetry. To achieve this goal, oral and maxillofacial surgeons need to consider overcorrection when establishing a surgical plan, and orthodontists need to plan decompensation to reduce unintentional canting correction during the preoperative preparation process.

Author Response File: Author Response.docx

Reviewer 2 Report

The present work extends the literature results by providing a description of the labial morphology through the variation of lip planes built over selected landmarks in the coronal, sagittal, and axial views, with the aim of evaluating the 3D changes produced by two-jaw surgery. The authors properly evaluated landmarks variation.

My main concerns regard the alignment process. In section 2.3, figure 1, is defined a surface orientation with ICP algorithm based on surface portion appropriately chosen far from the intervention area. The forehead represents a good constraint for the coronal plane direction, the nose represents a good constrain for the sagittal plane direction, but I'm concerned for the axial plane where the constrain is given by the tissue around nostrils. That is a surface portion with relevant curvatures, a narrow band of tissue and limited by the nostrils conduit. That geometrical condition is somewhat more difficult to acquire with Facial Scanner. Moreover, I suppose that the selection of the areas is manual on both surfaces, so the selected surfaces can be different and can vary for different operators doing the selection and with the same operator doing the two different selections on the two surfaces to be aligned. Even a slight difference in the ICP rigid transformation calculated on the selected areas with "uncertain" borders will be magnified toward the lower border of the chin, the farther the points are form the alignment surfaces.

It would be beneficial to perform a test in which the same pair preoperative 3D models (CT and FS) are aligned many times, with different selections, made possibly by different operators and to evaluate the errors introduced by the resulting (slightly) different transformation occurring.

In row 113 voxel based registration is declared: please elaborate on the volumes used for the registration PRE and POST surgery. The volume of the jaw should have been excluded to avoid that the discrepancies purposefully introduced by the surgery affects the overall transformation of the model.

Figure 2: same as figure 1, but with an aggravating factor: the selection of the nasion area is a shallow constrain for the axial plane and this probably leads to the distance visible around the eyes and the temples.

Finally: figure 1 and figure 2 provide the following information only: which areas of one model is in front of the other. It would be beneficial to add an image with a "shell to shell" deviation (for example using cloudcompare). For figure 1 it would also be possible to provide an estimation of the error between CT surface and more accurate FS surface.

[Note: software like cloudcompare returns mean and std deviation between the entire models. Those numbers are meaningless in a comparison of pre and post surgery surfaces. The statistics of the comparison between CT and FS in T1 (same geometry acquired with two different tools) could be altered by rigid translation component in the positions where the two surfaces intersect, because you get zero instead of the translation. So the only useful result is the distance color map. This is why I like the approach based on the landmarks, even if the landmarks identification is based on the ability of the operator and has its repeatability limits,that are well addressed in literature]

Author Response

<Reviewer 3>

1. Reviewer: My main concerns regard the alignment process. In section 2.3, figure 1, is defined a surface orientation with ICP algorithm based on surface portion appropriately chosen far from the intervention area. The forehead represents a good constraint for the coronal plane direction, the nose represents a good constrain for the sagittal plane direction, but I'm concerned for the axial plane where the constrain is given by the tissue around nostrils. That is a surface portion with relevant curvatures, a narrow band of tissue and limited by the nostrils conduit. That geometrical condition is somewhat more difficult to acquire with Facial Scanner. Moreover, I suppose that the selection of the areas is manual on both surfaces, so the selected surfaces can be different and can vary for different operators doing the selection and with the same operator doing the two different selections on the two surfaces to be aligned. Even a slight difference in the ICP rigid transformation calculated on the selected areas with "uncertain" borders will be magnified toward the lower border of the chin, the farther the points are form the alignment surfaces.

It would be beneficial to perform a test in which the same pair preoperative 3D models (CT and FS) are aligned many times, with different selections, made possibly by different operators and to evaluate the errors introduced by the resulting (slightly) different transformation occurring.

Figure 2: same as figure 1, but with an aggravating factor: the selection of the nasion area is a shallow constrain for the axial plane and this probably leads to the distance visible around the eyes and the temples.

Finally: figure 1 and figure 2 provide the following information only: which areas of one model is in front of the other. It would be beneficial to add an image with a "shell to shell" deviation (for example using cloudcompare). For figure 1 it would also be possible to provide an estimation of the error between CT surface and more accurate FS surface.

Response: In Section 2.3, we described how superimposition was performed in two steps using Geomagic Design X software. First, as suggested by the reviewer, the initial register was performed using an “approach based on landmarks.” Then, after the two 3D models were registered, the area to be used as an additional reference for more accurate superimposition was selected simultaneously from two overlapped 3D models. Within the preset “maximum average deviation range (0.05 mm)” based on the selected area, fine registration was performed by the ICP algorithm.

