- Email: [email protected]
Preoperative planning and intraoperative navigation in skull base surgery
Journal of Cranio-MaxillofacialSurgery (1998) 26, 220-225 © 1998 European Association for Cranio-Maxillofacial Surgery
Preoperative planning and intraoperative navigation in skull base surgery Stefan Hassfeld, I Joachim Z611er,I Friedrich K. Albert, 2 Christian R. Wirtz, 2 Michael Knauth, 3 Joachim Mtihling l
1Department of Oral and Maxillofacial Surgery (Head." Prof. Dr Dr J. Miihling), 2Department of Neurosurgery (Head." Prof. Dr S. Kunze), 3Department of Neuroradiology (Head." Prof. Dr K. Sartor), Ruprecht-Karls University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany SUMMARY. Experience with the commercially available, 3-D navigation systems Viewing Wand (ISG, Mississauga, Ontario, Canada) and SPOCS (Aesculap, Germany) in skull base surgery is presented. Having meanwhile been tested in over 60 clinical trials, the systems achieved an accuracy of _<2.7 mm which, at the moment, we deem sufficiently acceptable to proceed with their clinical evaluation. There was no difference in intraoperative accuracy between the mechanical and the optical navigation systems. The systems proved to be very helpful in identifying the extent of the tumours and in visualizing the proximity of vital structures. 3-D-planning, simulation and intraoperative navigation especially facilitates surgery in anatomically complicated situations, without risk of damaging neighbouring structures. The SPOCS (Surgical Planning and Orientation Computer System) revealed a considerably improved flexibility in handling and a better integration into the surgical procedure in comparison with the relatively inflexible and space-demanding Viewing Wand arm. Especially, the 'offset' function of the SPOCS offers the possibility of a virtual elongation of the instrument and thus, in combination with the on-line visualization of the corresponding images, of a 'look ahead' operation. By using computer-assisted simulation and navigation systems, we can expect quality improvement and risk reduction. More extensive and radical interventions seem possible.
INTRODUCTION
• Reconstructive plastic surgery following accidents and tumour resection.
Until today, the outcome of operations on the anatomically complex region of the skull base depended almost exclusively on the experience of the surgeon. Operation planning is based on the analysis of primarily 2-D and static sources of information such as X-ray images or outprints of CT and MRT data and their comparison with normal values. The employment of milled or stereolithographically-produced models on the basis of CT-scans facilitates spatial orientation (Lambrecht et al., 1995) but, as a static method, it is only of limited use for operation planning and simulation. Also, loss of time and high production costs are disadvantages. Therefore, we aim to support the 3-D planning and simulation of surgical operations by the employment of computers (Altobelli et al., 1995; Girod et al., 1995). Furthermore, the use of intraoperative instrument navigation will enable the exact surgical realization of the planning results (Hassfeld et al., 1995; Wagner et al., 1995; Vannier and Marsh, 1996; Watzinger et al., 1997). Possible fields of application of this computerassisted cranio-maxillofacial surgery are:
Since tumours of the anterior and the adjoining middle fossae with their invasive growth affecting the operational fields of several surgical disciplines and with their close contact with vital anatomical structures present a special challenge for treatment, they will be taken as examples. Operational procedures developed for the treatment of craniofacial malformations are the basis for the treatment of other diseases in this region (Tessier, 1967). A prime example is the temporary removal and subsequent replacement of large frontal and orbital bone segments (Marchac and Renier, 1979; Miihling et al., 1984). Depending on the turnout size and location the standardized removal of suitable bone segments allows a clear access (Zdller et al., 1995). In our interdisciplinary study, which we have been conducting at the Clinic for Head Diseases of the University of Heidelberg since mid 1993, we want to find out whether 'computer-assisted surgery' with its sophisticated methods of 3-D visualization, simulation of surgical procedures and intraoperative navigation can effectively support the surgeon in diagnosis as well as in planning and performing complex skull base surgery.
• The surgical correction of pronounced maxillary malpositions and facial asymmetries in the combined orthodontic, maxillo- and craniofacial treatment of congenital malformations of skull and face. • Surgery of turnouts (definition of resection boundaries) as well as
MATERIAL AND METHODS The navigation system SPOCS (Surgical Planning and Orientation Computer System) (Aesculap AG, 220
Preoperative planning and intraoperativenavigationin skullbase surgery 221
Navigationinstrumentwith infrared diodesfor localization.
Fig. 2 -
Infrared-opticalnavigationsystemSPOCS- Surgical Planning and OrientationComputer System(AesculapAG, Tuttlingen).
