3.2.1 Cranio-maxillofacial biomodelling

The complexity of cranio-maxillofacial anatomy combined with the morphological variation encountered by the reconstructive surgeon makes cranio-maxillofacial surgery a conceptually difficult task in explanation, planning and execution. The need for clear morphological understanding in such surgery played a large part in the development of both 3D imaging and solid biomodelling. The ideal method of displaying cranial anatomy is the patient's dried skull itself. The biomodelling system developed by Anatomics™ has now come very close to achieving this.

Cranio-maxillofacial surgeons have shown a high level of affinity for biomodelling, perhaps because they routinely use dental casting and the articulated models produced from such castings. The use of biomodels of the mandible and maxilla is a natural extension to the use of these dental castings. The biomodels have been commonly used to improve the diagnostic relevance of the data and for surgical simulation. A highlight of the use of biomodels in cranio-maxillofacial surgery has been the ready acceptance of biomodelling by the surgeons (D'Urso et al., 1998; Arvier et al., 1994; D'Urso et al., 1999a). Consequently they have developed many interesting applications.

In craniofacial surgery, biomodels have been traditionally used by surgeons to gain insight into unusual or particularly complex cases. Often in craniofacial surgery a multidisciplinary team is involved. The biomodels are often used to assist communication among team members, to achieve informed consent from patients and relatives and between surgeons intraoperatively. Biomodels have also been used to simulate craniofacial reconstruction. Standard surgical drills and saws can be used to fashion osteotomies. The bone fragments can be reconstructed using wire, plates and screws or glue. A technique that has been developed to assist with the reconstruction involves the manufacture of two biomodels (D'Urso et al., 1998). The first biomodel is reconstructed by surgeons and acts as an 'end-point' biomodel, illustrating the desired preplanned reconstruction. Surgeons then mark on the second biomodel the osteotomy lines that they intend to use. This biomodel becomes the 'start-point' biomodel. Intraoperatively the two biomodels are used in different ways. The start-point biomodel is used to navigate the anatomy and accurately to transfer the osteotomy lines onto the exposed patient's skull. The osteotomies are then made and the pieces of bone reconstructed according to the end-point biomodel. Such reconstruction, as well as the shaping of bone grafts, can be performed by a second surgeon working at a side table while the other surgeon continues to operate on the patient. This technique has been reported to improve the results of surgery as well as shorten the operating time. The end- and start-point biomodels have also proven very useful for informed consent and team communication.

Maxillofacial surgeons have also developed some interesting specific applications of biomodelling that warrant further discussion (Arvier et al., 1994). These are as follows:

  • integration of biomodels with dental castings to form articulated jaws;
  • the use of biomodels to shape prosthetic and autograph implants;
  • the use of biomodels to prefabricate templates and splints;
  • the use of biomodels in restorative prosthetics;
  • the use of biomodels to plan distraction osteogenesis. Integration of biomodels with dental castings

The combination of dental casting (replicating the teeth) with biomodels (replicating the jaws) may be advantageous for the following reasons:

  • The presence of metal artefact from dental fillings and braces is translated into the biomodel which can result in unacceptable inaccuracies. CT scanning in planes horizontal to the artefact source can minimize the number of planes in which image distortion occurs but may still not be sufficient.
  • The accuracy of CT scanning even at high resolution may not be sufficient for the biomodelling of opposing teeth to form a truly accurate occlusive bite.

Traditional plaster dental casting creates a highly accurate, relatively cheap model of the teeth. A combination of these castings with a biomodel mandible and maxilla forms a highly accurate hybrid with considerable benefits. Such a hybrid biomodel may be made by creating a wax key to localize the mandibular dentition in relation to the mandibular body (Arvier et al., 1994). The dentition is then cut away and the key used to locate the plaster casting in relation to the mandible biomodel. In a similar way the upper dental cast may be combined to the maxilla. A temporomandibular articulator may then be created by accurately occluding the dental arches about the temporomandibular joint. Such an articulated biomodel is particularly helpful in orthognathic surgery where the effects of osteotomies on the dental occlusion may be preoperatively examined.

