Egypt. J. Oral Maxillofac. Surg.

2010:1:7-11 (ISSN: 2090-097X)

Egyptian Journal of

Oral & Maxillofacial Surgery

Clinical Paper

Accuracy of profile prediction using computer software in orthognathic surgery

A. A. Barakat1, W. M. Refai2, S. T. Mekhemar1 Professor of Oral & Maxillofacial Surgery, Faculty of Oral & Dental Medicine, Cairo University Cairo, Egypt, 2Assistant Professor & Chairman, Orthodontic Department, Faculty of Dentistry, Minia University, Minia, Egypt.
1

A. A. Barakat1, W. M. Refai2, S. T. Mekhemar1

Postgraduate Student, Oral and Maxillofacial Surgery, 2Professor of Oral and Maxillofacial Surgery, 3Professor of Maxillofacial Radiology, 4 Lecturer of Oral and Maxillofacial Surgery, Faculty of Oral and Dental Medicine, Cairo University.
1

Abstract: Using computer softwares for prediction of orthognathic surgical outcome is gaining more popularity. Improving the accuracy of softwares and consequently improving the confidence in the surgical prediction is a crucial factor in the widespread use of such programs. This study aimed at studying the soft tissue profile prediction of Onyxceph software while eliminating any surgical inaccuracies in plan execution. The pre-operative cephalograms of 14 patients having orthognathic surgical procedure were superimposed on the post-operative ones and the exact surgical change in specific bony landmarks was calculated. Those measurements represented the exact surgical changes in the 2D cephalometric analysis. The surgical soft tissue profile prediction was based on this exact surgical change to yield a post-operative profile that was plotted against Key Words: Orthognathic surgery, the already achieved profiles. The possible reasons for the difference in measurements cephalometric analysis, presurgical prediction. between the surgical planning and the exact surgical changes were discussed and the accuracy of computer prediction was evaluated. Accepted for publication 25 June 2010.

Introduction Since the early 1990s surgical procedures had been employed to correct dentofacial abnormalities. As surgical techniques had improved, single and two jaw corrections had become common procedures to manage horizontal, vertical and transverse discrepancies in the facial skeleton. Modern orthognathic correction seeks to treat the facial deformity at its source by surgery to the relevant jaw. The criteria for success of an orthognathic surgical procedure are not only centered on the correction of skeletal and dental abnormalities, but also on an esthetic improvement as judged by both patients and practitioners. Defining esthetic aims obviously involves the problem of perception. What surgeons or orthodontists consider ideal may not be the same as the patients desires1. Prediction tracings should be an integral component in the planning for orthognathic surgery. Traditionally, cut-andpaste techniques of cephalometric acetate tracings had been used to develop a visual treatment objective2. There is now a correspondingly greater need for a method of a rapidly and accurately predicting the results of proposed treatment plans. Computer-aided diagnosis and treatment planning had received considerable attention lately as a mean of results prediction in orthognathic surgeries. The use of computers in orthognathic surgery is relatively recent. Schendel, Eisenfeld, Bell, and Epker3 were among the first to employ a computer system for the analysis of preoperative and postoperative soft tissue profile. Recent cephalometric investigations had found that movements of hard and soft tissue after orthognathic operations were strongly correlated horizontally but not vertically4. In addition, the position of the lips could not be predicted accurately5. It was also found that the dramatic improvement in the facial profiles of the bimaxillary surgery patients was primarily related to the backward movement of the mandible and the significant reduction in the superior lower lip area6.
2010 Egyptian Association of Oral & Maxillofacial Surgery. All rights reserved