Fig. 1 -
Tuttlingen) is based on the localization of diodes emitting an infrared light which is recorded by three linear cameras (Fig. 1). The technical precision of this system is + 0.8 mm. The cameras are attached to a freely movable tripod and can be swung into new positions during the operation according to the surgical situation. To eliminate time-consuming regauging procedures, a patient-tracking function is integrated. This automatically adjusts the patient's position if the operation table or the cameras are moved. Featuring a sufficient number of infrared diodes on the instrument, six degrees of freedom can be spatially detected and computed with the data of the patient's images on a UNIX graphic workstation. The localized instrument carries an automatic identification code and shows as an icon on the monitor. Thus, the surgeon always has a visual feedback as to which instrument is connected. Presently a pointer with three buttons to start software functions is available (Fig. 2). This allows the surgeon to handle the essential functions of the system at the operation table. The instruments will be supplemented by detectable coagulation-tweezers and endoscopes. The patient's head is fixed to the operating table by a Mayfield holder to allow accurate orientation. Subsequently the system is adjusted to the patient's position (Hassfeldet al., 1995, 1996). We prefer to register the position with three titanium mini-screws inserted into the bone under local anaesthesia before tomography, because they give higher precision than marks glued on the skin. The surgeon is now able to observe the actual spatial position and direction of his instruments working on the patient's head in relation to the patient images
of the CT- and/or MR-scans on the screen in 3-D reconstruction and in three reformated planes (axial, sagittal, coronal) of the dataset almost in real-time (Fig. 3). Between July 1993 and June 1997 an interdisciplinary team of cranio-maxillofacial surgeons and neurosurgeons operated on 106 tumours and tumour-like lesions of the anterior skull base and in the adjoining regions of the middle fossa. In 38 cases, the mechanical system was used and in 24 cases the optical navigation system was employed. The histological range of the lesions included, for example, meningioma, chordoma, chondrosarcoma, fibrosarcoma neurinoma, squamous cell carcinoma, adenoidcystic carcinoma and fibrous dysplasia.
RESULTS The application accuracy of the systems (intraoperative comparison with characteristic anatomical bony landmarks or fixed inserted marks) was found to be very high in skull base surgery, with a deviation of < 2.7 mm on average, based on CT-data. Also, during operations lasting several hours, virtually no further deviation occurred. There was no difference in intraoperative accuracy between the mechanical and the optical navigation systems. The navigation systems proved to be very helpful in identifying the extent of the tumours, thus increasing the extent of radical resection especially within the labyrinth of the paranasal sinuses, and in visualizing the proximity of vital structures, e.g. the carotid arteries, brain stem, or cranial nerves, thus decreasing morbidity. Using the MR-scan as the navigation basis, application accuracy values of _<4 mm were achieved in the soft tissue regions surrounding the skull base. In pure soft-tissue surgery, the flexibility of the tissue can render the preoperatively obtained data obsolete. To allow patient movement or to change the position of the camera array during the operation after
222 Journalof Cranio-Maxillofacial Surgery
CT or IVlRT 3D
o f instrument position and corresponding clinical view
- visualisation
Fig. 3 - The principle of intraoperativeinstrument navigation. the patient has been registered, a patient tracking device can be attached to the head fixation. Movements of the OR table as well as the camera boom are thus automatically compensated. Surgeons found the newly realized feature of monitoring certain software functions directly from the instrument very useful. This significantly accelerates the registration process and allows direct interaction with the image-data during the surgical operation for the first time. Especially the 'offset' function, i.e. the virtual elongation of the navigation instrument for a defined distance in combination with the online presentation of the reformated 3-D image data, turned out to be very useful.
CASE P R E S E N T A T I O N A 63-year-old patient suffered from an advanced meningioma of the skull base with intracranial extension and destruction of the central midface. Figure 4 shows the 3-D visualization of the MR-dataset on the screen of the navigation system. The preoperative 3-D reconstruction of the CT-scan (Fig. 5) reveals that the meningioma had destroyed the central mid-face (including the medial walls of the orbit and the nasal bone) and the anterior osseous cranial base. The resection of this monstrous tumour was done with a modified fronto-orbital osteotomy starting with a coronal incision without any further transfacial incisions. The intraoperative navigation system SPOCS
F i g . 4 - MR picture of an advanced meningioma of the cranial base with intracranial extensionand destruction of the central midface.
Preoperative planningand intraoperativenavigationin skull base surgery 223
F i g . 5 - Preoperative 3-D-reconstructionof the CT-data on the computer screen.
Intraoperativeviewof the infrared-opticalnavigation systemSPOCS during tumour resectionvia coronalincisionand subsequent fronto-orbitalosteotomy.
Fig. 6 -
Fig.
7 - Postoperative3-D-reconstructionof the CT-data.
able decrease in the patient's subsequent disability,
( Watanabe et al., 1991; M6sges, 1993; Sandeman et al., described above was applied in planning the access to and during the resection of the tumour in those parts of the cranial base that had been completely altered pathologically as compared with the normal anatomy (Fig. 6). Reconstruction of the cranial base, the medial orbits and the nasal bone was achieved with bilateral transplants of split calvarial bone pediculated to the temporal muscle after a temporary osteotomy of the zygomatic bone. The postoperative 3-D display shows the osseous reconstruction (Fig. 7). The preoperative picture (Fig. 8A) demonstrates the extent to which the tumour had displaced the two globes thus causing their protrusion with hypertelorism and impairment of vision. The postoperative picture (Fig. 8B) 8 weeks after surgery reveals the outward appearance achieved, with undisturbed vision.