Another way to avoid the difficulties associated with biomodelling the teeth and jaws is to use a laser scanner to create a surface file from the dental casting and then superimpose the file with data derived from CT. In this way an accurate dataset describing the teeth and jaws may be generated that could be used directly to biomodel the teeth and jaws from SL. Such a biomodel would accurately reproduce the dentition as the laser scanning may be performed with far higher resolution than CT scanning. The difficulties with this approach are the accurate co-registering of the two differing datasets in the correct anatomical position and acquisition of dental data that may be inaccessible to the laser scanner. Use of biomodels to shape maxillofacial implants

Biomodels may be used in several ways to shape maxillofacial prosthetic and autographic implants. A simple way to do this is to use the biomodel as a template on which a graft may be directly shaped intraoperatively (D'Urso et al., 1998; Arvier et al., 1994; D'Urso et al., 1999a). The bone graft may be harvested from the iliac crest and shaped directly on the sterilized biomodel. Once the contouring is satisfactory, the surgeon places the graft in situ and fixes it. This approach can dramatically reduce operating time while improving the end result. This can be achieved because the surgeon can shape the graft on the biomodel while the assistant simultaneously prepares the exposure of the donor site. This technique also avoids the need for repeated 'fitting and chipping' of the graft when the patient is directly used as the template, since direct shaping is restricted by soft tissue cover and limited surgical access.

Another approach is to use acrylic, or a similar material, preoperatively to create a master implant to serve as a guide for the shaping of the bone graft intraoperatively. This is particularly appropriate when the graft requires a somewhat complex shape. The surgeon can minimize operating time by preoperatively moulding the acrylic to the exact shape required, using the biomodel as the template.

More recently, synthetic bioabsorbable materials, such as polygalactic acid, have been introduced into cranio-maxillofacial surgery. Such materials in the form of sheets or plates can be intraoperatively shaped to fit specific anatomy. As with autograph implants, such bioab-sorbable implants can be shaped to a biomodel to save time and avoid difficulties with limited anatomical exposure. Use of biomodels to prefabricate templates and splints

Biomodels may be used to plan endosseous surgery and to create customized drill guide templates (Arvier et al., 1994). Edentulous patients may have teeth restored by mounting them on titanium pins which are implanted into the jaw. The implantation of the titanium pins, however, can be difficult and can be complicated by damage to the underlying dental nerve. Mandibular biomodels accurately replicate the neurovascular canal through which the mandibular nerve travels. The course of this nerve may easily be displayed by passing a malleable coloured wire along the neurovascular canal or replicated in a second colour using StereoCol resin. The biomodel can then be used to determine and rehearse the positioning and depth of the holes required to receive the titanium mounting pins using a standard dental drill. The pins can then be inserted into position and 'cold-cure' acrylic moulded around them and the mandibular contour to form a relocatable drill guide. The depth of each hole can also be determined relative to the drill guide and recorded. During the surgery the mucosa is stripped from the mandible and the drill guide matched using the reciprocal contours. While firmly held in place, the guide can be used to drill the holes with the correct positioning and depth as preplanned in the biomodel without risk of injury to the underlying mandibular nerve. More recently, SimPlant software has been used preoperatively to simulate the placement of implants in a virtual enviroment. Rapid prototyping is then used to generate a biomodel of the jaw as well as a custom drill guide template (Vrielinck et al., 2003).

Dental splints may be prefabricated using articulated biomodels of the teeth and jaws. Such splints may be useful to maintain the relative position of the dental arches after osteotomy surgery. The surgery is rehearsed on the biomodel and the relative position of the bones determined and set. A splint may then be moulded to fit to the reconstructed biomodel. At the end of surgery the splint may then be used to maintain positioning. This technique can save the time taken at the end of the procedure to mould such a splint. The risk of bony movements while moulding the splint directly is also avoided with this technique (Arvier et al., 1994).

