Neurosurg Focus 6 (3):Article 6, 1999 A mobile high-field magnetic resonance imaging system for neurosurgery

Garnette R. Sutherland, M.D., F.R.C.S.(C), Taro Kaibara, M.D., Deon Louw, M.D., F.R.C.S.(C), and John Saunders, Ph.D. Department of Clinical Neurosciences, Division of Neurosurgery, The University of Calgary, Calgary, Alberta, Canada

The authors' goal was to place a mobile, 1.5 tesla magnetic resonance (MR) imaging system into a neurosurgical operating room without adversely affecting established neurosurgical management. The system would help to plan accurate surgical corridors, confirm the accomplishment of operative objectives, and detect acute complications such as hemorrhage or ischemia. The authors used an actively shielded 1.5 tesla magnet, together with 15 m tesla/m gradients, MR console computers, gradient amplifiers, a titanium, hydraulic-controlled operating table, and a radio frequency coil that can be disassembled. The magnet is moved to and from the surgical field by using overhead crane technology. To date, the system has provided unfettered access to 46 neurosurgical patients.

In all patients, high-definition T1- and/or T2-weighted images were rapidly and reproducibly acquired at various stages of the surgical procedures. Eleven patients underwent craniotomy that was optimized after pre-incisional imaging. In four patients who harbored subtotally resected tumor, intraoperative MR imaging allowed removal of remaining tumor. Interestingly, the intraoperative administration of gadolinium in the management of patients with malignant glioma demonstrated a dynamic expansion of enhancement beyond the preoperative contrast contour. These zones of new enhancement proved, on examination of biopsy sample, to be tumor. The authors have demonstrated that high-quality MR images can be obtained within reasonable time constraints in the operating room. Procedures can be conducted without compromising or altering traditional neurosurgical, nursing, or anesthetic techniques. It is feasible that within the next decade intraoperative MR imaging may become the standard of care in neurosurgery. Key Words * magnetic resonance imaging * neurosurgery * intraoperative imaging * radio frequency coil

Lars Leksell advocated the philosophy that no technique in neurosurgery could be too refined,

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC particularly in reference to the ability to localize lesions.[25] In fact, it is a truism to state that the entire progression of neurosurgery has paralleled, and depended on, increasingly sophisticated neuroimaging. Random trephinations of Incan patients with epilepsy likely met with modest success, and it is only in the 19th century that Paul Broca[7] and others challenged phrenologic maxims with concepts of cortical compartmentation of function. This spawned a symbiotic relationship of neurologist and neurosurgeon, the former often guiding the latter to a presumptive operative target. Intraoperative imaging was revolutionized by the pioneering efforts of Walter Dandy[10] in pneumoencephalography. In 1927, the neurologist Egas Moniz and the neurosurgeon Almeida Lima performed cerebral angiography,[11] ushering in a new era in the operative management of patients with cerebrovascular disease.[5,12,33] Intraoperative ultrasound, both cranial and spinal, has also been successfully applied to lesion localization.[8,13,42] However, it was the invention of computerized tomography (CT) scanning and magnetic resonance (MR) imaging that heralded the modern epoch of neurosurgery. The use of these technologies has avoided the side effects and complications of pneumoencephalography and provided exquisite images of parenchymal pathology. It was not long before the CT scanner was entrained into the operating room, as reported by Lunsford and associates.[26,29,30] Lunsford and coworkers[28] have suggested that CT scanning would probably remain the preeminent modality for conventional intraoperative image-guided stereotactic surgery, because of the enhanced access it provides to patient and its enhanced instrument compatibility. However, the use of MR imaging promised radiation-free, multiplanar images with superior spatial resolution. Intraoperative MR imaging would thus allow accurate and conservative craniotomy placement, as well as planning of safer surgical corridors. Maximum resection of tumor could be confirmed, and complications such as hemorrhage or ischemia could be detected in a timely fashion. Moreover, its use could update the archived images that plague the currently popular frameless stereotactic systems, which are unable to compensate for varying degrees of intraoperative brain shift and distortion.[14,36] Investigators, in collaboration with General Electric Medical Systems (Milwaukee, WI) took up the challenge. They developed a 0.5 tesla, vertical, biplanar Double Doughnut configuration that has been used to obtain stereotactic biopsy samples and to perform craniotomy, cyst drainage, and laminectomy.[2,6] Although the system is seminal in concept, it lacks the field strength needed for higher quality images, functional MR imaging, and MR spectroscopy. Surgical access is limited, and the immobility of the system necessitates nonferromagnetic instrumentation, including the microscope and drill. Siemens AG (Erlangen, Germany) produced a 0.2 tesla, horizontal, biplanar intraoperative MR imaging system, the Magnetom Open, which is positioned adjacent to the operating theater.[37] This physical separation of surgery and imaging was devised to solve the problems incurred by the General Electric Double Doughnut system, because it allows standard surgical technique along with unfettered access. Instrument compatibility is less relevant, resulting in cost savings. However, the field-strength is only 0.2 tesla, and patient transit is time consuming. It is with these concerns in mind that we developed a movable, high-field 1.5 tesla actively shielded magnet. The system allows rapid acquisition of high-quality images without the need for potentially hazardous mobilization of the patient. Surgical access to the neurosurgical patient remains unhindered. CLINICAL MATERIALS AND METHODS Patient Population Table 1 provides a summary of characteristics of the first 46 patients to undergo surgery in which our

