- Email: [email protected]
Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study
YIJOM-3543; No of Pages 7
Int. J. Oral Maxillofac. Surg. 2016; xxx: xxx–xxx http://dx.doi.org/10.1016/j.ijom.2016.10.016, available online at http://www.sciencedirect.com
Clinical Paper Orthognathic Surgery
Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study
C. M. Resnick1, R. R. Dang1, S. J. Glick2, B. L. Padwa1 1
Department of Plastic and Oral Surgery, Harvard School of Dental Medicine, Boston Children’s Hospital, Boston, MA, USA; 23D Systems, Inc., Denver, CO, USA
C. M. Resnick, R. R. Dang, S. J. Glick, B. L. Padwa: Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study. Int. J. Oral Maxillofac. Surg. 2016; xxx: xxx–xxx. # 2016 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Abstract. Three-dimensional (3D) soft tissue prediction is replacing two-dimensional analysis in planning for orthognathic surgery. The accuracy of different computational models to predict soft tissue changes in 3D, however, is unclear. A retrospective pilot study was implemented to assess the accuracy of Dolphin 3D software in making these predictions. Seven patients who had a single-segment Le Fort I osteotomy and had preoperative (T0) and >6-month postoperative (T1) cone beam computed tomography (CBCT) scans and 3D photographs were included. The actual skeletal change was determined by subtracting the T0 from the T1 CBCT. 3D photographs were overlaid onto the T0 CBCT and virtual skeletal movements equivalent to the achieved repositioning were applied using Dolphin 3D planner. A 3D soft tissue prediction (TP) was generated and differences between the TP and T1 images (error) were measured at 14 points and at the nasolabial angle. A mean linear prediction error of 2.91 2.16 mm was found. The mean error at the nasolabial angle was 8.1 5.68. In conclusion, the ability to accurately predict 3D soft tissue changes after Le Fort I osteotomy using Dolphin 3D software is limited.
Planning for orthognathic surgery has historically been facilitated by two-dimensional (2D) soft tissue prediction of simulated osseous movements.1 As the availability of cone beam computed tomography (CBCT) and three-dimensional (3D) photography has increased, many software packages have added capability for 3D prediction. The ability to accurately 0901-5027/000001+07
simulate skeletal movements in 3D is invaluable in helping surgeons plan orthognathic procedures, inform patients of the expected results of their operations, and teach trainees. Current software can reliably simulate hard tissue movements of the maxilla and mandible in 3D.2 A linear correlation between hard and soft tissue changes has
Key words: 3D prediction; orthognathic surgery; Le Fort I osteotomy; virtual treatment planning. Accepted for publication 29 October 2016
been established in 2D,3 but these relationships have not been completely determined in 3D. In addition, prior studies of 3D prediction outcomes for orthognathic surgery have not incorporated colorized 3D photographs, an essential component to creating life-like predictions for treatment planning and patient education (Fig. 1).
# 2016 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
2
Resnick et al. Materials and methods
To address the study question, a retrospective case series of patients who had a single-segment LFI osteotomy was implemented. This was designed as a pilot study with strict inclusion criteria in order to test the concept of 3D photographic prediction using Dolphin 3D software. This study was approved by the Institutional Review Board of the Center for Applied Clinical Investigation at Boston Children’s Hospital. Study sample
Fig. 1. (A) Soft tissue prediction from the CBCT volume. (B) Soft tissue prediction with 3D photographic overlay onto the CBCT volume.
Three broad categories of computational modelling methods have been applied to facial soft tissue morphing: mass spring model (MSM), finite element model (FEM), and mass tensor model (MTM).4 Each has advantages and weaknesses, and no method has been accepted as the gold standard. These approaches have been shown to produce soft tissue predictions for orthognathic surgery that are accurate to within 0.27–1.17 mm.5–8 All of these modelling methods require a large amount of graphics and computational resources and are therefore difficult to apply in real time. In order to accommodate real-time soft tissue prediction during virtual treatment planning and consultations with patients and families, Dolphin 3D Imaging (Dolphin Imaging & Management Solutions, Chatsworth, CA, USA), the market leader in orthognathic surgical planning, uses a landmark-based photographic morphing algorithm that was developed for 2D
prediction and has been extrapolated to 3D. This system requires the user to plot 79 landmarks on the CT volume (42 bony, 37 soft tissue) and generates adjustable curves connecting these points, similar to the tracing of a lateral cephalometric radiograph (Fig. 2). The purpose of this study was to assess the accuracy of soft tissue prediction for Le Fort I osteotomy (LFI) using Dolphin 3D software. The hypotheses were (1) that Dolphin 3D would produce clinically useful 3D photographic predictions for LFI osteotomies, and (2) that Dolphin 3D would more accurately predict soft tissue changes for midline structures than for lateral facial points because the algorithms for 3D soft tissue morphing are extrapolated from experience with 2D changes in the midline. The specific aim was to measure differences in the predicted 3D soft tissue image compared to the actual result at multiple midline and lateral facial points in a series of patients after LFI osteotomy.
The study population included patients who had a single-segment LFI osteotomy at Boston Children’s Hospital from March 2008 to June 2014. To be included, subjects had to have both preoperative (T0) and at least 6-month postoperative (T1) CBCTs and 3D photographs, and had to have completed orthodontic treatment by the time of the T1 records. Patients were excluded if they had (1) craniofacial anomalies including cleft lip/palate, (2) additional operations at the time of LFI such as malar implants, mandibular or chin osteotomies, (3) multi-segment LFI osteotomies, (4) orthodontic appliances in place at the time of T1 records, or (5) inadequate records. This study was limited to single-segment LFI osteotomy in order to facilitate the evaluation of soft tissue differences in one facial region without the influence from other osseous changes. All patients underwent pre- and postoperative orthodontic treatment and had fixed orthodontic appliances in place at the time of LFI osteotomy. All T1 3D photographs were taken after orthodontic appliances had been removed. Image acquisition and preparation
Fig. 2. Method to assign landmarks and adjust morphing curves (green curves) in Dolphin 3D. Blue points are user-defined landmarks and yellow points are positions for adjustment of morphing curves. (A) Lateral view, and (B) frontal view. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
CBCTs were obtained using either an iCAT scanner (3D Imaging Systems, Imaging Sciences International Inc., Hatfield, PA, USA) or a Planmeca Promax 3D Max scanner (Planmeca USA, Inc., Roselle, IL, USA), using standard CBCT exposure settings. Prior to obtaining each image, the patients were placed in natural head position and asked to close the teeth and relax the facial muscles. All CBCT volumes were captured from at least the glabella to the hyoid bone. 3D photographs were acquired using a 3D VECTRA M3 imaging system and VECTRA software version 5.5 (Canfield Scientific, Inc., Fairfield, NJ, USA). This system comprises five pod-mounted cameras that are positioned in an arc around
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I Segment T0 and T1 CBCT volumes and save as separate STL files (Mimics)
(Table 1, Fig. 4). Twelve of these points are standard anthropometric or cephalometric landmarks.10–12 Two lateral points were created for this study: (1) lateral ala (LA), defined as the intersection of lines tangent to endocanthion (en) and subalare (sbal), and (2) maxillary buttress (MB), found at the intersection of lines tangent to exocanthion (ex) and sbal (Fig. 4A). Linear errors exceeding 2 mm were considered clinically significant, as this magnitude of difference has been suggested as the threshold for a visually perceptible facial difference.13 All imaging manipulation and analysis were performed by a single investigator (S.J.G.) who is a biomedical engineer and routinely performs virtual surgical planning for orthognathic surgery.