Distraction osteogenesis is a technique used to promote the remodelling and lengthening of bones. The technique uses an implantable device slowly to distract an osteotomized bone by about 1 mm per day. Biomodels have been used to plan the positioning of the distraction device as well as to determine the extent of distraction to achieve the desired result. The use of biomodels has proven extremely valuable in this regard (Yamaji et al., 2004; Whitman and Connaughton, 1999). Use of biomodels in restorative prosthetics

Yet another use of biomodels in maxillofacial surgery is for the prefabrication of restorative prosthetics. Surgery usually requires titanium fixative implants to be inserted on which a prosthesis is mounted. A biomodel can be used to determine the best location for these implants and to construct the overlying prosthesis. This may be performed by inserting the mounts into the biomodel, constructing the overlying wire scaffolding and then adding the plastic nasal and dental prostheses. A biomodel is therefore used not only to plan and rehearse implantation but also for the construction of the prosthesis. This improves the ability to form an accurate prosthesis, which enhances the cosmetic result and shortens the operative time. Another advantage is that less time is required to shape and trial fit the prosthesis, as much of this can be performed on the biomodel.

Biomodelling has also been used to generate a prosthetic ear replacement. CT scanning is performed and the normal ear is mirror imaged. A biomodel is then used as a master to cast a synthetic substitute ear which is fastened via osteointegrated titanium screws.

3.2.2 Use of real virtuality in customized cranio-maxillofacial prosthetics

The sight of unfortunate patients with disfiguring skull defects on the neurosurgical and rehabilitation wards has always been disturbing. These defects are not only cosmetically displeasing but also pose a significant risk to the patient's underlying brain should trauma occur. The method traditionally used by neurosurgeons, before the advent of biomodelling, to repair cran-iotomy defects (for which no autologous bone is available) is to shape cold-cure acrylic to fit the defect. This acrylic is moulded and polymerized in situ to form the implant. The technique is limited by the artistic skill of the surgeon to achieve the desired contour of the implant within the short time before polymerization. This moulding process can be time consuming, especially if the surgeon takes several attempts to achieve the desired contour. Longer surgical time increases infection risk. Infection is a major complication in any implant surgery as its presence will usually necessitate removal of the implant. Another disadvantage of cold-cure acrylic monomer is its toxicity. If polymerization, is incomplete, monomer can leach into the patients tissues with detrimental effects.

The application of real virtuality to assist in the manufacture of customized prostheses was realized at an early stage (D'Urso et al., 1994). Early techniques were based on the creation of CNC milled models that could be made into a mould from which the implant could be cast (White, 1982; Toth, ellis and stewart, 1988). A model could also act as a template. Wax could be moulded to create a master implant that could then be used to create a mould from which the implant could be cast. Alternatively, the model could act as a template over which titanium (Blake, MacFarlaneandHinton, 1990;Joffe etal., 1992; Abbott etal., 1994)orhydroxylapatite (Waite et al., 1989) could be moulded to generate an implant. Such techniques have also been applied to construct prostheses in maxillofacial surgery and orthopaedic surgery (Rhodes et al., 1987).

New methods, using biomodelling, to construct cranioplastic prostheses have been developed. A master is generated by a computer graphic mirroring process whereby the normal side of the cranium can be manipulated to produce an exact contour of the 'missing part' of the skull. This 'missing part' and the biomodel of the craniotomy defect thus form both the male master implant and the female biomodel of the patient's skull. The SL master implant can be hand finished exactly to fit the biomodel of the craniotomy defect, which has a high degree of accuracy (D'Urso et al., 2000; Barker, Earwaker and Lisle, 1994). Any small irregular features around the edge of the master implant are smoothed to create a fit with the biomodel achieving maximum contact. The SL master implant may then be used to create a mould from which an acrylic implant is cast.

If the defect extends beyond the mid-line, several techniques can be used to generate a master implant. Computer interpolation from existing anatomy can be used to fill the deficit. The superimposition of a standard skull dataset with the patient's skull can be performed. The patient's anatomy is then subtracted to leave a master implant. The simplest and most effective technique involves the direct sculpting of a wax master to fill the defect.

The use of rapid prototyping to manufacture both male and female prosthetic components simultaneously has significant advantages over previous techniques:

  • it negates the need for wax sculpting of the master;
  • it directly provides a master implant;
  • it can be used with all known prosthetic materials.