DISCUSSION The goal of techniques of computer-assisted surgery is to provide the surgeon with continuously updated information about the location of critical structures in relation to his instruments. The employment of such techniques aims at the reduction of risks in surgery as well as operation time and, consequently, a consider-
1994). The navigation systems Viewing Wand and SPOCS, presented above, provide direct intraoperative support of the surgeon by tools for localization and navigation. The selected surgical procedure can be realized in a spatially accurate way with the help of the intraoperative navigation technique. Tumours of the cranial base that extend intracranially and invade the nose, the paranasal sinuses and/or the orbits require a combined intra- and extracranial procedure. In such cases today, surgical techniques are applied that were initially developed for the treatment of cranio-facial malformations (Marchac et al., 1994). With growing experience, the majority of tumours of the cranial base can be surgically accessed and dissected by fronto-orbital osteotomy or by one of two standardized modifications - the fronto-orbito-nasal and the fronto-orbitozygomatic osteotomies (Z61ler et al., 1995). The advantage of this surgical technique lies in the good extra- and intracranial visualization, also in cases of advanced tumours that can be reached causing only minimal trauma to the brain. Resection of bone and soft tissue in otherwise hardly accessible regions of the skull base is hereby considerably facilitated thus allowing more radical surgery whilst simultaneously reducing the risk.
224 Journalof Cranio-MaxillofacialSurgery
Fig. 8 The preoperativepicture (A) showsthe pronouncedprotrusion of the gIobeswith hypertelorism;the visionwas impaired.The postoperative picture (B) revealsthe outward appearanceachievedeight weeksafter operation; there was no impairmentof vision. Further reduction of surgical risk as well as of postoperative morbidity is to be achieved by the application of computer-assisted surgery techniques. The 3-D visualization of CT- and MR-data supports radiologists and surgeons in the analysis of complex pathological changes (Vannier and Marsh, 1996). Pre and intraoperative simulation procedures help to plan and evaluate diverse surgical methods with computermodels (Altobelli et al., 1993; Girod et al., 1995; Pommert et al., 1996). Subsequently, intraoperative instrument-navigation directly supports the surgeon during the operation. The surgeon can observe on the screen the present spatial position and direction of the instruments led by his hand on the patient's head in relation to the patient's image data. The preoperatively determined procedure can be accurately realized spatially with the help of the intraoperative navigation-technique. The precision of _<2.7 mm (intraoperative comparisons with characteristic anatomical landmarks or inserted marks) achieved in 38 clinical procedures with the mechanical arm and, up to now, in 24 trials of the infrared light-based system on the patient, appears sufficient to us at the moment. The application accuracy precision thus corresponds to the values given in reports on comparable systems (M6sges,
1993; Zinreich et al., 1993; Sandeman et al., 1994; Zamorano et al., 1994; Wagner et al., 1995; Bucholz and Gmco, 1996; Hauser et al., 1997). Though the discussion is still open on the accuracy of surgical navigation systems (Watzinger et al., 1997; Marmulla and Niederdellmann, 1998) it has to be emphasized from a surgical point of view that a system accuracy in the range of 2 mm or better is desirable. However, we found that rigid fixation of the patient is a severe limitation during operation. We would prefer a tracking device, which can be mounted directly onto the patient which will compensate for movements. Still, basic problems stem from differences between the preoperatively created model of the patient and the intraoperative situation. Pathological swellings and alterations, deformations, which, for example, result from the removal of bone segments and cannot be precisely predicted, and, generally, inaccuracies in data registration, processing and calculation add to this problem. These deviations are considerably less in interventions on bony structures or adjacent tissue, as in the case presented of a skull base turnout, than in soft tissue areas. Future intraoperative scanning techniques such as ultrasound and open M R appliances will lead to improvements in this field.
