Abstract
This study tests computer imaging software (SurgiCase-CMF ® , Materialise) that enables surgeons to perform virtual orthognathic surgical planning using a three dimensional (3D) utility that previews the final shape of hard and soft tissues. It includes a soft tissue simulation module that has created images of soft tissues altered through bimaxillary orthognathic surgery to correct facial deformities. Cephalometric radiographs and CT scans were taken of each patient before and after surgery. The surgical planning system consists of four stages: CT data reconstruction; 3D model generation of facial hard and soft tissue; different virtual surgical planning and simulation modes; and various preoperative previews of the soft tissues. Surgical planning and simulation is based on a 3D CT reconstructed bone model and soft tissue image generation is based on physical algorithms. The software rapidly follows clinical options to generate a series of simulations and soft tissue models; to avoid TMJ functional problems, pre-surgical plans were evaluated by an orthodontist. Comparing simulation results with postoperative CT data, the reliability of the soft tissues preview was >91%. SurgiCase ® software can provide a realistic, accurate forecast of the patient’s facial appearance after surgery.
Interest in modelling human soft tissue is growing for a wide range of applications such as physiological analysis, surgery planning, or interactive simulation for training purposes. In orthognathic surgery, facial deformities are corrected by repositioning bone segments. The aim is to correct skull deformities and dental occlusion to improve the aesthetic appearance of the patient’s face. Modelling of the visco-elastic behaviour of soft tissue is a key element in accurately predicting the surgical outcome.
Three-dimensional (3D) computer simulation can be useful to preview the surgical outcome in orthognathic surgery. C utting et al. first described a method for computer-assisted design for craniofacial surgical procedures considering 3D cephalometric constraints. Y asuda et al. proposed a computer system for craniofacial surgical planning based on computer tomography (CT) images in brachycephalic cases. A ltobelli et al. applied interactive repositioning of hard tissues using cephalometry and anthropometric datasets.
Several models of soft tissue deformation have been documented in previous research: K och et al. proposed a mass-spring model as a collection of point masses connected by linear or non-linear springs in a lattice structure, with no realistic description of the physical behaviour of the human tissue. Several methods to simulate the elastic nature of soft tissues have been proposed. T erzopoulos et al. , P latt & B arr and W aters used deformable models. B ro -N ielsen & C otin studied the problem of reducing computational time using a condensation technique. D elingette et al. represented the skull, muscles and skin in the patient’s face as 3D surfaces of complex topology, with three or four connected simplex meshes and characterized with a constant vertex-to-vertex connectivity. The tissues between the skin and skull were modelled with muscles organized in layers that would deform the patient’s face in response to surgery on the skull. Delingette’s method requires the operator’s intervention for the application of muscles.
More accurate simulations were based on the finite element technique. K eeve et al. first presented an anatomy-based 3D finite element tissue model integrated into a computer-aided surgical planning system. This method forecasts alterations in soft tissue resulting from the realignment of the underlying hard tissue, without considering the exact anatomical structures. G ladilin et al. described the biomechanical behaviour of soft biological tissues as an isotropic homogenous postoperatively and linear elastic continuum with the linear elastic model based on efficient finite element methods. Heuristic modelling of the facial muscles was applied to estimate the patient’s facial appearance. This method produced accurate results and W estermark et al. have validated it in a number of cases .
M ollemans et al. compared the use of four different computational strategies for 3D soft tissue measurements in maxillofacial orthognathic surgery. They found that the highest accuracy was obtained by using the linear finite element model or the linear mass tensor model. The average median distance comparing the preoperative and postoperative position was only 0.60 mm and the average 90% percentile stayed below 1.5 mm.
S arti et al. proposed a method for computing soft tissue deformation in craniofacial surgery from CT images without any intermediate geometric mesh. That system worked on the fine anatomical structure generated by CT. The model was generated from the segmentation of the CT volume in three different types of tissues: bones, soft tissues and embedding materials. Each region was modelled with a different type of equation to obtain a ‘simulation algorithm’ to predict the behaviour of the soft tissue. A physically based simulation module computed the soft tissue deformations caused by the new bone geometry. The displacement of soft tissues has been modelled as classic continuous mechanical equations. In particular, the modified linear elasticity equations system was adopted for isotropic materials with embedded boundary conditions , according to the in vivo mechanical properties of skin and muscle described by B lack & H astings . Soft tissue is usually assumed to behave as a linear, elastic, isotropic material. This assumption is probably reasonable for small strains and at a spatial scale that is large compared with the relative correlation length of the elastic variability in the tissue sample as proposed by K rouskop et al.
S arti et al. suggested that simulation of morphological modifications in the patient’s face following bone repositioning could greatly improve surgical planning. The most straightforward approach for validating the reliability and reproducibility of surgical planning software is to compare it with preoperative and postoperative data from clinical cases.