Register T1 segments to T0 position and export as STL files (Mimics)
Align and link T0 3D photograph to T0 CBCT (Dolphin)
Create virtual LFI osteotomies on T0 segments (Dolphin)
Assign hard and soft tissue landmarks, adjust morphing curves (Dolphin)
Move T0 maxillary segment to T1 position, autorotate T0 mandibular segment to T1 position (Dolphin)
Prediction image (TP) generated by software, exported as STL file (Dolphin)
Align and register T1 and TP phographs for measurement of differences(Mimics)
Data analysis
Fig. 3. Process for image preparation. The software program used for each step is given in parentheses.
the subject. System calibration was performed before each image capture. The patient was asked to fixate on a point at eye level on the wall in their direct vision prior to the image capture in order to establish natural head position.9 Images were saved as .OBJ files. Each CBCT volume was digitally segmented into relevant anatomical units (maxilla, mandible) using Mimics software (Materialise Inc., Leuven, Belgium) (Fig. 3). Each segmented unit was then converted to an STL file. The T1 maxillary and mandibular segments were aligned to the T0 CBCT using the skull base as a reference point. The segments were saved in these registered positions as new STL files and exported. These STL files, along with photographic and CBCT data, were imported into Dolphin 3D software version 11.8 (Dolphin Imaging & Management Solutions). In Dolphin 3D, virtual bone cuts mimicking the true LFI osteotomies were created on the T0 CBCT volumes using the virtual surgical planning module. The T0 3D photographs were linked to the T0 CBCT volumes using cranial and forehead landmarks. Seventy-nine points (42 bony, 37 soft tissue) were plotted onto the T0 linked CBCT/3D photograph as prompted by the Dolphin software, and curves connecting these points were generated automatically (Fig. 2). These curves were adjusted to accurately align with the 3D photograph. The T0 maxillary segment was virtually moved to match the T1 position. The T0
3
mandibular segment was then auto-rotated as necessary to align with the T1 position. A 3D soft tissue prediction image (TP) was generated using the native Dolphin 3D algorithm. The TP image was exported to Mimics and aligned to the T1 3D photograph using the nasion and ears as reference landmarks.
Linear and angular errors were assessed using one-tailed t-tests and Wilcoxon signed rank tests, respectively. Accuracy was defined for linear measurements as error of less than 2 mm.13 Statistical analyses were performed using IBM SPSS Statistics version 21.0 (IBM Corp., Armonk, NY, USA). A P-value of <0.05 was considered statistically significant. Results Sample characteristics
Image analysis
Differences between the TP and T1 facial images (error) were measured for each subject at 14 points (six midline and eight lateral) and at the nasolabial angle
Seven subjects were included in this pilot study (Table 2). Five (71.4%) were female. The mean age for the sample was 18.1 1.0 years at the time of LFI. All subjects presented with maxillary sagittal
Table 1. Points used for analysis. Group
Point
Description
Midline
Rhinion (R) Pronasale (P) Subnasale (sn)
Junction of bony and cartilaginous nasal dorsum Most protruded point of the apex nasi Midpoint of the angle at the columella base where the lower border of the nasal septum and the surface of the upper lip meet Point of greatest concavity on the contour of the upper lip Most anterior point of the upper lip at the mucocutaneous junction Lowest point of the upper lip vermilion
Soft tissue A point (A0 ) Labiale superius (ls) Stomion superius (stms) Lateral
Subalare (sbal) Chelion (ch) Lateral ala (LA)a Maxillary buttress (MB)a
Most lateral point in the curved base line of each ala Point located at each labial commissure Intersection of lines tangent to endocanthionb and sbal Intersection of lines tangent to exocanthionb and sbal
a
Points derived for this study. Endocanthion, the point at the inner commissure of the eye fissure; exocanthion, the point at the outer commissure of the eye fissure. b
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
4
Resnick et al. Table 3. Error and accuracy measurements.
Fig. 4. Points for measurement. (A) Frontal view. The two points derived for this study are demonstrated: lateral ala point (LA), defined as the intersection of lines tangent to endocanthion (en) and subnasale (sbal), and maxillary buttress point (MB), defined as the intersection of lines tangent to exocanthion (ex) and sbal. (B) Lateral view showing all points.
hypoplasia. Additional diagnoses included vertical maxillary excess (n = 3), maxillary vertical hypoplasia (n = 1), and anterior open bite (n = 1). The T0 images were obtained at a mean of 3.4 0.7 months (range 2.1–4.5 months) prior to LFI, and the T1 data were collected at a mean of 14.1 4.3 months (range 7.7–20.2 months) postoperatively. The average operative movements of the maxilla were as follows: 5.8 3.1 mm of sagittal advancement (range 3.7–9.0 mm, measured at anterior nasal spine (ANS)); 1.1 2.8 mm of vertical shortening (range 5.3 mm of shortening to 1.7 mm of lengthening, measured at ANS); steepening of the occlusal plane by 2.4 2.18 (range 0.3–5.28 of steepening, measured as the angle formed by the intersection of a line drawn through ANS and posterior nasal spine in the midsagittal plane and the Frankfort horizontal); and 1.7 1.3 mm of rotation and/or yaw correction (range 0.2–3.6 mm, measured
at the contact point of the central incisors) (Table 2).
