This study presents a computer-assisted planning system for dysgnathia treatment. It describes the process of information gathering using a virtual articulator and how the splints are constructed for orthognathic surgery. The deviation of the virtually planned splints is shown in six cases on the basis of conventionally planned cases. In all cases the plaster models were prepared and scanned using a 3D laser scanner. Successive lateral and posterior-anterior cephalometric images were used for reconstruction before surgery. By identifying specific points on the X-rays and marking them on the virtual models, it was possible to enhance the 2D images to create a realistic 3D environment and to perform virtual repositioning of the jaw. A hexapod was used to transfer the virtual planning to the real splints. Preliminary results showed that conventional repositioning could be replicated using the virtual articulator.
In many fields, computer-assisted planning has enabled surgeons to perform complex interventions . New techniques, such as navigated surgery have been established in craniofacial surgery.
To improve treatment and results, systems such as Gaertner’s virtual articulator have been used to simulate static or dynamic occlusion. This project used natural reference points acquired by means of a pointer device. The relation between the models is established by means of a registration imprint and movement is recorded by a jaw-motion-analyzer (Zebris Medical GmbH, Isny, Germany), but the results cannot be retransferred to reality.
In another project , plaster models were set up in the articulator; they were conventional in respect to the calibrated double-base method – muensteraner model operation system (KD-MMS). The models were later placed into the hexapod and transformed by an operation planning tool to find the optimum position. The pneumatically fixed models were fixed into their final position by applying plaster. The surgeon can reference the real cement models and the displayed transformation data in the program.
Most comparable projects are restricted to soft-tissue simulation based on CT data with the corresponding exposure dose. Even if the jaw is reposition, the main aim of these projects is improved postoperative aesthetic appearance and function of the patient’s stomathognatic system.
The project presented here requires only two routine cephalometric images, significantly minimizing the X-ray and radiation dose to the patient, and non-invasive fixation of the patient during exposure. Using the two images, the authors developed a virtual articulator including anatomical registration and controlled repositioning of the jaws. The transformation ends with the creation of a real-world splint for intra-operative use. The aim is to improve the accuracy of orthognathic surgery by advancing 2D images with the addition of the third dimension, and using the enhanced 3D environment in the computer-assisted planning system.
Material and methods
Six patients whose treatment had been planned conventionally were divided into two groups of three patients; one group requiring reposition of a single jaw and the other repositioning both jaws. The radiological images described below were taken from all six patients, and two pairs of plaster models of the jaws were built. One model pair was used for conventional planning using a semi-adjustable articulator and the other pair for the virtual planning method. The implemented virtual articulator orientates on the selected rotation points of the caput mandibulae.
In the conventional method, a wax imprint of the patient was created and, to achieve an exact anatomical constellation, a face bow (KaVo, Biberbach, Germany) was used.
Patients were examined routinely by means of a digital radiological device (ORTHOPHOS XG Plus , Sirona, Bensheim, Germany) with a cephalometric attachment using mainly Sidex version 5.54 (Sirona, Bensheim, Germany). This attachment has a three-point fixation unit (left/right porion and glabella) which is fixed to a rotation table at the top of the cephalometric attachment giving a virtual rotation axis through the patient. With the help of this unit the patient is always positioned at the same distance from the image-plane and in the field of view of the X-ray. For this study, one digital image was taken from the posterior-anterior and one from the lateral direction.
Positioner and initialization object
In order to be able to transfer the virtually planned splints into reality, the plaster models have to have a well-defined position and orientation, which has to be reproducible in the virtual environment.
The X1med3D positioner (Med3D, Heidelberg, Germany; Fig. 1 ) has six arms enabling the user to set up every movement (including rotations, tilting, moving) with an accuracy of 1/10 mm (manufacturer’s information). On the arms, there is a scale for setting up defined positions of the table. The table consists of an embedded Adesso-split (Baumann Dental GmbH, Keltern-Ellmendingen, Germany) and three linear, independent landmarks that simplify the referencing of the center and orientation of the table. Above the table, there is a stamp-like attachment ( Fig. 1 ) with an Adesso-split.
For specifying a fixed working area and to fix the orientation of the stamp, an initializing object was created ( Fig. 1 ). It consists of a cylinder with a height of 10 cm with a fixed Adesso-split plate on either side. The split plates are mirrored on an axis going through the center of the cylindrical object. By positioning this object on the table, setting all arms of the hexapod to the same length and pushing the stamp onto the object, the working area was defined and was fixed by the distance holder ( Fig. 1 ). The adjustable range is between 7.5 cm and 14 cm.
3D laser scanner
Scanning of the defined models was performed by a DigiScope (3D Alliance Inc., Bischoffen, Germany), which is a laser-based 3D coordinate-measuring device with a measuring volume of 200 × 200 × 200 to 600 × 600 × 600 mm with 3 (maximum 6) movement axes. The digital optical sensors have a recurrence accuracy of 8 μm.
For measurement of the deviation between conventional and computer-assisted planned splints, a MicroScribe G2X (Immersion Corporation, San Jose, USA) was used. This is a high-performance sensor for tracking the position and orientation of the stylus tip. Its working space is a sphere of 1.27 m and it has an accuracy of 0.23 mm measured on 100 points ANSI sphere (manufacturer’s information).
The scheme ( Fig. 2 ) shows the different stages, such as data acquisition, preparation of registration, reposition and return of the planned splints into the real world. Data acquisition and preparation of the models are described above.
Planning was carried out by a surgeon and a technician. The surgeon identified the points on the models and images and undertook the planning. The technician used the same segmented points and could refer to the real models and the conventionally planned splints.
Preparation of the models
After initializing the X1med3D ( Fig. 1 ), the mandible and the maxilla were cemented to the Adesso assembly disc using the wax imprint or, in two cases, the conventionally planned starting splint for fixing the relation. The stamp was fixed in this position, resulting in a horizontal-only volume expansion while the cement dries.
Through this procedure, the authors held the stamp at a set distance and fixed the models at a specific distance and defined unambiguous orientation to the plate.
Transfer of the models into the virtual word
Before the scanning procedure, the plate of the hexapod carrying the cement model was mounted onto the table of the scanner. Starting with the three landmarks on the plate, each model was scanned from different sides with a small advancement of the scanner and a high resolution using the ScanOS (3D Alliance Inc., Bischoffen, Germany) software.
For performance reasons, the number of points was reduced from 2.75 million to a maximum of 362,000. This was achieved by removing doubled points, unneeded points and by a substantial reduction of points at less important sites, such as the palate. The software used was Rapidform2006 (INUS Technology, Seoul, Korea). The number of points was reduced and the surface smoothened, without geometric loss.
Transformation of the 2D images into a 3D environment
In order to be able to reference radiological images to each other at the time the patient was scanned, information on the fixed geometric metrics of the ORTHOFLOW XG PLUS and on the images’ size and resolution was required. With the information on the size and resolution, which can be found in the Sidex files, the images can be aligned ( Fig. 3 ). For finding the rotation axis of the images, the user has to pick the porions in each of the images, which can be accomplished in a separate dialog showing the appropriate image in full resolution. Reconstructing the path of the X-ray, beginning at the center of the line connecting the left and right porions of each image and the virtual source of the X-ray, at a distance of 230 mm (manufacturer’s information) the intersection with the virtual rotation axis can be found. By bringing these points onto one point and rotating one of the images by 90° the 3D environment is established ( Fig. 3 ). The scene now shows the situation at the time the patient was scanned, including the rotation the patient performed before the second image was taken.