Preoperative planning and intraoperative navigation in skull base surgery 225
The newly realized option with the SPOCS to trigger certain software functions directly with the surgical instrument has turned out to be very useful. The patient registration process is accelerated. For the first time, the surgeon now has the opportunity to interact directly with the patient's dataset during the operation. Especially the 'offset' function, i.e. the virtual elongation of the instrument tip for a defined distance in combination with the online presentation of the corresponding reformated 3-D image data, allows us to gain further information about the surgical site and to estimate possible risks of the next surgical steps more accurately. The computer-assisted surgery techniques presented also allow operations in anatomically difficult situations without endangering neighbouring structures. Operation planning is improved significantly. We expect computer-assisted simulation and navigation systems to improve quality and to reduce risks of surgery. In future, they should also lead to optimizing operating time and thus operation costs. REFERENCES Altobelli, D.E., R. Kikinis, JB. Mulliken, H. Cline, W. Lorensen, F Yolesz: Computer-assisted three- dimensional planning in craniofacial surgery. Plast. Reconstr. Surg. 92 (1993) 576-585. Bucholz, R.D., D.J Greco: Image guided surgical techniques for infections and trauma of the central nervous system. Neurosurg. Clin. N. Am. 7 (1996) 187 200. Gimd, S., E. Keeve, B Ghvd: Advances in interactive craniofacial surgery planning by 3D simulation and visualization. Int. J. Oral Maxillofac. Surg. 24 (1996) 120 125. Hassfeld, S., J. Miihling, J. Zdller: Intraoperative Navigation in Oral and Maxillofacial Surgery. Int. J. Oral Maxillofac. Surg. 24(1995) 111 119. Hassfeld, S., J Zdller, C.R. Wirtz, M, Knauth, ,~ Mfihling: Intraoperative Navigation in Maxillofacial Surgery - Clinical Experiences, Demands and Developments, In: Lemke, Inamura, Jaffe, Vannier (Eds). Proceedings of the International Symposium on Computer and Communication Systems for Image Guided Diagnosis and Therapy CAR '96. Amsterdam, Elsevier, (1996) 739-744. Hauser R., R Westermann, R. Probst." Non-invasive 3D Patient Registration for Image-Guided. Intranasal Surgery Experimental and Clinical Results, In: Troccaz, Grimson, M6sges (Eds). First Joint Conference Computer Vision, Virtual reality and Robotics in Medicine and Medical Robotics and Computer-Assisted Surgery - CVRMedMRCAS '97. Berlin: Springel; (1997) 327-336. Lambrecht J. T., H. Schiel, A. Jacob, Z Kreusch: CAR-CAD-CAMCAS: 3D Perspectives, In: Lemke, Inamura, Jaffe, Vannier (Eds). Proceedings of the International Symposium on Computer and Communication Systems for Image Guided Diagnosis and Therapy CAR '95. Berlin: Springer, (1995) 1364-1368.
Marchac, 12, D. Renier: Le front flottant. Traitement pr6coce des faciocraniostenoses. Ann. Chit. Plast. 24 (1979) 121-126. Marchac, D., D. Renier, S. Broumand: Timing of treatment for craniosynostosis and faeio-craniosynostosis: a 20-year experience. Br. J. Plast. Surg. 47 (1994) 211-222. Marmulla R., H. Niederdellmann: Letter to the Editor. J. Craniomaxillofac. Surg. 26 (1998) 68-69. M6sges; R: Computergestiitzte Chirurgie (CAS) der Schfidelbasisregion. Erg~inzung, Revolution oder Sciencefiction? Eur. Arch. Otorhinolaryngology 250/Suppl I (1993) 373-383. Miihling, J, J. Reuther, N Sdrensen: Operative Behandlung craniofacialer Fehlbildungen. Kinderarzt 15 (1984) 1022-1023. Pommert A., M. Riemer, ~E Schiemann, R. Schubert, U Tiede, K. H. H6hne: Three-dimensional imaging in medicine: methods and applications, In: Taylor, Lavall6e, Burdea, M6sges (Eds). Computer-integrated Surgery. Cambridge: The Mit Press, (1996) 155 174. Sandeman, D.R., N Patel, C. Chandler, R.~ Nelson, H.B. Coakham, H. B Gr~'th: Advances in image-directed neurosurgery: preliminary experience with the ISG Viewing Wand compared with the Leksell G frame. Br. J. Neurosurg. 8 (1994) 529-544. Tessier, P: Osteotomies totales de la face: syndrome de Crouzon, syndrome d'Apert; oxyc6phalies, scaphoc6phalies, turric6phalies. Ann. Chir. Plast. 12 (1967) 273 279. Vannier, M. W.., J.L. Marsh: Three-dimensional imaging, surgical planning, and image-guided therapy. Radiol. Clin. N. Am. 34 (1996) 545-563. Wagner, A., O. Ploder, E. Enislidis, M. Truppe, R. Ewers: Virtual image guided navigation in tumor surgery technical innovation. J. Cranio-Maxillofac. Surg. 23 (1995) 271-273. Watzinger, F., F Wanschitz, A. Wagner, G. Enislidis, FV. Millesi, A. Baumann, R. Ewers: Computer-aided navigation in secondary reconstruction of post-traumatic deformities of the zygoma. J. Cranio-Maxillofac. Surg. 25 (1997) 198-202, Watanabe, E., Y. Mayanagi, Y Kosugi, S. Manaka, K. Takakura: Open surgery assisted by the neuronavigator, a stereotactic, articulated, sensitive arm. Neuro surgery 28 (i 991 ) 792-799. Zamorano, L., Z. Jiang, A.M. Kadi." Computer-assisted neurosurgery system: Wayne State University hardware and software configuration. Comput. Med. Imaging Graph. 18 (1994) 257-271. Zinreich, S.J, S.A. Tebo, D.M. Long et al.: W.M. Frameless stereotaxic integration of CT imaging data: accuracy and initial applications. Radiology 188 (1993) 735 742. Z61ler, J, E Albert, J. Mfihling: Die modifizierte fronto-orbitale Osteotomie als Zugangsweg zur operativen Therapie des Albright-Syndroms. Dtsch. Z. Mund Kiefer Gesichts Chir. 19 (1995) 268-272.