Point/measure
Mean error
Accuracya
R P sn A0 ls stms sbalb chb LAb MBb Nasolabial angle
0.62 0.30 1.00 0.89 2.84 2.55 2.05 2.59 1.92 1.92 1.54 1.16 4.15 2.21 3.37 1.53 4.54 1.97 3.31 1.81 8.1 5.68
100% 85.7% 57.1% 71.4% 85.7% 71.4% 14.2% 7.14% 14.2% 28.5% –
a Accuracy = percentage average absolute error <2 mm. b Average of error from left and right sides.
lateral points, however, consistently exceeded the 2 mm linear error threshold (P < 0.05). The mean error at the nasolabial angle was 8.1 5.68 (z = 0.676, P = 0.499). Discussion
Image analysis
The overall mean error across all linear measurements was 2.91 2.16 mm (Table 3, Fig. 5). The mean error for points in the midline was 1.66 1.82 mm and for lateral points it was 3.84 1.92 mm. In the midline, the closest correlation between the TP and T1 images was seen at rhinion (mean error 0.62 0.30 mm) and the poorest agreement was at subnasale (mean error 2.84 2.55 mm). Of the lateral points, the most accurate was at the buttress (mean error maxillary 3.31 1.81 mm) and the least accurate was the lateral ala (mean error 4.54 1.97 mm). Despite two (33%) midline points with mean errors exceeding 2 mm (subnasale and soft tissue A point), no midline points had statistically significant inaccuracy (P > 0.05). Six (75%)
As the technology for capturing 3D osseous and soft tissue anatomy has become widely available and the popularity of virtual 3D orthognathic treatment planning has burgeoned, surgeons have begun to use soft tissue prediction software in treatment planning and patient education. Accurate soft tissue predictions are achievable with sophisticated software packages,5–8 but the ability to generate this level of predictive accuracy ‘at the chairside’ has not been evaluated. The purpose of this pilot study was to test the ability of a popular commercially available software that can generate 3D predictions in real time (Dolphin 3D) to accurately predict the soft tissue response to LFI osteotomy. It was expected that the Dolphin 3D algorithm, which utilizes fewer morphing points and requires less
Table 2. Characteristics of the study sample. ID
1 2 3 4 5 6 7
Sex
M F M F F F F
Age (years)
19.1 18.0 18.9 18.6 18.0 16.0 18.7
Diagnosis
Maxillary Maxillary Maxillary Maxillary Maxillary Maxillary Maxillary
sagittal sagittal sagittal sagittal sagittal sagittal sagittal
hypoplasia hypoplasia, open bite hypoplasia hypoplasia, vertical maxillary excess and vertical hypoplasia hypoplasia, vertical maxillary excess hypoplasia, vertical maxillary excess
Le Fort I osteotomy osseous movements Sagittal (mm)a
Vertical (mm)b
Occlusal plane (8)c
Rotation/yaw (mm)d
4.0 5.1 9.0 7.3 8.3 3.7 8.8
0.3 1.3 1.7 2.8 0.9 4.6 5.3
0.7 4.4 0.3 0.9 2.2 5.2 4.9
2.8 1.6 0.2 3.3 0.7 3.6 1.1
M, male; F, female. a Measured at the anterior nasal spine; a positive number indicates advancement. b Measured at the anterior nasal spine; a positive number indicates lengthening and a negative number indicates shortening. c Measured as the angle formed by the intersection of a line drawn through anterior nasal spine and posterior nasal spine in the midsagittal plane and the Frankfort horizontal. d Measured at the contact point of the central incisors.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I
5
Fig. 5. Differences between the predicted (Tp) and actual (T1) postoperative 3D images: (A) lateral tracings for midline measurements; (B) topographic maps.
computational power than other models, would introduce more error than some other software packages, but would be accurate to within a 2-mm threshold. As the methodology for photographic morphing in Dolphin software was extrapolated from 2D treatment planning, it was also hypothesized that the prediction accuracy for midline facial structures, which are
similar to those used for 2D prediction, would exceed that for lateral points, which are novel to 3D. In this sample, the prediction photographs generated by Dolphin 3D had an overall linear mean error of 2.9 mm. Much of this error was introduced at the lateral facial points; six of the eight lateral points (75%) had errors of more than 2 mm. For
midline points, the mean error was 1.7 mm, which may be clinically acceptable.13 In the midline, changes of the nasal base were associated with the largest prediction errors: mean errors at subnasale and the nasolabial angle were 2.8 mm and 88, respectively, though the former did not reach statistical significance for exceeding the 2-mm accuracy threshold. While the
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
6
Resnick et al.
magnitude of clinical significance for angular measurements has not been defined in the literature, the authors submit that 88 is a clinically important error, as the difference between a nasolabial angle of 908 and nearly 1008 is apparent in routine clinical practice. These results are not surprising – the prediction of nasolabial changes after orthognathic surgery has been notoriously difficult.14 Several studies have evaluated 3D soft tissue outcomes using CBCT soft tissue densities without 3D photographic overlays.1–8,13,15–22 Only three of these have been limited to patients who had LFI osteotomies.7,19,21 Wang and Yang found a finite element model to produce soft tissue deformation with absolute error under 0.5 mm,8 while several authors have tested predictions for LFI osteotomies using the mass spring model and found errors ranging from 0.27 to 1.17 mm.6,7 Liebregts and colleagues found a mean error of only 0.81 mm for predictions of bimaxillary surgery generated using a mass tensor model.5 Wide variability in inclusion criteria, computational models for soft tissue morphing, and methods of error analysis, however, make these publications difficult to compare. This study incorporated 3D photographic overlays in the analysis of prediction accuracy. This is an important modification, as the diagnostic and educational power of a 3D photograph significantly exceeds that of a monochromatic CBCT soft tissue density (Fig. 1). The ability for planning software to accurately morph the CBCT volume may not fully determine the secondary response to the overlaid 3D photograph. There are two likely reasons for the larger errors seen in the present sample compared to previous studies. First, because error is introduced with each step in the imaging registration process,23,24 inferior accuracy can be expected as steps such as the overlaying of 3D photographs onto CBCT volumes are added to the technique. Second, the methodology used by Dolphin 3D to generate photographic predictions differs considerably from the computational models used by many other software packages.4 Dolphin software utilizes a landmark-based morphing algorithm, which has been validated using lateral cephalometric radiographs and 2D photographs.1–3 As the number of available points and the vectors of their potential movement move from finite in 2D to nearly infinite in 3D, it is unclear if this methodology can be accurately extrapolated to 3D. The only previous study that has evaluated 3D predictions using
Dolphin software found prediction errors to be clinically acceptable except for at the nasolabial and mentolabial angles, where mean error was 98.18 This study has several limitations. First, the small sample size and retrospective design limit the ability to draw strong conclusions. One reason for the small sample size was the development of strict inclusion and exclusion criteria, which resulted in the majority of the orthognathic procedures completed in the unit during the study period being eliminated. It was felt, however, that, for this pilot study, a uniform population was critical to minimizing confounding variables in the analysis. A larger study sample could allow additional important questions to be addressed, such as the effect of the magnitude and direction of osseous movement on the predictive accuracy of the software. Second, the introduction of error during the multiple overlays of CBCT volumes and 3D photographs is inherent to the process. To minimize this, all image manipulation for this study was performed by a single investigator (S.J.G.) who is a biomedical engineer and routinely performs virtual surgical planning. Third, all preoperative images from which the predictions were generated were obtained with fixed orthodontic appliances in place, but the postoperative images were taken after these appliances had been removed. This difference introduces additional error, as the orthodontic appliances may provide added support to the upper lip. While stated as a limitation, this could be viewed as a strength of the study, as the ‘real’ result of LFI osteotomies at the endtreatment stage has been evaluated. Opportunities for the introduction of minor discrepancies between predicted and real images are innumerable, and unnaturally controlling for them decreases the clinical applicability of these techniques to routine practice. In conclusion, this preliminary analysis demonstrated that the ability to accurately predict 3D soft tissue changes after Le Fort I osteotomy using Dolphin planning software is limited. Prediction accuracy may be acceptable for linear changes in the midline, but not for lateral facial points. In the midline, changes at the nasal base are most prone to prediction error. It is planned to continue to study the predictive accuracy of this software with a larger sample.