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC mobile high-field MR system was used. The majority of patients harbored glioma, meningioma, or vascular disease. The remaining patients harbored intracranial and skull base lesions.

Imaging Magnet Doors separate the magnet from the remainder of the operating room when not in use (Fig. 1). The magnet, a superconducting, 1.5 tesla is now commercially available (Magnex Scientific, Abingdon, Oxon, England), is designed to accept acceleration and deceleration stresses when moving. To decrease fringe fields it has been actively shielded. This results in a 5-Gauss line with a radial dimension of only 2.9 m, negating the need for reinforced steel in the operating room walls. The magnet subtends an inner cylindrical bore of a diameter of 80 cm. The installation of gradient inserts and radiofrequency (RF) shield reduces the usable diameter to 62.1 cm. The self-shielded gradient coils generate 15 mT/m at 250 A with a 500-microsecond rise time. Setup time is minimized with patented automatic tuning and matching of the RF coil.[20,21] The magnet has a length of 1.5 meters, which is short enough to allow it to be positioned so that the part of the patient to be imaged is at the center of the bore both longitudinally and vertically. It is mobilized longitudinally on a ceiling support track by an electric motor.[19]

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC Fig. 1. Diagram illustrating the neurosurgical operating room showing the magnet in storage (A) or imaging positions (B). The 5 (yellow) and 50 (red) gauss lines are indicated, together with the observation room, where the images are first displayed. Radiofrequency Coil The radiofrequency (RF) coil was designed to provide a high signal-to-noise ratio while maintaining complete access to the patient. The design was based on four free-standing tuned rings on the surface of a sphere that are inductively coupled.[18] This ensemble resonates at the Larmor frequency and gives an RF field that is homogeneous. The modified design has four rings: one at 13 cm and another at 17 cm

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC above the patient's head, and two coils placed symmetrically below the head. Each ring is tuned with four capacitors placed 90° apart to cancel electrical field effects at the center. Fine tuning is accomplished by adjusting the angle of tilt of a fifth ring placed below the head. The coil is matched to 50 ohms by the vertical movement of the fifth ring, which is the only ring to have any electrical connections. One of the benefits of this type of design is that the bottom rings can be incorporated into the operating room table and the upper two rings can easily be removed to allow complete access to the patient's head (Fig 2). A new technology to construct the rings was developed to eliminate any image distortion due to magnetic susceptibility. The method involves electroplating copper onto 2.23-cm fiberglass resin rings that have been coated with conductive paint.

Fig. 2. Photograph showing placement of the upper half of the RF coil onto the bottom half, which is incorporated into the operating room table. The ability to disassemble the RF coil provides free surgical access to the patient between MR examinations. Most commercial MR imaging systems provide for automatic tuning of the RF coil, but the matching is fixed. Because surgery would have significant impact on coil loading, both automatic tuning and matching were developed. The major obstacle to automatic tuning and matching is the interaction of the two mechanisms. The use of a doubly resonant, inductively coupled coil overcomes this problem because the tuning and matching are then orthogonal for small perturbations. When such a coil is equipped with a negative feedback control of tuning and matching elements, stable and rapid automatic tuning and matching are accomplished.[20] Magnetic Resonance Imaging Console Standard Surrey Medical Imaging Systems, Limited, Guildford, Surrey, U.K., hardware has been incorporated into the intraoperative MR imaging system. The software is being rationalized, and all computers are being converted to use Microsoft NT. The console computers, together with gradient amplifiers, are located in a room below the operating theater. A workstation remotely connected to the MR console is located in a viewing room adjacent to the operating theater. Acquired images are displayed on liquid crystal monitors located both in the viewing room and operating theater.