Stefan Hassfeld, MD, DMD, PhD Oral and Maxillofacial Surgery Im Neuenheimer Feld 400 69120 Heidelberg, Germany Tel:+49 6221 567335 Fax:+49 6221 564375 E-mail [email protected]. Paper received 24 March 1998 Accepted 15 June 1998
Preoperative planning and intraoperative navigation in skull base surgery Stefan Hassfeld, I Joachim Z611er,I Friedrich K. Albert, 2 Christian R. Wirtz, 2 Michael Knauth, 3 Joachim Mtihling l
1Department of Oral and Maxillofacial Surgery (Head." Prof. Dr Dr J. Miihling), 2Department of Neurosurgery (Head." Prof. Dr S. Kunze), 3Department of Neuroradiology (Head." Prof. Dr K. Sartor), Ruprecht-Karls University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany SUMMARY. Experience with the commercially available, 3-D navigation systems Viewing Wand (ISG, Mississauga, Ontario, Canada) and SPOCS (Aesculap, Germany) in skull base surgery is presented. Having meanwhile been tested in over 60 clinical trials, the systems achieved an accuracy of _<2.7 mm which, at the moment, we deem sufficiently acceptable to proceed with their clinical evaluation. There was no difference in intraoperative accuracy between the mechanical and the optical navigation systems. The systems proved to be very helpful in identifying the extent of the tumours and in visualizing the proximity of vital structures. 3-D-planning, simulation and intraoperative navigation especially facilitates surgery in anatomically complicated situations, without risk of damaging neighbouring structures. The SPOCS (Surgical Planning and Orientation Computer System) revealed a considerably improved flexibility in handling and a better integration into the surgical procedure in comparison with the relatively inflexible and space-demanding Viewing Wand arm. Especially, the 'offset' function of the SPOCS offers the possibility of a virtual elongation of the instrument and thus, in combination with the on-line visualization of the corresponding images, of a 'look ahead' operation. By using computer-assisted simulation and navigation systems, we can expect quality improvement and risk reduction. More extensive and radical interventions seem possible.
INTRODUCTION
• Reconstructive plastic surgery following accidents and tumour resection.
Until today, the outcome of operations on the anatomically complex region of the skull base depended almost exclusively on the experience of the surgeon. Operation planning is based on the analysis of primarily 2-D and static sources of information such as X-ray images or outprints of CT and MRT data and their comparison with normal values. The employment of milled or stereolithographically-produced models on the basis of CT-scans facilitates spatial orientation (Lambrecht et al., 1995) but, as a static method, it is only of limited use for operation planning and simulation. Also, loss of time and high production costs are disadvantages. Therefore, we aim to support the 3-D planning and simulation of surgical operations by the employment of computers (Altobelli et al., 1995; Girod et al., 1995). Furthermore, the use of intraoperative instrument navigation will enable the exact surgical realization of the planning results (Hassfeld et al., 1995; Wagner et al., 1995; Vannier and Marsh, 1996; Watzinger et al., 1997). Possible fields of application of this computerassisted cranio-maxillofacial surgery are:
Since tumours of the anterior and the adjoining middle fossae with their invasive growth affecting the operational fields of several surgical disciplines and with their close contact with vital anatomical structures present a special challenge for treatment, they will be taken as examples. Operational procedures developed for the treatment of craniofacial malformations are the basis for the treatment of other diseases in this region (Tessier, 1967). A prime example is the temporary removal and subsequent replacement of large frontal and orbital bone segments (Marchac and Renier, 1979; Miihling et al., 1984). Depending on the turnout size and location the standardized removal of suitable bone segments allows a clear access (Zdller et al., 1995). In our interdisciplinary study, which we have been conducting at the Clinic for Head Diseases of the University of Heidelberg since mid 1993, we want to find out whether 'computer-assisted surgery' with its sophisticated methods of 3-D visualization, simulation of surgical procedures and intraoperative navigation can effectively support the surgeon in diagnosis as well as in planning and performing complex skull base surgery.
• The surgical correction of pronounced maxillary malpositions and facial asymmetries in the combined orthodontic, maxillo- and craniofacial treatment of congenital malformations of skull and face. • Surgery of turnouts (definition of resection boundaries) as well as
MATERIAL AND METHODS The navigation system SPOCS (Surgical Planning and Orientation Computer System) (Aesculap AG, 220
Preoperative planning and intraoperativenavigationin skullbase surgery 221
Navigationinstrumentwith infrared diodesfor localization.
Fig. 2 -
Infrared-opticalnavigationsystemSPOCS- Surgical Planning and OrientationComputer System(AesculapAG, Tuttlingen).