Methyl methacrylate (acrylic) has been long accepted for use in cranioplasties (Manson, Crawley and Hoopes, 1986; Remsen, Dawson and Biller, 1986). Other materials allow the ingrowth of bone, such as hydroxylapatite (Waite et al., 1989), ceramic (Kobayashi et al., 1987) and ionomeric bone cement (Ramsden, Herdman and dye, 1992). These materials are highly expensive, somewhat difficult to mould or shape and, if infected, can be difficult to remove in toto. Titanium is also used commonly (Eufinger et al., 1995; Blake, MacFarlane and Hinton, 1990; Joffe et al., 1992; Abbott et al., 1994) and has the advantage of being biologically inert, although this is offset by cost, difficulties in moulding, casting or milling and the artefacts generated in CT and MR after implantation. Computer mirroring techniques for the generation of prostheses

Mirror imaging techniques have been applied frequently to assist the assessment of asymmetrical deformities of bone and soft tissues in the cranio-maxillofacial region. In order to isolate the dimensions of a missing part of anatomy, digital subtraction mirror imaging may be used (Fukuta and Jackson, 1992). In this technique the CT data are segmented to isolate the tissue of interest, usually bone. The original image is saved. A mirror image of the image about the sagittal plane is then created and its abnormal half erased. The normal half of the original image is also erased. The original image (containing the missing anatomy) is then subtracted from the normal mirrored image. This should leave the missing anatomy alone, which can then be used to generate a 'master' implant. The master implant data and original data are then used to manufacture biomodels. Alternatively, the master implant data may be used directly to fabricate the implant. CNC milling of titanium or the use of selective laser sintering have been reported in this regard.

As the mirroring technique is essentially a two-dimensional approach to a three-dimensional problem, the symmetry of data acquisition from the patient is critical. CT scanning must be performed without gantry tilt and patient movement, and with isoaxial positioning of the patient. These criteria are difficult to attain in routine scanning as it is extremely difficult to place the patient's head in the scanner without slight rotation or movement occurring. Even one or two degrees of misalignment is significant. The plane of symmetry in the images is especially difficult to determine when the routine reference landmarks such as orbit or nose are deformed. Natural asymmetry may complicate digital subtraction mirror imaging in any particular patient. Each slice needs to be checked by the operator to determine the influence of the natural asymmetry, and allowances and alterations may be necessary to accommodate for this which can become time consuming. If rotation of the patient occurs during CT scanning it is possible to reformat an asymmetrical CT scan. Complex algorithms have been developed to solve many of these issues, and the virtual creation of the 'missing' anatomy is possible but often time consuming and limited by possible error.

It should also be noted that any inaccuracies created in a master implant can be readily corrected by burring away material and/or adding wax to it so that it fits exactly into the craniotomy defect reproduced in the biomodel. In this way, hours of computer work to attain an accurate master can be accomplished within minutes by hand in the workshop. This observation highlights the utility of the solid master implant compared with the virtual master implant. The generation of a master implant has another advantage in that a thoroughly validated material, such as acrylic, can be used to fashion the final prosthesis.

In our experience the best method for the generation of customized implants is for a pros-thetist to sculpt the master from wax or directly from titanium on the biomodel itself (D'Urso etal., 2000). This allows for subtle anatomical variations to be incorporated as well as allowing for the surgeon's input in the customization process. Again, the use of real virtuality appears to have more utility than a virtual alternative.

Once generated, the implant is sterilized by gas or autoclaving. The implant is used by the surgeon in conjunction with the biomodel of the region to determine the exact attachment sites and means. Attachment is usually achieved with titanium miniplates and screws. Titanium miniplates and screws can also be preoperatively attatched to the implant to minimize operating time. Results of implantation

In our experience of over 160 implants manufactured by Anatomics™, all have fitted well except for one. The implants required only minor trimming to achieve good contact between the bevelled edge of the implant and the bone defect. The fit after such minor trimming was consistently within 1.5 mm. The use of prefabricated implants was reported to reduce operating time compared with the traditional cold-cure method. Surgeons have reported that the ability to study the implant and the biomodel preoperatively allows fixation to be exactly planned and possible problems to be identified, e.g. the site of cranial sinuses to be determined (D'Urso etal., 2000).