Accuracy of profile prediction using computer software in orthognathic surgery

The prediction of treatment outcome is an important part of orthognathic planning. The orthodontic and surgical changes must be accurately described prior treatment in order to assess the treatment feasibility and optimize case management. Currently, there are several computer software packages programmed for such planning, claiming to predict the soft tissue outcome from surgical intervention. It would be important to assess the validity of these programs. The known ratios of soft-to-hard tissue movements can vary considerably in each individual7. Posposil8 investigated the reliability and feasibility of prediction tracing in orthognathic surgery and found errors to be more common in bimaxillary osteotomy. Nowadays, a variety of computerized analyses of lateral cephalographs are used to predict treatment changes in the antero-posterior and vertical facial planes, e.g. Dentofacial Planner, Opal TM Quick Ceph TM and Tipos TM9. However, most of these systems are still based on the lateral cephalometric radiographs. It is now possible to make large corrections of the jaws in the three planes of space and so pre-treatment planning is of paramount importance. More recently, there have been exciting developments in 3D and video imaging techniques but as yet the 2D profile system still remains the most widely employed prediction method. A series of programs had been developed since then for surgery prediction10 two-dimensional planning of the maxillofacial reconstruction with Le Fort I, II and III11, interactive threedimensional graphic techniques for the simulation of facial surgery12 and a video imaging technique as an aid for planning and counseling in orthognathic surgery13. The validity of the predicted soft tissue changes after Le Fort I osteotomy using the dentofacial planner (DFP) was investigated. The program proved to serve as good predictor of the individual soft tissue points. However, when these landmarks were used to construct a facial profile, a considerable difference in the predicted and actual profiles could occur. This was due to the significant mean differences detected between actual and predicted values for some landmarks. It was concluded that DFP needed better refined estimates of the intercepts in its regression equations associated with these landmarks1. Long-term changes in the soft tissue profile following mandibular setback surgery was investigated. The presence of factors that may influence the soft response to skeletal repositioning was also investigated. It was proposed that the database used in the prediction software be adjusted to account in an attempt to improve the accuracy of computerized treatment simulations14. The accuracy of preoperative OPALTM orthognathic predictions was assessed by retrospective analysis of 25 Class II patients who had orthodontic treatment combined with mandibular advancement osteotomy. There was a bias towards under-prediction of the vertical skeletal changes when there was more backward mandibular rotation than anticipated. Immediate postoperative cephalographs were also affected by a 2.1 mm mean downward displacement of the mandible as a result of the surgical wafer15. Moreover, the accuracy of the two-dimensional profile prediction produced by the computer-assisted simulation system for orthognathic surgery (CASSOS), for the correction of class III facial deformities. Correction was performed by

maxillary advancement or bimaxillary surgery. The results showed that a 2 mm horizontal difference in maxillary or mandibular soft tissue position was necessary before the expert and lay panel could detect a change. A 3 mm change was therefore defined as clinically significant16. A new approach using not only 3D surface models of the patients anatomy, but also a corresponding volumetric model, was discussed. This 3D planning software was used in the treatment of 15 patients and was found to provide a good correlation between simulation and postoperative outcome17. One of the most recently available software is the onyxceph. The accuracy of the computer systems currently in use for predicting the soft tissue profile after orthognathic surgery had not been fully investigated. It was the purpose of the present study to examine the validity of prediction of soft tissue changes after performing various types of orthognathic surgeries using the Onyxceph program. Material and Methods The study is a retrospective preliminary one.14 Patients were included (9 females and 5 males). Concerning surgical procedures, 8 patients had bimaxillary osteotomies while the other 6 ones had single jaw surgeries. Their age ranged from 20 28 years. Standard pre- and post-operative lateral cephalometric radiographs were obtained for each patient on the same cephalometric radiographic machine. Both the pre- and postoperative cephalograms were imported into orthognathic treatment-planning computer software OnyxCeph, Image Instruments GmbH. Both cephalograms were adjusted to scale using the rulers in the radiograph. The cephalograms for each patient were digitized and traced by identifying and placing a series of cephalometric points, both for the bony and soft tissue profiles. The computer software then performed a standard cephalometric analysis for each of the cephalograms. The preoperative tracing was then superimposed over the postoperative cephalogram using the S-N plane as the reference line. According to the type of osteotomy that has been actually carried out, a similar virtual osteotomy was selected from the software menu. The cut-out tracing of the virtually osteotomized bone segment was then manually translated and rotated to fit as accurately as possible on the actual postoperative bone position in the cephalogram, while the computer-software was allowed to predict the soft tissue profile changes for that movement. When a satisfactory result was achieved, the prediction tracing was then superimposed over the actual postoperative soft tissue profile tracing using the same plane; S-N; as a reference plane. The superimposed tracings were then printed as a hard copy in a 1:1 scale. On the hard copy a corrected S-N plane (SN-70) was used a reference for the vertical point measurements while a line perpendicular to the corrected S-N plane was used as a reference for the horizontal point measurements (Fig. 1).

A. A. Barakat et al.

Linear measurements were then taken between these reference lines and the following points: nasal tip, subnasale, upper and lower lips margins, stomion, soft tissue pogonion, soft tissue gnathion and finally soft tissue menton (Fig. 2) After measuring vertical and horizontal coordinates given for the soft tissue profile points, the deviation of the predicted coordinates from the actual ones were calculated. The whole procedure is shown in (Figs. 1-6).

Fig. 4: Treatment simulation with profile prediction.

Fig. 1: Preoperative cephalomatry.

Fig 5: Predicted profile superimposed on actual profile.

Fig. 2: Post-operative cephalometry.