Fig. 1 -
Tuttlingen) is based on the localization of diodes emitting an infrared light which is recorded by three linear cameras (Fig. 1). The technical precision of this system is + 0.8 mm. The cameras are attached to a freely movable tripod and can be swung into new positions during the operation according to the surgical situation. To eliminate time-consuming regauging procedures, a patient-tracking function is integrated. This automatically adjusts the patient's position if the operation table or the cameras are moved. Featuring a sufficient number of infrared diodes on the instrument, six degrees of freedom can be spatially detected and computed with the data of the patient's images on a UNIX graphic workstation. The localized instrument carries an automatic identification code and shows as an icon on the monitor. Thus, the surgeon always has a visual feedback as to which instrument is connected. Presently a pointer with three buttons to start software functions is available (Fig. 2). This allows the surgeon to handle the essential functions of the system at the operation table. The instruments will be supplemented by detectable coagulation-tweezers and endoscopes. The patient's head is fixed to the operating table by a Mayfield holder to allow accurate orientation. Subsequently the system is adjusted to the patient's position (Hassfeldet al., 1995, 1996). We prefer to register the position with three titanium mini-screws inserted into the bone under local anaesthesia before tomography, because they give higher precision than marks glued on the skin. The surgeon is now able to observe the actual spatial position and direction of his instruments working on the patient's head in relation to the patient images
of the CT- and/or MR-scans on the screen in 3-D reconstruction and in three reformated planes (axial, sagittal, coronal) of the dataset almost in real-time (Fig. 3). Between July 1993 and June 1997 an interdisciplinary team of cranio-maxillofacial surgeons and neurosurgeons operated on 106 tumours and tumour-like lesions of the anterior skull base and in the adjoining regions of the middle fossa. In 38 cases, the mechanical system was used and in 24 cases the optical navigation system was employed. The histological range of the lesions included, for example, meningioma, chordoma, chondrosarcoma, fibrosarcoma neurinoma, squamous cell carcinoma, adenoidcystic carcinoma and fibrous dysplasia.
RESULTS The application accuracy of the systems (intraoperative comparison with characteristic anatomical bony landmarks or fixed inserted marks) was found to be very high in skull base surgery, with a deviation of < 2.7 mm on average, based on CT-data. Also, during operations lasting several hours, virtually no further deviation occurred. There was no difference in intraoperative accuracy between the mechanical and the optical navigation systems. The navigation systems proved to be very helpful in identifying the extent of the tumours, thus increasing the extent of radical resection especially within the labyrinth of the paranasal sinuses, and in visualizing the proximity of vital structures, e.g. the carotid arteries, brain stem, or cranial nerves, thus decreasing morbidity. Using the MR-scan as the navigation basis, application accuracy values of _<4 mm were achieved in the soft tissue regions surrounding the skull base. In pure soft-tissue surgery, the flexibility of the tissue can render the preoperatively obtained data obsolete. To allow patient movement or to change the position of the camera array during the operation after
222 Journalof Cranio-Maxillofacial Surgery
CT or IVlRT 3D
o f instrument position and corresponding clinical view
- visualisation
Fig. 3 - The principle of intraoperativeinstrument navigation. the patient has been registered, a patient tracking device can be attached to the head fixation. Movements of the OR table as well as the camera boom are thus automatically compensated. Surgeons found the newly realized feature of monitoring certain software functions directly from the instrument very useful. This significantly accelerates the registration process and allows direct interaction with the image-data during the surgical operation for the first time. Especially the 'offset' function, i.e. the virtual elongation of the navigation instrument for a defined distance in combination with the online presentation of the reformated 3-D image data, turned out to be very useful.
CASE P R E S E N T A T I O N A 63-year-old patient suffered from an advanced meningioma of the skull base with intracranial extension and destruction of the central midface. Figure 4 shows the 3-D visualization of the MR-dataset on the screen of the navigation system. The preoperative 3-D reconstruction of the CT-scan (Fig. 5) reveals that the meningioma had destroyed the central mid-face (including the medial walls of the orbit and the nasal bone) and the anterior osseous cranial base. The resection of this monstrous tumour was done with a modified fronto-orbital osteotomy starting with a coronal incision without any further transfacial incisions. The intraoperative navigation system SPOCS
F i g . 4 - MR picture of an advanced meningioma of the cranial base with intracranial extensionand destruction of the central midface.
Preoperative planningand intraoperativenavigationin skull base surgery 223
F i g . 5 - Preoperative 3-D-reconstructionof the CT-data on the computer screen.
Intraoperativeviewof the infrared-opticalnavigation systemSPOCS during tumour resectionvia coronalincisionand subsequent fronto-orbitalosteotomy.
Fig. 6 -
Fig.
7 - Postoperative3-D-reconstructionof the CT-data.
able decrease in the patient's subsequent disability,
( Watanabe et al., 1991; M6sges, 1993; Sandeman et al., described above was applied in planning the access to and during the resection of the tumour in those parts of the cranial base that had been completely altered pathologically as compared with the normal anatomy (Fig. 6). Reconstruction of the cranial base, the medial orbits and the nasal bone was achieved with bilateral transplants of split calvarial bone pediculated to the temporal muscle after a temporary osteotomy of the zygomatic bone. The postoperative 3-D display shows the osseous reconstruction (Fig. 7). The preoperative picture (Fig. 8A) demonstrates the extent to which the tumour had displaced the two globes thus causing their protrusion with hypertelorism and impairment of vision. The postoperative picture (Fig. 8B) 8 weeks after surgery reveals the outward appearance achieved, with undisturbed vision.