All the patients were happy with the cosmetic result, and several commented that they could not even tell that a cranial implant was present. The patients were particularly pleased to have the opportunity to examine their implant before surgery. The implant and biomodel were helpful in obtaining informed consent.

Only one surgeon reported that significant trimming of around 5 mm was required to fit the implant to a large frontal defect. It is not clear why this happened. The most likely cause was that the biomodel warped after being removed from the SLA machine or that the surgeon failed to remove all of the soft tissues from around the craniotomy defect. Warpage has been reported as a source of error by other investigators. Advantages of prefabricated customized cranioplastic implants

The advantages of the biomodel-generated cranioplasties compared with the traditional techniques used for cranioplastic surgery where autologous bone is not available can be summarized as follows:

  • improved cosmesis and fit;
  • non-toxic thermally polymerized acrylic used;
  • reduced operating time and risk of infection;
  • opportunity for surgeons to assess the defect and implant preoperatively, and improve fixation;
  • improved patient informed consent;
  • allows one-stage resection and implantation procedures to be performed. 3.2.3 Biomodel-guided stereotaxy

Stereotactic (Greek stereo 3D, and tactic to touch) surgery can be defined as the ability accurately to localize brain structures after locating the discrete structures by means of an apparatus co-registered with 3D coordinates. Biomodel-guided stereotaxy uses a physical phantom as the source of the 3D coordinates for the precise localization of intracranial anatomy. This original technique truly reflects the Greek origins of the word stereotactic. As this application of biomodelling is novel, the background of stereotaxy will be discussed before detailing the experiments that were conducted. Development of stereotaxy

Prior to the development of CT, stereotactic techniques were based on head frame systems calibrated to the anterior-posterior commissural (AC-PC) line visualized from intraoperative ventriculography (Anichkov, Polonsky and Usov, 1977). The advent of CT enabled the development of the 'modern art' of stereotaxy. CT digitally displayed anatomy in detail and with only minor geometrical distortion. By combining CT with a localizing frame and fiducials, Brown (Brown, 1979) developed CT stereotaxy. The fiducial system serves as a reference by which any target on a CT/MR scan can be defined in terms of its location with respect to the head frame.

Frame-based stereotaxy is disadvantaged by the traumatic attachment of a head frame, cluttering of the operative field and limited application of evolving 3D imaging techniques. As many neurosurgical procedures do not require pinpoint accuracy, the inconvenience of the frame is often not warranted. The need for a more ergonomic system that easily integrates the anatomical information derived from 3D imaging has spawned the development of frameless stereotaxy. Three or more small stereotactic fiducials are fixed to the patient's scalp prior to volumetric CT or MR. The data are transferred to a operating room workstation. Infrared light projected from a mounted emitter/receiver device is used to localize a wand with reflective fiducials in relation to the patient fiducials. The triangulation of the transmitter positions and the wand tip allows the co-registration of the wand position and the 3D image derived from scan data. This sophisticated apparatus and 3D computer software thus allow the surgeon surgically to navigate with the wand by visualizing its position in relation to the image data.

Sources of error inherent in all frameless stereotaxy systems are related to the quality of image data, the localizing apparatus, registration error (caused by movement of fiducials attached to skin or movement of skin surface landmarks), the deformation of soft tissues during scanning, operator error and movement of the brain relative to the markers or 'brain shift' (caused by craniotomy, tumour resection, CSF aspiration, brain swelling and retraction). Such a system is also expensive and complicated and requires significant training and know-how. Development of biomodel-guided stereotactic surgery

Biomodel-guided stereotactic surgery uses a solid replica of the patient (biomodel) to provide the anatomical data needed for trajectory planning. This concept forms the basis of a method of stereotaxy that can be performed using a number of different apparatus (D'Urso, 1993). The method is as follows:

  1. The patient undergoes CT scanning and a biomodel is manufactured.
  2. An apparatus is attached to definable points on the replica.
  3. The apparatus is used to create an operative plan so that interventional coordinates are able to be saved (e.g. a trajectory is defined to localize a target).
  4. The apparatus is transferred and attached to the patient via common definable points.
  5. The intervention is repeated on the patient using the saved coordinates.