Fig. 6: Superimposed profiles analyzed.

Results Our study included 14 patients, 9 females and 5 males. Eight patients received bimaxillary osteotomies while six had only single jaw surgeries. The patients ranged in age between 20-28 years. The mean differences between the predicted and actual soft tissue changes are explained in (Table 1).

Fig. 3: Preoperative tracing superimposed on postoperative cephalogram.

10

Accuracy of profile prediction using computer software in orthognathic surgery

Table 1. Pre and post-operative changes in the vertical and horizontal directions. Vertical Tip of Nose Subnasale Labrale Superioris Labrale Inferioris Labiomental fold Soft Tissue Pog Soft Tissue Gn Soft tissue Me Horizontal Tip of Nose Subnasale Labrale Superioris Labrale Inferioris Labiomental fold Soft Tissue Pog Soft Tissue Gn Soft tissue Me -2.33 0.167 -0.5 -1.875 -2 -3.25 -2.125 -3.125 Fig. 8: Mean difference between changes in the horizontal plane. 0.66 1.5 1.33 0.375 0 0.875 -0.25 0.75

However, when comparing between measurements in the horizontal and vertical planes, it was found that the predicted profiles were more accurate in the vertical than the horizontal plane. Although these detected accuracies and inaccuracies, the differences in both the vertical and horizontal planes were nonsignificant. Discussion From all the available computer prediction programs Quick Ceph, DFP, and CASSOS were the most frequently used programs. Each of these prediction programs has its own limitations, but ultimately one has to realize that the accuracy of the prediction software depends on the records obtained from the patients and their standardization15. Our review was aimed at finding the accuracy of soft tissue profile prediction using a rarely investigated computer software; OnyxCeph. In accordance with most articles, our technique of superimposing the preoperative hard tissue tracing onto that of the actual postoperative tracing before allowed accurate testing of the software soft tissue prediction accuracy away from the effects of intraoperative surgical errors and orthodontic tooth movements. Other authors implemented the laboratory data or intraoperative surgical records to simulate the hard tissue movements on the preoperative cephalogram and this might have been a source of error. Although most of our postoperative cephalograms were obtained 6 months after the surgery, we now think that this might have not been enough time for total disappearance of soft tissue edema that was reported to take as long as 1-2 years. Almost all studies in general; and again similar to our findings; showed accurate prediction outcomes (< 2 mm) when compared with actual results in both horizontal and vertical directions. Only for a few specific cephalometric points was the difference larger than 2 mm. The facial area where most of the larger prediction differences appear is the lower lip. The clinical implication is that we need to caution our patients about potential limitations in the prediction of horizontal and vertical changes in the lower lip area. Stated reasons for the discrepancies

The differences among all points ranged between 0 and 3.25 with a median of 1.1025 mm. The largest difference detected was in the chin area with mean horizontal differences between -2 and -3.25 mm. The program seemed to under-estimate the soft tissue profile changes. Comparing the mean difference between the predicted and actual changes in the soft tissue profile points in the vertical plane, the results revealed that highest accuracy was found in Sm, Gn and Me in the vertical plane (Fig. 7) In addition, when comparing the mean difference between the predicted and actual changes in the soft tissue profile points in the horizontal plane, it was detected that the highest inaccuracy occurred in points Sm, Pog and Me in the horizontal plane. (Fig. 8)

Fig. 7: Mean difference between changes in the vertical plane.

A. A. Barakat et al.

11

are different lip tonicity, length, posture, and mass. Although minor changes (< 2 mm) of the soft tissue profile have been shown not to be clinically detectable, accumulation of a number of these minor point inaccuracies in the vertical and horizontal directions may produce a clinically significant difference that can be detected by both professionals and laypersons equally. Another significant problem that is worth discussing is that the actual prediction software used subjects of different racial and ethnic backgrounds as norms in preprogramming the software database than our patient population. This point has been raised previously. The real impact of this variable is unknown. It is impossible to find an answer with the currently available data. Most reports; including ours, included a heterogenous patient population who received various combinations of osteotomies to correct a multitude of facial deformities18-20. However, it can generally be concluded that soft tissue prediction for single jaw surgeries was slightly more accurate than those for bimaxillary osteotomies. Further studies with larger samples size are required to figure out if those programs offer more accurate predictions for either maxillary or mandibular osteotomies. For clinicians it is important to understand the limitations we identified with the current software but also understand what the future may bring. With the progressive use of 3-dimensional (3D) technology, both to quantify hard tissue changes (CBCT) and soft tissue changes (stereophotogrammetry and/or laser scanning), the prediction programs currently being used will become obsolete in the near future. Absence of significant patient databases, limitations in the current software capabilities to manage 3D data, and lack of a full and proper understanding of how to analyze and manage the 3D data are still major barriers for its widespread use. Conclusions 1. The predicted profiles were more accurate in the vertical than the horizontal plane. 2. The highest inaccuracy occurred in points Sm, Pog and Me in the horizontal plane. 3. The highest accuracy was found in Sm, Gn and Me in the vertical plane. 4. Further refinement in the calculation algorithms and more race specific data will greatly improve the prediction outcome. 5. Despite those discrepancies still this software (Onyxceph software) is quite a helpful tool in orthognathic surgical planning. References
1. Konstiantos KA, OReilly MT, Close J. The validity of the prediction of soft tissue profile changes after LeFort I osteotomy using the dentofacial planner (computer software). Am. J. Orthod. Dentofac.Orthop. 1994:(3):241-9.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14.