Although this method was used to obtain more accurate superposition based on previous studies, including reference [20], the limitations mentioned by the reviewers remains in this study. To supplement the limitations described in the discussion section, the following was added as suggested by the reviewer (lines 424-430):

• When taking 3D images with an FS, errors may occur due to the scanning ability of the equipment, and there is a possibility that the scanning accuracy may be further reduced in tissues with complex geometric curvature such as around the In addition, in this study, the areas and landmarks selected as references in the process for soft tissue superposition may cause more errors because they have insufficient constraints, especially in the axial plane direction.

 

2. Reviewer: In row 113 voxel based registration is declared: please elaborate on the volumes used for the registration PRE and POST surgery. The volume of the jaw should have been excluded to avoid that the discrepancies purposefully introduced by the surgery affects the overall transformation of the model.

Response: The superimposition for hard tissue was described based on the anterior cranial base; we further discussed that the volume of the jaw affected by orthognathic surgery was excluded, per reviewer 3's suggestion. We have added the following text to Section 2.3 (lines 115-116):

• “…and voxel-based registration was performed on the anterior cranial base to exclude the volume of the jaw affected by orthognathic surgery…”

Author Response File: Author Response.docx

Reviewer 3 Report

This work presented is on the changes to the lip following bimax surgery. The paper would be of interest to clinicians particularly OMFS and orthodontists.

Few points which could be improved:

1) Line 98: Add abbreviations to the list of landmarks in text.

2) Line 130: I noted 7 hard tissue landmarks, 3 soft tissue landmarks and 3 reference planes in Table 2. Please correct the number of landmarks in the text.

3) Line 174: How did you decide on the effect size? Is there a reference to this?

4) I am confused by the arrangement of Table 1. The p-value is at the bottom line which before that were comparisons of other variables but the p-value is only referring to Menton. Could you rearrange the table?

5) Could you discuss not the limitations of mixing Class I and Class III cases in the analysis of the asymmetry? If the cases are concerning only on asymmetry of the menton, why did the patient have Le Fort 1?

Author Response

<Reviewer 2>

1. Reviewer: Line 98: Add abbreviations to the list of landmarks in text.

Response: Abbreviations were added as recommended.

 

2. Reviewer: Line 130: I noted 7 hard tissue landmarks, 3 soft tissue landmarks and 3 reference planes in Table 2. Please correct the number of landmarks in the text.

Response: The number of landmarks was corrected as “11” including only the left and right landmarks (U3-D/ND, U6-D/ND, Mf-D/ND, Me, Ch-D/ND, Ls, Li) where the amount of change was measured.

 

3. Reviewer: Line 174: How did you decide on the effect size? Is there a reference to this?

Response: The effect size was decided based on a previous study that used a similar method as in this study. We have specified this point:

• Based on a preliminary study [21],

Reference: Choi, S.-H.; Lee, H.; Hwang, J.J.; Jung, H.-D.; Hwang, C.-J.; Cha, J.-Y. Differences in soft-tissue thickness changes after bimaxillary surgery between patients with vertically high angle and normal angle, Am. J. Orthod. Dentofac. Orthop. 2021, 159, 30-40

 

4. Reviewer: I am confused by the arrangement of Table 1. The p-value is at the bottom line which before that were comparisons of other variables but the p-value is only referring to Menton. Could you rearrange the table?

Response: In Table 1, p-value values that may cause confusion were deleted and replaced with comments below the Table.

 

5. Reviewer: Could you discuss not the limitations of mixing Class I and Class III cases in the analysis of the asymmetry? If the cases are concerning only on asymmetry of the menton, why did the patient have Le Fort 1?

Response: We have added the following text to the Discussion (lines 380-384):

• This study was conducted on patients diagnosed with skeletal class I or class III facial asymmetry who had maxillary occlusal plane cant. To improve facial asymmetry, the impaction of the non-deviated side of the maxilla and the differential setback of the mandible were commonly performed; thus, the patients were analyzed together without dividing them into Class I and Class III groups.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thank you for your revision.

After your corrections and clarifications - my opinion that iy may be accepted for publication. 

Still, I'd like a better clarification of the issue of 1 month post-op soft tissue assessment - as it is in controversy to well published and established knowledge of orthognathic postoperative soft tissue evaluation.

All the best!

Author Response

Thank you for your comments. The relevant information has been added to the discussion as follows:

According to our protocol, the evaluation of post-operative hard tissue stabilization began 4 weeks after orthognathic surgery, and facial scanning was performed to evaluate soft tissue changes in consideration of individual differences in each patient. Several studies reported that facial morphology recovers to approximately 83~90% within 3 months following surgery [23,24], although many studies suggest that the soft tissues stabilize 6 months after surgery [25]. The timing of soft tissue evaluation remains controversial. In this study, we intended to evaluate the change in the hard tissue landmarks by orthognathic surgery rather than the change caused by orthodontic treatment.