Funding
None declared.
Competing interests
One author, Sadie Glick, is an employee of 3D Systems, Inc., which uses Dolphin 3D imaging for virtual surgical planning. There were no financial incentives for participation in this study. Ethical approval
This study was approved by the Institutional Review Board of the Center for Applied Clinical Investigation at Boston Children’s Hospital (Protocol #00019505). Patient consent
Written consent was obtained for publication of the patient photographs. References 1. Mankad B, Cisneros GJ, Freeman K, Eisig SB. Prediction accuracy of soft tissue profile in orthognathic surgery. Int J Adult Orthodon Orthognath Surg 1999;14:19–26. 2. Tucker S, Cevidanes LH, Styner M, Kim H, Reyes M, Proffit W, et al. Comparison of actual surgical outcomes and 3-dimensional surgical simulations. J Oral Maxillofac Surg 2010;68:2412–21. 3. Chew MT, Sandham A, Wong HB. Evaluation of the linearity of soft- to hard-tissue movement after orthognathic surgery. Am J Orthod Dentofacial Orthop 2008;134:665–70. 4. Kim H, Jurgens P, Nolte LP, Reyes M. Anatomically-driven soft-tissue simulation strategy for cranio-maxillofacial surgery using facial muscle template model. Med Image Comput Comput Assist Interv 2010;13:61–8. 5. Liebregts J, Xi T, Timmermans M, de Koning M, Berge S, Hoppenreijs T, et al. Accuracy of three-dimensional soft tissue simulation in bimaxillary osteotomies. J Craniomaxillofac Surg 2015;43:329–35. 6. Schendel SA, Jacobson R, Khalessi S. 3dimensional facial simulation in orthognathic surgery: is it accurate? J Oral Maxillofac Surg 2013;71:1406–14. 7. Ullah R, Turner PJ, Khambay BS. Accuracy of three-dimensional soft tissue predictions in orthognathic surgery after Le Fort I advancement osteotomies. Br J Oral Maxillofac Surg 2015;53:153–7. 8. Wang S, Yang J. An improved finite element model for craniofacial surgery simulation. Int J Comput Assist Radiol Surg 2009;4:579–87. 9. Meiyappan N, Tamizharasi S, Senthilkumar KP, Janardhanan K. Natural head position: an overview. J Pharm Bioallied Sci 2015;7:S424–7. 10. Farkas LG. Anthropometry of the head and face. second ed. New York: Raven Press; 1994.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I 11. Jacobson A, Caufield PW. Introduction to radiographic cephalometry. Philadelphia: Lea & Febiger; 1985. 12. Jagadish Chandra H, Ravi MS, Sharma SM, Rajendra Prasad B. Standards of facial esthetics: an anthropometric study. J Maxillofac Oral Surg 2012;11:384–9. 13. Kaipatur NR, Flores-Mir C. Accuracy of computer programs in predicting orthognathic surgery soft tissue response. J Oral Maxillofac Surg 2009;67:751–9. 14. Yamada T, Mishima K, Moritani N, Janune D, Matsumura T, Ikeya Y, et al. Nasolabial morphologic changes after a Le Fort I osteotomy: a three-dimensional anthropometric study. J Craniofac Surg 2010;21:1089–95. 15. Jabar N, Robinson W, Goto TK, Khambay BS. The validity of using surface meshes for evaluation of three-dimensional maxillary and mandibular surgical changes. Int J Oral Maxillofac Surg 2015;44:914–20. 16. Khambay B, Ullah R. Current methods of assessing the accuracy of three-dimensional soft tissue facial predictions: technical and clinical considerations. Int J Oral Maxillofac Surg 2015;44:132–8. 17. Liebregts JH, Timmermans M, De Koning MJ, Berge SJ, Maal TJ. Three-dimensional facial simulation in bilateral sagittal split
18.
19.
20.
21.
22.
osteotomy: a validation study of 100 patients. J Oral Maxillofac Surg 2015; 73:961–70. Nadjmi N, Tehranchi A, Azami N, Saedi B, Mollemans W. Comparison of soft-tissue profiles in Le Fort I osteotomy patients with Dolphin and Maxilim softwares. Am J Orthod Dentofacial Orthop 2013;144:654– 62. Nkenke E, Vairaktaris E, Kramer M, Schlegel A, Holst A, Hirschfelder U, et al. Threedimensional analysis of changes of the malar-midfacial region after LeFort I osteotomy and maxillary advancement. Oral Maxillofac Surg 2008;12:5–12. Plooij JM, Swennen GR, Rangel FA, Maal TJ, Schutyser FA, Bronkhorst EM, et al. Evaluation of reproducibility and reliability of 3D soft tissue analysis using 3D stereophotogrammetry. Int J Oral Maxillofac Surg 2009;38:267–73. Shafi MI, Ayoub A, Ju X, Khambay B. The accuracy of three-dimensional prediction planning for the surgical correction of facial deformities using Maxilim. Int J Oral Maxillofac Surg 2013;42:801–6. Terzic A, Combescure C, Scolozzi P. Accuracy of computational soft tissue predictions in orthognathic surgery from
7
three-dimensional photographs 6 months after completion of surgery: a preliminary study of 13 patients. Aesthetic Plast Surg 2014;38:184–91. 23. Naudi KB, Benramadan R, Brocklebank L, Ju X, Khambay B, Ayoub A. The virtual human face: superimposing the simultaneously captured 3D photorealistic skin surface of the face on the untextured skin image of the CBCT scan. Int J Oral Maxillofac Surg 2013;42:393–400. 24. Xin P, Yu H, Cheng H, Shen S, Shen SG. Image fusion in craniofacial virtual reality modeling based on CT and 3dMD photogrammetry. J Craniofac Surg 2013; 24:1573–6.