DISCUSSION The goal of techniques of computer-assisted surgery is to provide the surgeon with continuously updated information about the location of critical structures in relation to his instruments. The employment of such techniques aims at the reduction of risks in surgery as well as operation time and, consequently, a consider-
1994). The navigation systems Viewing Wand and SPOCS, presented above, provide direct intraoperative support of the surgeon by tools for localization and navigation. The selected surgical procedure can be realized in a spatially accurate way with the help of the intraoperative navigation technique. Tumours of the cranial base that extend intracranially and invade the nose, the paranasal sinuses and/or the orbits require a combined intra- and extracranial procedure. In such cases today, surgical techniques are applied that were initially developed for the treatment of cranio-facial malformations (Marchac et al., 1994). With growing experience, the majority of tumours of the cranial base can be surgically accessed and dissected by fronto-orbital osteotomy or by one of two standardized modifications - the fronto-orbito-nasal and the fronto-orbitozygomatic osteotomies (Z61ler et al., 1995). The advantage of this surgical technique lies in the good extra- and intracranial visualization, also in cases of advanced tumours that can be reached causing only minimal trauma to the brain. Resection of bone and soft tissue in otherwise hardly accessible regions of the skull base is hereby considerably facilitated thus allowing more radical surgery whilst simultaneously reducing the risk.
224 Journalof Cranio-MaxillofacialSurgery
Fig. 8 The preoperativepicture (A) showsthe pronouncedprotrusion of the gIobeswith hypertelorism;the visionwas impaired.The postoperative picture (B) revealsthe outward appearanceachievedeight weeksafter operation; there was no impairmentof vision. Further reduction of surgical risk as well as of postoperative morbidity is to be achieved by the application of computer-assisted surgery techniques. The 3-D visualization of CT- and MR-data supports radiologists and surgeons in the analysis of complex pathological changes (Vannier and Marsh, 1996). Pre and intraoperative simulation procedures help to plan and evaluate diverse surgical methods with computermodels (Altobelli et al., 1993; Girod et al., 1995; Pommert et al., 1996). Subsequently, intraoperative instrument-navigation directly supports the surgeon during the operation. The surgeon can observe on the screen the present spatial position and direction of the instruments led by his hand on the patient's head in relation to the patient's image data. The preoperatively determined procedure can be accurately realized spatially with the help of the intraoperative navigation-technique. The precision of _<2.7 mm (intraoperative comparisons with characteristic anatomical landmarks or inserted marks) achieved in 38 clinical procedures with the mechanical arm and, up to now, in 24 trials of the infrared light-based system on the patient, appears sufficient to us at the moment. The application accuracy precision thus corresponds to the values given in reports on comparable systems (M6sges,
1993; Zinreich et al., 1993; Sandeman et al., 1994; Zamorano et al., 1994; Wagner et al., 1995; Bucholz and Gmco, 1996; Hauser et al., 1997). Though the discussion is still open on the accuracy of surgical navigation systems (Watzinger et al., 1997; Marmulla and Niederdellmann, 1998) it has to be emphasized from a surgical point of view that a system accuracy in the range of 2 mm or better is desirable. However, we found that rigid fixation of the patient is a severe limitation during operation. We would prefer a tracking device, which can be mounted directly onto the patient which will compensate for movements. Still, basic problems stem from differences between the preoperatively created model of the patient and the intraoperative situation. Pathological swellings and alterations, deformations, which, for example, result from the removal of bone segments and cannot be precisely predicted, and, generally, inaccuracies in data registration, processing and calculation add to this problem. These deviations are considerably less in interventions on bony structures or adjacent tissue, as in the case presented of a skull base turnout, than in soft tissue areas. Future intraoperative scanning techniques such as ultrasound and open M R appliances will lead to improvements in this field.
Preoperative planning and intraoperative navigation in skull base surgery 225
The newly realized option with the SPOCS to trigger certain software functions directly with the surgical instrument has turned out to be very useful. The patient registration process is accelerated. For the first time, the surgeon now has the opportunity to interact directly with the patient's dataset during the operation. Especially the 'offset' function, i.e. the virtual elongation of the instrument tip for a defined distance in combination with the online presentation of the corresponding reformated 3-D image data, allows us to gain further information about the surgical site and to estimate possible risks of the next surgical steps more accurately. The computer-assisted surgery techniques presented also allow operations in anatomically difficult situations without endangering neighbouring structures. Operation planning is improved significantly. We expect computer-assisted simulation and navigation systems to improve quality and to reduce risks of surgery. In future, they should also lead to optimizing operating time and thus operation costs. REFERENCES Altobelli, D.E., R. Kikinis, JB. Mulliken, H. Cline, W. Lorensen, F Yolesz: Computer-assisted three- dimensional planning in craniofacial surgery. Plast. Reconstr. Surg. 92 (1993) 576-585. Bucholz, R.D., D.J Greco: Image guided surgical techniques for infections and trauma of the central nervous system. Neurosurg. Clin. N. Am. 7 (1996) 187 200. Gimd, S., E. Keeve, B Ghvd: Advances in interactive craniofacial surgery planning by 3D simulation and visualization. Int. J. Oral Maxillofac. Surg. 24 (1996) 120 125. Hassfeld, S., J. Miihling, J. Zdller: Intraoperative Navigation in Oral and Maxillofacial Surgery. Int. J. Oral Maxillofac. Surg. 24(1995) 111 119. Hassfeld, S., J Zdller, C.R. Wirtz, M, Knauth, ,~ Mfihling: Intraoperative Navigation in Maxillofacial Surgery - Clinical Experiences, Demands and Developments, In: Lemke, Inamura, Jaffe, Vannier (Eds). Proceedings of the International Symposium on Computer and Communication Systems for Image Guided Diagnosis and Therapy CAR '96. Amsterdam, Elsevier, (1996) 739-744. Hauser R., R Westermann, R. Probst." Non-invasive 3D Patient Registration for Image-Guided. Intranasal Surgery Experimental and Clinical Results, In: Troccaz, Grimson, M6sges (Eds). First Joint Conference Computer Vision, Virtual reality and Robotics in Medicine and Medical Robotics and Computer-Assisted Surgery - CVRMedMRCAS '97. Berlin: Springel; (1997) 327-336. Lambrecht J. T., H. Schiel, A. Jacob, Z Kreusch: CAR-CAD-CAMCAS: 3D Perspectives, In: Lemke, Inamura, Jaffe, Vannier (Eds). Proceedings of the International Symposium on Computer and Communication Systems for Image Guided Diagnosis and Therapy CAR '95. Berlin: Springer, (1995) 1364-1368.