This original method has been validated by the use of apparatus in phantom and patient studies.

Figure 3.1 1 Stereotactic template; 2 trajectory barrel; 3 moulded support; 4 contoured base plate; 5 localizing holes to marker points Biomodel-guided stereotactic surgery with a template and markers

The surgeon draws the resection margin of the tumour on a biomodel of the patient (see Figure 3.4). The resection is performed on the biomodel. A custom-made template is made to fit the operative plan on the biomodel by moulding acrylic to the biomodel so that its outer boundary replicates the surgeon's original marked resection. The resection is rehearsed and the biomodel left with a large defect. The template is used to guide the resection on a mirrored biomodel of the cranium. The 'normal' resected plate is then used as a master implant and is hand finished and used to mould and cast an acrylic custom implant.

Intraoperatively, the surgeon places the template over the skull and contour localizes it. The template is held firmly to the skull and the resection margin is traced from the boundary of the template onto the patient's cranium using diathermy. The tumour is then resected using the traced margin as previously rehearsed on the biomodel. The preoperatively customized acrylic implant is used to close the defect.

The transfer template to perform biomodel-guided stereotaxic surgery may take several forms, each with its own merits:

  1. A simple impression template may be produced. Moulded plastic can be used to trace out a boundary or to guide an intracranial trajectory. Such a template can be fabricated on the biomodel and then be transferred to the patient and aligned by the matching of surface contours alone. This simple device may be difficult to localize if the contour is not sufficiently distinctive. This technique was mentioned previously for the placement of endosseous implants.
  2. Surface markers can be used to locate a template precisely. The template is located to surface contour and at least two markers. The template may be used to trace a boundary for a craniotomy and preoperatively customize an implant to close the defect. One or multiple intracranial trajectories may be incorporated into the template by means of barrels. The template may be made to encase multiple implants by means of a lid with screws. This technique has special application for interstitial brachytherapy (Poulsen et al., 1999).

The accuracy of these experiments was comparable with the previously described stereotactic techniques. Biomodel-guided stereotactic surgery using the D'Urso frame Frame description

The desire to improve the apparatus used for biomodel-guided stereotactic surgery led to the development of a new stereotactic frame (Figure 3.2). A prototype was made using SL and, after design modifications, a stainless steel frame was manufactured. This D'Urso stereotactic frame has been designed to be usable, simple and accurate. The device is ergonomic in the surgical practice of biomodel-guided stereotactic surgery. The reusable frame has three legs which allow exact marker point attachment and a ball-in-socket instrument guide which affords a wide trajectory angle. Adjustable feet easily attach to surface marker pins.

To validate the accuracy of the D'Urso frame, a phantom study was performed. The mean measurement error between the biomodel trajectory and the skull target was 0.89 mm with a 95% confidence interval between 0.86 mm and 0.93 mm (D'Urso et al., 1999b).

The patient is examined by the surgeon and the best position of the frame is determined using standard imaging data or previous biomodels see Figure 3.5. The frame is then held against the patient's head and the scalp where the feet of the frame rest is shaved and infiltrated with local

Figure 3.2 Diagram of the original D'Urso frame: 1 instrument in barrel; 2 trajectory barrel; 3 pivot ball; 4 base plate; 5 lock nut to hold trajectory; 6 tripod base; 7 marker pin; 8 adjustable foot; 9 cranial surface


Was this article helpful?

0 0
Arthritis Relief and Prevention

Arthritis Relief and Prevention

This report may be oh so welcome especially if theres no doctor in the house Take Charge of Your Arthritis Now in less than 5-Minutes the time it takes to make an appointment with your healthcare provider Could you use some help understanding arthritis Maybe a little gentle, bedside manner in your battle for joint pain relief would be great Well, even if you are not sure if arthritis is the issue with you or your friend or loved one.

Get My Free Ebook

Post a comment