Address: Cory M. Resnick Boston Children’s Hospital Department of Plastic and Oral Surgery 300 Longwood Avenue Boston MA 02115 USA Tel: +1 617 355 6082; Fax: +1 617 738 1657 E-mail: [email protected]
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
Int. J. Oral Maxillofac. Surg. 2016; xxx: xxx–xxx http://dx.doi.org/10.1016/j.ijom.2016.10.016, available online at http://www.sciencedirect.com
Clinical Paper Orthognathic Surgery
Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study
C. M. Resnick1, R. R. Dang1, S. J. Glick2, B. L. Padwa1 1
Department of Plastic and Oral Surgery, Harvard School of Dental Medicine, Boston Children’s Hospital, Boston, MA, USA; 23D Systems, Inc., Denver, CO, USA
C. M. Resnick, R. R. Dang, S. J. Glick, B. L. Padwa: Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study. Int. J. Oral Maxillofac. Surg. 2016; xxx: xxx–xxx. # 2016 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Abstract. Three-dimensional (3D) soft tissue prediction is replacing two-dimensional analysis in planning for orthognathic surgery. The accuracy of different computational models to predict soft tissue changes in 3D, however, is unclear. A retrospective pilot study was implemented to assess the accuracy of Dolphin 3D software in making these predictions. Seven patients who had a single-segment Le Fort I osteotomy and had preoperative (T0) and >6-month postoperative (T1) cone beam computed tomography (CBCT) scans and 3D photographs were included. The actual skeletal change was determined by subtracting the T0 from the T1 CBCT. 3D photographs were overlaid onto the T0 CBCT and virtual skeletal movements equivalent to the achieved repositioning were applied using Dolphin 3D planner. A 3D soft tissue prediction (TP) was generated and differences between the TP and T1 images (error) were measured at 14 points and at the nasolabial angle. A mean linear prediction error of 2.91 2.16 mm was found. The mean error at the nasolabial angle was 8.1 5.68. In conclusion, the ability to accurately predict 3D soft tissue changes after Le Fort I osteotomy using Dolphin 3D software is limited.
Planning for orthognathic surgery has historically been facilitated by two-dimensional (2D) soft tissue prediction of simulated osseous movements.1 As the availability of cone beam computed tomography (CBCT) and three-dimensional (3D) photography has increased, many software packages have added capability for 3D prediction. The ability to accurately 0901-5027/000001+07
simulate skeletal movements in 3D is invaluable in helping surgeons plan orthognathic procedures, inform patients of the expected results of their operations, and teach trainees. Current software can reliably simulate hard tissue movements of the maxilla and mandible in 3D.2 A linear correlation between hard and soft tissue changes has
Key words: 3D prediction; orthognathic surgery; Le Fort I osteotomy; virtual treatment planning. Accepted for publication 29 October 2016
been established in 2D,3 but these relationships have not been completely determined in 3D. In addition, prior studies of 3D prediction outcomes for orthognathic surgery have not incorporated colorized 3D photographs, an essential component to creating life-like predictions for treatment planning and patient education (Fig. 1).
# 2016 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
2
Resnick et al. Materials and methods
To address the study question, a retrospective case series of patients who had a single-segment LFI osteotomy was implemented. This was designed as a pilot study with strict inclusion criteria in order to test the concept of 3D photographic prediction using Dolphin 3D software. This study was approved by the Institutional Review Board of the Center for Applied Clinical Investigation at Boston Children’s Hospital. Study sample
Fig. 1. (A) Soft tissue prediction from the CBCT volume. (B) Soft tissue prediction with 3D photographic overlay onto the CBCT volume.
Three broad categories of computational modelling methods have been applied to facial soft tissue morphing: mass spring model (MSM), finite element model (FEM), and mass tensor model (MTM).4 Each has advantages and weaknesses, and no method has been accepted as the gold standard. These approaches have been shown to produce soft tissue predictions for orthognathic surgery that are accurate to within 0.27–1.17 mm.5–8 All of these modelling methods require a large amount of graphics and computational resources and are therefore difficult to apply in real time. In order to accommodate real-time soft tissue prediction during virtual treatment planning and consultations with patients and families, Dolphin 3D Imaging (Dolphin Imaging & Management Solutions, Chatsworth, CA, USA), the market leader in orthognathic surgical planning, uses a landmark-based photographic morphing algorithm that was developed for 2D
prediction and has been extrapolated to 3D. This system requires the user to plot 79 landmarks on the CT volume (42 bony, 37 soft tissue) and generates adjustable curves connecting these points, similar to the tracing of a lateral cephalometric radiograph (Fig. 2). The purpose of this study was to assess the accuracy of soft tissue prediction for Le Fort I osteotomy (LFI) using Dolphin 3D software. The hypotheses were (1) that Dolphin 3D would produce clinically useful 3D photographic predictions for LFI osteotomies, and (2) that Dolphin 3D would more accurately predict soft tissue changes for midline structures than for lateral facial points because the algorithms for 3D soft tissue morphing are extrapolated from experience with 2D changes in the midline. The specific aim was to measure differences in the predicted 3D soft tissue image compared to the actual result at multiple midline and lateral facial points in a series of patients after LFI osteotomy.
The study population included patients who had a single-segment LFI osteotomy at Boston Children’s Hospital from March 2008 to June 2014. To be included, subjects had to have both preoperative (T0) and at least 6-month postoperative (T1) CBCTs and 3D photographs, and had to have completed orthodontic treatment by the time of the T1 records. Patients were excluded if they had (1) craniofacial anomalies including cleft lip/palate, (2) additional operations at the time of LFI such as malar implants, mandibular or chin osteotomies, (3) multi-segment LFI osteotomies, (4) orthodontic appliances in place at the time of T1 records, or (5) inadequate records. This study was limited to single-segment LFI osteotomy in order to facilitate the evaluation of soft tissue differences in one facial region without the influence from other osseous changes. All patients underwent pre- and postoperative orthodontic treatment and had fixed orthodontic appliances in place at the time of LFI osteotomy. All T1 3D photographs were taken after orthodontic appliances had been removed. Image acquisition and preparation
Fig. 2. Method to assign landmarks and adjust morphing curves (green curves) in Dolphin 3D. Blue points are user-defined landmarks and yellow points are positions for adjustment of morphing curves. (A) Lateral view, and (B) frontal view. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
CBCTs were obtained using either an iCAT scanner (3D Imaging Systems, Imaging Sciences International Inc., Hatfield, PA, USA) or a Planmeca Promax 3D Max scanner (Planmeca USA, Inc., Roselle, IL, USA), using standard CBCT exposure settings. Prior to obtaining each image, the patients were placed in natural head position and asked to close the teeth and relax the facial muscles. All CBCT volumes were captured from at least the glabella to the hyoid bone. 3D photographs were acquired using a 3D VECTRA M3 imaging system and VECTRA software version 5.5 (Canfield Scientific, Inc., Fairfield, NJ, USA). This system comprises five pod-mounted cameras that are positioned in an arc around
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I Segment T0 and T1 CBCT volumes and save as separate STL files (Mimics)
(Table 1, Fig. 4). Twelve of these points are standard anthropometric or cephalometric landmarks.10–12 Two lateral points were created for this study: (1) lateral ala (LA), defined as the intersection of lines tangent to endocanthion (en) and subalare (sbal), and (2) maxillary buttress (MB), found at the intersection of lines tangent to exocanthion (ex) and sbal (Fig. 4A). Linear errors exceeding 2 mm were considered clinically significant, as this magnitude of difference has been suggested as the threshold for a visually perceptible facial difference.13 All imaging manipulation and analysis were performed by a single investigator (S.J.G.) who is a biomedical engineer and routinely performs virtual surgical planning for orthognathic surgery.