Marchac, 12, D. Renier: Le front flottant. Traitement pr6coce des faciocraniostenoses. Ann. Chit. Plast. 24 (1979) 121-126. Marchac, D., D. Renier, S. Broumand: Timing of treatment for craniosynostosis and faeio-craniosynostosis: a 20-year experience. Br. J. Plast. Surg. 47 (1994) 211-222. Marmulla R., H. Niederdellmann: Letter to the Editor. J. Craniomaxillofac. Surg. 26 (1998) 68-69. M6sges; R: Computergestiitzte Chirurgie (CAS) der Schfidelbasisregion. Erg~inzung, Revolution oder Sciencefiction? Eur. Arch. Otorhinolaryngology 250/Suppl I (1993) 373-383. Miihling, J, J. Reuther, N Sdrensen: Operative Behandlung craniofacialer Fehlbildungen. Kinderarzt 15 (1984) 1022-1023. Pommert A., M. Riemer, ~E Schiemann, R. Schubert, U Tiede, K. H. H6hne: Three-dimensional imaging in medicine: methods and applications, In: Taylor, Lavall6e, Burdea, M6sges (Eds). Computer-integrated Surgery. Cambridge: The Mit Press, (1996) 155 174. Sandeman, D.R., N Patel, C. Chandler, R.~ Nelson, H.B. Coakham, H. B Gr~'th: Advances in image-directed neurosurgery: preliminary experience with the ISG Viewing Wand compared with the Leksell G frame. Br. J. Neurosurg. 8 (1994) 529-544. Tessier, P: Osteotomies totales de la face: syndrome de Crouzon, syndrome d'Apert; oxyc6phalies, scaphoc6phalies, turric6phalies. Ann. Chir. Plast. 12 (1967) 273 279. Vannier, M. W.., J.L. Marsh: Three-dimensional imaging, surgical planning, and image-guided therapy. Radiol. Clin. N. Am. 34 (1996) 545-563. Wagner, A., O. Ploder, E. Enislidis, M. Truppe, R. Ewers: Virtual image guided navigation in tumor surgery technical innovation. J. Cranio-Maxillofac. Surg. 23 (1995) 271-273. Watzinger, F., F Wanschitz, A. Wagner, G. Enislidis, FV. Millesi, A. Baumann, R. Ewers: Computer-aided navigation in secondary reconstruction of post-traumatic deformities of the zygoma. J. Cranio-Maxillofac. Surg. 25 (1997) 198-202, Watanabe, E., Y. Mayanagi, Y Kosugi, S. Manaka, K. Takakura: Open surgery assisted by the neuronavigator, a stereotactic, articulated, sensitive arm. Neuro surgery 28 (i 991 ) 792-799. Zamorano, L., Z. Jiang, A.M. Kadi." Computer-assisted neurosurgery system: Wayne State University hardware and software configuration. Comput. Med. Imaging Graph. 18 (1994) 257-271. Zinreich, S.J, S.A. Tebo, D.M. Long et al.: W.M. Frameless stereotaxic integration of CT imaging data: accuracy and initial applications. Radiology 188 (1993) 735 742. Z61ler, J, E Albert, J. Mfihling: Die modifizierte fronto-orbitale Osteotomie als Zugangsweg zur operativen Therapie des Albright-Syndroms. Dtsch. Z. Mund Kiefer Gesichts Chir. 19 (1995) 268-272.
Stefan Hassfeld, MD, DMD, PhD Oral and Maxillofacial Surgery Im Neuenheimer Feld 400 69120 Heidelberg, Germany Tel:+49 6221 567335 Fax:+49 6221 564375 E-mail [email protected]. Paper received 24 March 1998 Accepted 15 June 1998