15. 16.

17. 18.

19.

20.

Fish LC, Epker BN. Surgical-orthodontic cephalometric prediction tracing. J.Clin.Orthod. 1980: (1):36-52. Vannier MW, Marsh JL, Warren JO. Three dimensional CT reconstruction images for craniofacial surgical planning and evaluation. Radiology 1984:(1):179-84. Chew MT. Soft and hard tissue changes after bimaxillary surgery in Chinese Class III patients. Angle Orthod. 2005:(6):959-63. Koh CH, Chew MT. Predictability of soft tissue profile changes following bimaxillary surgery in skeletal Class III Chinese patients. J.Oral Maxillofac.Surg. 2004:(12):1505-9. Altug-Atac AT, Bolatoglu H, Memikoglu UT. Facial soft tissue profile following bimaxillary orthognathic surgery. Angle Orthod. 2008:(1):50-7. Freihofer HP,Jr. Changes in nasal profile after maxillary advancement in cleft and non-cleft patients. J.Maxillofac.Surg. 1977 :(1):20-7. Pospisil OA. Reliability and feasibility of prediction tracing in orthognathic surgery. J.Craniomaxillofac.Surg. 1987: (2):79-83. Cousley RR, Grant E, Kindelan JD. The validity of computerized orthognathic predictions. J. Orthod. 2003: (2):149,54; discussion 128. Harradine NW, Birnie DJ. Computerized prediction of the results of orthognathic surgery. J. Maxillofac. Surg. 1985:(6):245-9. Walters H, Walters DH. Computerised planning of maxillo-facial osteotomies: The program and its clinical applications. Br.J.Oral Maxillofac.Surg. 1986: (3):178-89. Moss JP, Grindrod SR, Linney AD, Arridge SR, James D. A computer system for the interactive planning and prediction of maxillofacial surgery. Am. J. Orthod. Dentofac.Orthop. 1988: 4(6):469-75. Sarver DM, Johnston MW, Matukas VJ. Video imaging for planning and counseling in orthognathic surgery. J.Oral Maxillofac.Surg. 1988:(11):939-45. Mobarak KA, Krogstad O, Espeland L, Lyberg T. Factors influencing the predictability of soft tissue profile changes following mandibular setback surgery. Angle Orthod. 2001: (3):216-27. Cousley RR, Grant E. The accuracy of preoperative orthognathic predictions. Br.J.Oral Maxillofac.Surg. 2004: (2):96-104. Jones RM, Khambay BS, McHugh S, Ayoub AF. The validity of a Computer-Assisted Simulation System for Orthognathic Surgery (CASSOS) for planning the surgical correction of Class III skeletal deformities: Single-jaw versus bimaxillary surgery. Int.J.Oral Maxillofac.Surg. 2007:(10):900-8. Westermark A, Zachow S, Eppley BL. Three-dimensional osteotomy planning in maxillofacial surgery including soft tissue prediction. J.Craniofac.Surg. 2005 :(1):100-4. Kolokitha OE, Athanasiou AE, Tuncay OC. Validity of computerized predictions of dentoskeletal and soft tissue profile changes after mandibular setback and maxillary impaction osteotomies. Int.J.Adult Orthodon. Orthognath. Surg. 1996: (2):137-54. Gerbo LR, Poulton DR, Covell DA, Russell CA. A comparison of a computer-based orthognathic surgery prediction system to postsurgical results. Int. J. Adult Orthodon. Orthognath.Surg. 1997:(1):55-63. Loh S, Heng JK, Ward Booth P, Winchester L, McDonald F. A radiographic analysis of computer prediction in conjunction with orthognathic surgery. Int.J.Oral Maxillofac.Surg. 2001: (4):25963.