Register T1 segments to T0 position and export as STL files (Mimics)
Align and link T0 3D photograph to T0 CBCT (Dolphin)
Create virtual LFI osteotomies on T0 segments (Dolphin)
Assign hard and soft tissue landmarks, adjust morphing curves (Dolphin)
Move T0 maxillary segment to T1 position, autorotate T0 mandibular segment to T1 position (Dolphin)
Prediction image (TP) generated by software, exported as STL file (Dolphin)
Align and register T1 and TP phographs for measurement of differences(Mimics)
Data analysis
Fig. 3. Process for image preparation. The software program used for each step is given in parentheses.
the subject. System calibration was performed before each image capture. The patient was asked to fixate on a point at eye level on the wall in their direct vision prior to the image capture in order to establish natural head position.9 Images were saved as .OBJ files. Each CBCT volume was digitally segmented into relevant anatomical units (maxilla, mandible) using Mimics software (Materialise Inc., Leuven, Belgium) (Fig. 3). Each segmented unit was then converted to an STL file. The T1 maxillary and mandibular segments were aligned to the T0 CBCT using the skull base as a reference point. The segments were saved in these registered positions as new STL files and exported. These STL files, along with photographic and CBCT data, were imported into Dolphin 3D software version 11.8 (Dolphin Imaging & Management Solutions). In Dolphin 3D, virtual bone cuts mimicking the true LFI osteotomies were created on the T0 CBCT volumes using the virtual surgical planning module. The T0 3D photographs were linked to the T0 CBCT volumes using cranial and forehead landmarks. Seventy-nine points (42 bony, 37 soft tissue) were plotted onto the T0 linked CBCT/3D photograph as prompted by the Dolphin software, and curves connecting these points were generated automatically (Fig. 2). These curves were adjusted to accurately align with the 3D photograph. The T0 maxillary segment was virtually moved to match the T1 position. The T0
3
mandibular segment was then auto-rotated as necessary to align with the T1 position. A 3D soft tissue prediction image (TP) was generated using the native Dolphin 3D algorithm. The TP image was exported to Mimics and aligned to the T1 3D photograph using the nasion and ears as reference landmarks.
Linear and angular errors were assessed using one-tailed t-tests and Wilcoxon signed rank tests, respectively. Accuracy was defined for linear measurements as error of less than 2 mm.13 Statistical analyses were performed using IBM SPSS Statistics version 21.0 (IBM Corp., Armonk, NY, USA). A P-value of <0.05 was considered statistically significant. Results Sample characteristics
Image analysis
Differences between the TP and T1 facial images (error) were measured for each subject at 14 points (six midline and eight lateral) and at the nasolabial angle
Seven subjects were included in this pilot study (Table 2). Five (71.4%) were female. The mean age for the sample was 18.1 1.0 years at the time of LFI. All subjects presented with maxillary sagittal
Table 1. Points used for analysis. Group
Point
Description
Midline
Rhinion (R) Pronasale (P) Subnasale (sn)
Junction of bony and cartilaginous nasal dorsum Most protruded point of the apex nasi Midpoint of the angle at the columella base where the lower border of the nasal septum and the surface of the upper lip meet Point of greatest concavity on the contour of the upper lip Most anterior point of the upper lip at the mucocutaneous junction Lowest point of the upper lip vermilion
Soft tissue A point (A0 ) Labiale superius (ls) Stomion superius (stms) Lateral
Subalare (sbal) Chelion (ch) Lateral ala (LA)a Maxillary buttress (MB)a
Most lateral point in the curved base line of each ala Point located at each labial commissure Intersection of lines tangent to endocanthionb and sbal Intersection of lines tangent to exocanthionb and sbal
a
Points derived for this study. Endocanthion, the point at the inner commissure of the eye fissure; exocanthion, the point at the outer commissure of the eye fissure. b
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
4
Resnick et al. Table 3. Error and accuracy measurements.
Fig. 4. Points for measurement. (A) Frontal view. The two points derived for this study are demonstrated: lateral ala point (LA), defined as the intersection of lines tangent to endocanthion (en) and subnasale (sbal), and maxillary buttress point (MB), defined as the intersection of lines tangent to exocanthion (ex) and sbal. (B) Lateral view showing all points.
hypoplasia. Additional diagnoses included vertical maxillary excess (n = 3), maxillary vertical hypoplasia (n = 1), and anterior open bite (n = 1). The T0 images were obtained at a mean of 3.4 0.7 months (range 2.1–4.5 months) prior to LFI, and the T1 data were collected at a mean of 14.1 4.3 months (range 7.7–20.2 months) postoperatively. The average operative movements of the maxilla were as follows: 5.8 3.1 mm of sagittal advancement (range 3.7–9.0 mm, measured at anterior nasal spine (ANS)); 1.1 2.8 mm of vertical shortening (range 5.3 mm of shortening to 1.7 mm of lengthening, measured at ANS); steepening of the occlusal plane by 2.4 2.18 (range 0.3–5.28 of steepening, measured as the angle formed by the intersection of a line drawn through ANS and posterior nasal spine in the midsagittal plane and the Frankfort horizontal); and 1.7 1.3 mm of rotation and/or yaw correction (range 0.2–3.6 mm, measured
at the contact point of the central incisors) (Table 2).
Point/measure
Mean error
Accuracya
R P sn A0 ls stms sbalb chb LAb MBb Nasolabial angle
0.62 0.30 1.00 0.89 2.84 2.55 2.05 2.59 1.92 1.92 1.54 1.16 4.15 2.21 3.37 1.53 4.54 1.97 3.31 1.81 8.1 5.68
100% 85.7% 57.1% 71.4% 85.7% 71.4% 14.2% 7.14% 14.2% 28.5% –
a Accuracy = percentage average absolute error <2 mm. b Average of error from left and right sides.