M archetti et al. in a clinical and surgical study reported a reliability and reproducibility rate of 80% using VISU software. The Belgian company, Materialise, has developed a simulation software package, SurgiCase-CMF ® , that includes VISU soft tissue simulation algorithms. The aim of this paper is to use statistical analysis to evaluate whether this software is a useful tool for forecasting the aesthetic impact of soft tissue movements in dento-skeletal malocclusions, such as maxillary hypoplasia and mandibular prognathism.
Patients and methods
10 patients (5 men; 5 women) were recruited for the study. They had dento-facial deformities and had undergone 3D CT scanning before, and 6 months after, surgical correction with classic osteotomies: Le Fort I maxillary osteotomy, bilateral sagittal split mandibular osteotomy and genioplasty ( Table 1 ). 3D CT images of hard and soft tissues had been recorded before surgery. These were submitted to simulate orthognathic surgery, which made it possible to observe the postoperative surgical outcome and the quality of the surgical simulations. It was assumed that after 6 months, soft tissue postoperative oedema was resolved.
| Pts | Maxillary reposition Le Fort I | Mandibular reposition Obwegeser – Dal Pont | Genioplasty |
|---|---|---|---|
| 1 | 4 mm advancement 3.5 mm left shifting |
Mandibular set back occlusion |
|
| 2 | 4 mm advancement 3 mm vertical reduction |
Mandibular set back occlusion |
3 mm advancement |
| 3 | 5 mm advancement 3 mm vertical reduction |
Mandibular set back occlusion |
4 mm vertical reduction 2 mm advancement |
| 4 | 6 mm advancement 4 mm vertical increase |
Mandibular set back occlusion |
6 mm vertical increase 4 mm advancement |
| 5 | 5 mm advancement 3 mm vertical reduction |
Mandibular set back occlusion |
|
| 6 | 5 mm advancement 5 mm transversal posterior expansion |
Mandibular anti clock wise rotation |
|
| 7 | 8 mm advancement | Mandibular set back occlusion |
|
| 8 | 5 mm advancement 2 mm vertical right reduction 2 mm vertical left increase |
Mandibular set back occlusion |
|
| 9 | 6 mm advancement | Mandibular set back occlusion |
5 mm vertical reduction |
| 10 | 2 mm advancement 7 mm vertical increase |
Mandibular set back occlusion |
6 mm vertical increase 4 mm advancement |
Multi-slice imaging data were obtained before surgery using a CT unit (16 slices Light Speed; General Electric Healthcare, Milwaukee, WI, USA) in helical mode. The following parameters were applied: pitch 1, 120 kV, 90 mAs, helicoidal st 0.6, 16 × 1.25, 0.8 mm.
The Frankfort plane of the supine patient was vertical in relation to the horizontal plane of the CT. The CT sections of the field of view must have the same field of vision which must include all the areas of interest and must be acquired in the same direction, maintaining space between the sections. Request of inclination gantry was equal to 0°.
The data were stored in DICOM format and automatically imported into SurgiCase-CMF ® PRO 1.2. SurgiCase’s semiautomatic tools processed two-dimensional (2D) CT slices from real patients to obtain 3D volumes for surgery simulation. The image was processed, cleaned of artefacts, and the hard and soft tissues were segmented. Threshold levels were set for creating a virtual 3D model of the hard and soft tissues. The texture mapping utility made it possible to superimpose a photo on the CT scan, which allowed a view of the patient’s skin tone. The user indicates a number of landmark points on the frontal image along with their corresponding points on the 3D model of the soft tissues. After indicating a set of points, the software automatically wraps the 2D picture on top of the 3D surface. The wrapping algorithm is a radial base function-based interpolation algorithm that matches the texture perfectly at corresponding points and minimizes the deformation in the area between the points .
Osteotomy lines were traced on the 3D virtual skeletal CT image so that anatomical regions could be moved and relocated. The user can easily indicate cutting paths that simulate the bone cuts the surgeon will perform during the operation. The size of the cuts and the thickness of the cutting blade can be specified. After making the necessary cuts, the user can move different 3D objects in relation to one another. The user can move parts by entering the desired direction (anterior, posterior, left, right, cranially, caudally) or can select them with the mouse. At any time, the user can obtain an overview of the total overall movement of a specific anatomical region. Bone segment movements are quantified in terms of translational and rotational parameters (millimetres and degrees). For example, it is possible to relocate the centre of rotation to a different area of teeth or palate, which allows highly precise planning and ensures the accuracy of the simulation. The surgeon is able to check new occlusion and skeletal balance, and any bone interference or asymmetry in the new bone and dental position. After surgical planning a hypothesis for new bone geometry is put forward.
SurgiCase-CMF ® computed the soft tissue deformation caused by new bone geometry using a physically based, previously published simulation module. The same graphical interface makes it possible to preview the patient’s new facial appearance.
The authors had to calculate errors by comparing the distance between the soft tissue simulation and the 6-month postoperative 3D soft tissues CT data for each point ( Fig. 1 ).