lateral points, however, consistently exceeded the 2 mm linear error threshold (P < 0.05). The mean error at the nasolabial angle was 8.1 5.68 (z = 0.676, P = 0.499). Discussion
Image analysis
The overall mean error across all linear measurements was 2.91 2.16 mm (Table 3, Fig. 5). The mean error for points in the midline was 1.66 1.82 mm and for lateral points it was 3.84 1.92 mm. In the midline, the closest correlation between the TP and T1 images was seen at rhinion (mean error 0.62 0.30 mm) and the poorest agreement was at subnasale (mean error 2.84 2.55 mm). Of the lateral points, the most accurate was at the buttress (mean error maxillary 3.31 1.81 mm) and the least accurate was the lateral ala (mean error 4.54 1.97 mm). Despite two (33%) midline points with mean errors exceeding 2 mm (subnasale and soft tissue A point), no midline points had statistically significant inaccuracy (P > 0.05). Six (75%)
As the technology for capturing 3D osseous and soft tissue anatomy has become widely available and the popularity of virtual 3D orthognathic treatment planning has burgeoned, surgeons have begun to use soft tissue prediction software in treatment planning and patient education. Accurate soft tissue predictions are achievable with sophisticated software packages,5–8 but the ability to generate this level of predictive accuracy ‘at the chairside’ has not been evaluated. The purpose of this pilot study was to test the ability of a popular commercially available software that can generate 3D predictions in real time (Dolphin 3D) to accurately predict the soft tissue response to LFI osteotomy. It was expected that the Dolphin 3D algorithm, which utilizes fewer morphing points and requires less
Table 2. Characteristics of the study sample. ID
1 2 3 4 5 6 7
Sex
M F M F F F F
Age (years)
19.1 18.0 18.9 18.6 18.0 16.0 18.7
Diagnosis
Maxillary Maxillary Maxillary Maxillary Maxillary Maxillary Maxillary
sagittal sagittal sagittal sagittal sagittal sagittal sagittal
hypoplasia hypoplasia, open bite hypoplasia hypoplasia, vertical maxillary excess and vertical hypoplasia hypoplasia, vertical maxillary excess hypoplasia, vertical maxillary excess
Le Fort I osteotomy osseous movements Sagittal (mm)a
Vertical (mm)b
Occlusal plane (8)c
Rotation/yaw (mm)d
4.0 5.1 9.0 7.3 8.3 3.7 8.8
0.3 1.3 1.7 2.8 0.9 4.6 5.3
0.7 4.4 0.3 0.9 2.2 5.2 4.9
2.8 1.6 0.2 3.3 0.7 3.6 1.1
M, male; F, female. a Measured at the anterior nasal spine; a positive number indicates advancement. b Measured at the anterior nasal spine; a positive number indicates lengthening and a negative number indicates shortening. c Measured as the angle formed by the intersection of a line drawn through anterior nasal spine and posterior nasal spine in the midsagittal plane and the Frankfort horizontal. d Measured at the contact point of the central incisors.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I
5
Fig. 5. Differences between the predicted (Tp) and actual (T1) postoperative 3D images: (A) lateral tracings for midline measurements; (B) topographic maps.
computational power than other models, would introduce more error than some other software packages, but would be accurate to within a 2-mm threshold. As the methodology for photographic morphing in Dolphin software was extrapolated from 2D treatment planning, it was also hypothesized that the prediction accuracy for midline facial structures, which are
similar to those used for 2D prediction, would exceed that for lateral points, which are novel to 3D. In this sample, the prediction photographs generated by Dolphin 3D had an overall linear mean error of 2.9 mm. Much of this error was introduced at the lateral facial points; six of the eight lateral points (75%) had errors of more than 2 mm. For
midline points, the mean error was 1.7 mm, which may be clinically acceptable.13 In the midline, changes of the nasal base were associated with the largest prediction errors: mean errors at subnasale and the nasolabial angle were 2.8 mm and 88, respectively, though the former did not reach statistical significance for exceeding the 2-mm accuracy threshold. While the
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
6
Resnick et al.
magnitude of clinical significance for angular measurements has not been defined in the literature, the authors submit that 88 is a clinically important error, as the difference between a nasolabial angle of 908 and nearly 1008 is apparent in routine clinical practice. These results are not surprising – the prediction of nasolabial changes after orthognathic surgery has been notoriously difficult.14 Several studies have evaluated 3D soft tissue outcomes using CBCT soft tissue densities without 3D photographic overlays.1–8,13,15–22 Only three of these have been limited to patients who had LFI osteotomies.7,19,21 Wang and Yang found a finite element model to produce soft tissue deformation with absolute error under 0.5 mm,8 while several authors have tested predictions for LFI osteotomies using the mass spring model and found errors ranging from 0.27 to 1.17 mm.6,7 Liebregts and colleagues found a mean error of only 0.81 mm for predictions of bimaxillary surgery generated using a mass tensor model.5 Wide variability in inclusion criteria, computational models for soft tissue morphing, and methods of error analysis, however, make these publications difficult to compare. This study incorporated 3D photographic overlays in the analysis of prediction accuracy. This is an important modification, as the diagnostic and educational power of a 3D photograph significantly exceeds that of a monochromatic CBCT soft tissue density (Fig. 1). The ability for planning software to accurately morph the CBCT volume may not fully determine the secondary response to the overlaid 3D photograph. There are two likely reasons for the larger errors seen in the present sample compared to previous studies. First, because error is introduced with each step in the imaging registration process,23,24 inferior accuracy can be expected as steps such as the overlaying of 3D photographs onto CBCT volumes are added to the technique. Second, the methodology used by Dolphin 3D to generate photographic predictions differs considerably from the computational models used by many other software packages.4 Dolphin software utilizes a landmark-based morphing algorithm, which has been validated using lateral cephalometric radiographs and 2D photographs.1–3 As the number of available points and the vectors of their potential movement move from finite in 2D to nearly infinite in 3D, it is unclear if this methodology can be accurately extrapolated to 3D. The only previous study that has evaluated 3D predictions using
Dolphin software found prediction errors to be clinically acceptable except for at the nasolabial and mentolabial angles, where mean error was 98.18 This study has several limitations. First, the small sample size and retrospective design limit the ability to draw strong conclusions. One reason for the small sample size was the development of strict inclusion and exclusion criteria, which resulted in the majority of the orthognathic procedures completed in the unit during the study period being eliminated. It was felt, however, that, for this pilot study, a uniform population was critical to minimizing confounding variables in the analysis. A larger study sample could allow additional important questions to be addressed, such as the effect of the magnitude and direction of osseous movement on the predictive accuracy of the software. Second, the introduction of error during the multiple overlays of CBCT volumes and 3D photographs is inherent to the process. To minimize this, all image manipulation for this study was performed by a single investigator (S.J.G.) who is a biomedical engineer and routinely performs virtual surgical planning. Third, all preoperative images from which the predictions were generated were obtained with fixed orthodontic appliances in place, but the postoperative images were taken after these appliances had been removed. This difference introduces additional error, as the orthodontic appliances may provide added support to the upper lip. While stated as a limitation, this could be viewed as a strength of the study, as the ‘real’ result of LFI osteotomies at the endtreatment stage has been evaluated. Opportunities for the introduction of minor discrepancies between predicted and real images are innumerable, and unnaturally controlling for them decreases the clinical applicability of these techniques to routine practice. In conclusion, this preliminary analysis demonstrated that the ability to accurately predict 3D soft tissue changes after Le Fort I osteotomy using Dolphin planning software is limited. Prediction accuracy may be acceptable for linear changes in the midline, but not for lateral facial points. In the midline, changes at the nasal base are most prone to prediction error. It is planned to continue to study the predictive accuracy of this software with a larger sample.
Funding
None declared.
Competing interests
One author, Sadie Glick, is an employee of 3D Systems, Inc., which uses Dolphin 3D imaging for virtual surgical planning. There were no financial incentives for participation in this study. Ethical approval
This study was approved by the Institutional Review Board of the Center for Applied Clinical Investigation at Boston Children’s Hospital (Protocol #00019505). Patient consent
Written consent was obtained for publication of the patient photographs. References 1. Mankad B, Cisneros GJ, Freeman K, Eisig SB. Prediction accuracy of soft tissue profile in orthognathic surgery. Int J Adult Orthodon Orthognath Surg 1999;14:19–26. 2. Tucker S, Cevidanes LH, Styner M, Kim H, Reyes M, Proffit W, et al. Comparison of actual surgical outcomes and 3-dimensional surgical simulations. J Oral Maxillofac Surg 2010;68:2412–21. 3. Chew MT, Sandham A, Wong HB. Evaluation of the linearity of soft- to hard-tissue movement after orthognathic surgery. Am J Orthod Dentofacial Orthop 2008;134:665–70. 4. Kim H, Jurgens P, Nolte LP, Reyes M. Anatomically-driven soft-tissue simulation strategy for cranio-maxillofacial surgery using facial muscle template model. Med Image Comput Comput Assist Interv 2010;13:61–8. 5. Liebregts J, Xi T, Timmermans M, de Koning M, Berge S, Hoppenreijs T, et al. Accuracy of three-dimensional soft tissue simulation in bimaxillary osteotomies. J Craniomaxillofac Surg 2015;43:329–35. 6. Schendel SA, Jacobson R, Khalessi S. 3dimensional facial simulation in orthognathic surgery: is it accurate? J Oral Maxillofac Surg 2013;71:1406–14. 7. Ullah R, Turner PJ, Khambay BS. Accuracy of three-dimensional soft tissue predictions in orthognathic surgery after Le Fort I advancement osteotomies. Br J Oral Maxillofac Surg 2015;53:153–7. 8. Wang S, Yang J. An improved finite element model for craniofacial surgery simulation. Int J Comput Assist Radiol Surg 2009;4:579–87. 9. Meiyappan N, Tamizharasi S, Senthilkumar KP, Janardhanan K. Natural head position: an overview. J Pharm Bioallied Sci 2015;7:S424–7. 10. Farkas LG. Anthropometry of the head and face. second ed. New York: Raven Press; 1994.
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016
YIJOM-3543; No of Pages 7
Dolphin 3D soft tissue prediction for Le Fort I 11. Jacobson A, Caufield PW. Introduction to radiographic cephalometry. Philadelphia: Lea & Febiger; 1985. 12. Jagadish Chandra H, Ravi MS, Sharma SM, Rajendra Prasad B. Standards of facial esthetics: an anthropometric study. J Maxillofac Oral Surg 2012;11:384–9. 13. Kaipatur NR, Flores-Mir C. Accuracy of computer programs in predicting orthognathic surgery soft tissue response. J Oral Maxillofac Surg 2009;67:751–9. 14. Yamada T, Mishima K, Moritani N, Janune D, Matsumura T, Ikeya Y, et al. Nasolabial morphologic changes after a Le Fort I osteotomy: a three-dimensional anthropometric study. J Craniofac Surg 2010;21:1089–95. 15. Jabar N, Robinson W, Goto TK, Khambay BS. The validity of using surface meshes for evaluation of three-dimensional maxillary and mandibular surgical changes. Int J Oral Maxillofac Surg 2015;44:914–20. 16. Khambay B, Ullah R. Current methods of assessing the accuracy of three-dimensional soft tissue facial predictions: technical and clinical considerations. Int J Oral Maxillofac Surg 2015;44:132–8. 17. Liebregts JH, Timmermans M, De Koning MJ, Berge SJ, Maal TJ. Three-dimensional facial simulation in bilateral sagittal split
18.
19.
20.
21.
22.
osteotomy: a validation study of 100 patients. J Oral Maxillofac Surg 2015; 73:961–70. Nadjmi N, Tehranchi A, Azami N, Saedi B, Mollemans W. Comparison of soft-tissue profiles in Le Fort I osteotomy patients with Dolphin and Maxilim softwares. Am J Orthod Dentofacial Orthop 2013;144:654– 62. Nkenke E, Vairaktaris E, Kramer M, Schlegel A, Holst A, Hirschfelder U, et al. Threedimensional analysis of changes of the malar-midfacial region after LeFort I osteotomy and maxillary advancement. Oral Maxillofac Surg 2008;12:5–12. Plooij JM, Swennen GR, Rangel FA, Maal TJ, Schutyser FA, Bronkhorst EM, et al. Evaluation of reproducibility and reliability of 3D soft tissue analysis using 3D stereophotogrammetry. Int J Oral Maxillofac Surg 2009;38:267–73. Shafi MI, Ayoub A, Ju X, Khambay B. The accuracy of three-dimensional prediction planning for the surgical correction of facial deformities using Maxilim. Int J Oral Maxillofac Surg 2013;42:801–6. Terzic A, Combescure C, Scolozzi P. Accuracy of computational soft tissue predictions in orthognathic surgery from
7
three-dimensional photographs 6 months after completion of surgery: a preliminary study of 13 patients. Aesthetic Plast Surg 2014;38:184–91. 23. Naudi KB, Benramadan R, Brocklebank L, Ju X, Khambay B, Ayoub A. The virtual human face: superimposing the simultaneously captured 3D photorealistic skin surface of the face on the untextured skin image of the CBCT scan. Int J Oral Maxillofac Surg 2013;42:393–400. 24. Xin P, Yu H, Cheng H, Shen S, Shen SG. Image fusion in craniofacial virtual reality modeling based on CT and 3dMD photogrammetry. J Craniofac Surg 2013; 24:1573–6.
Address: Cory M. Resnick Boston Children’s Hospital Department of Plastic and Oral Surgery 300 Longwood Avenue Boston MA 02115 USA Tel: +1 617 355 6082; Fax: +1 617 738 1657 E-mail: [email protected]
Please cite this article in press as: Resnick CM, et al. Accuracy of three-dimensional soft tissue prediction for Le Fort I osteotomy using Dolphin 3D software: a pilot study, Int J Oral Maxillofac Surg (2016), http://dx.doi.org/10.1016/j.ijom.2016.10.016