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC Operating Room Table An important element in the success of a neurosurgical procedure is a movable operating room table that has many degrees of freedom. With the placement of an MR imaging system into the operating room, it is crucial that the table be completely MR compatible. Our table was designed and constructed from titanium and fiberglass and is sufficiently cantilevered as to allow the patient's head to be at the center of the magnet during imaging. The various movements of the table are driven hydraulically rather than by electrical motors at various times during the operation, which is necessitated by the presence of the magnetic field. The use of hydraulics provides smooth movement through all degrees of freedom. Operating Room Ceiling support track beams have been attached to the building support walls. These beams are able to support the magnet, which has a weight of four tons. The magnet is moved into and out of the operative field by using a small electrical motor (Fig. 3). The operating theater is a standard sized 7.58 X 10.4 m room that includes an alcove of 2.4 X 3.8 m which houses the magnet when not in use. For safety, the 50 and 5-Gauss lines for both the docked and imaging magnet positions are clearly marked on the floor.

Fig. 3. Photograph showing the ceiling-mounted 1.5 tesla magnet in its imaging position. In conventional MR imaging suites, traditionally located in radiology departments, the magnet is placed in a shielded room to stop RF noise from interfering with image quality. To attenuate atmospheric RF sufficiently, heavy metal doors with high electrical contact with the door frame and, hence the rest of the room, are used. Because this would be difficult to achieve in an operating room, local RF shielding methods were developed. A tube, the outer surface of which has been sprayed with copper paint, was placed inside the gradient coil tube and connected at the non-patient end of the magnet to a copper mesh. This not only electrically closes this end of the magnet but also acts as a shield for the RF coil. When the patient has been positioned for imaging, a silver-impregnated RF tent is placed around the operating table, over the patient, and attached to the magnet, thereby establishing electrical contact between the tent and copper paint. All patient-monitoring equipment, as well as the RF transmission and receiving signals, pass through a filter box, which in turn is connected to the RF tent.

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC Imaging Prior to an MR imaging study a final safety examination is performed by the operating room charge nurse. All MR-incompatible equipment such as the operating microscope, the ultrasonic homogenizer, the high-speed air drill, and bipolar cautery are moved from the operative field to beyond the 5-Gauss line. To obtain images during the surgical procedure, sterile drapes are placed over the operative field and the patient. The upper half of the RF coil is placed over the head and secured to its bottom half. The magnet bay doors are opened, and the magnet is moved over the patient for imaging. At the completion of the study the magnet is taken back to its alcove, the overdrapes removed, all instrumentation returned to the surgical field, and surgery resumed. Preparation for imaging, including safety inspection, draping, and positioning of the magnet requires approximately 5 minutes. Axial, coronal, or sagittal T1-weighted 5-mm image sequences are acquired in 4.5 minutes. Currently, T2-weighted image acquisition time is 10 minutes. After the imaging procedures, the time required to remove the magnet, drapes, and to resume surgery is approximately 2 minutes. Oil-based vitamin E capsules are placed in the proximity of a putative scalp incision to act as fiducial markers. The capsules have a characteristic, bright signal on T1-weighted images. They may also be used intracranially for interdissection imaging by covering them with sterile plastic, because they will not withstand the autoclaving process. RESULTS Patient Characteristics and Safety Table 1 presents the first 46 patients by age, gender, and pathological diagnosis. The majority of patients harbored glioma, meningioma, or vascular pathology. The remainder comprised a heterogeneity of intracranial and skull base lesions. In total, 158 MR imaging studies were performed on the 46 patients (Table 2).

The studies included 67 T1-, 62 T1-(after gadolimium administration), and 29 T2 -weighted sequences. In all studies, the signal-to-noise ratio was sufficient to provide high-quality images that enhanced the surgical procedure. Surgical-planning intraoperative imaging was performed prior to surgical preparation and draping, interdissection images at various stages of the operative procedure, and quality assurance imaging following wound closure. There have been no adverse events related to the high magnetic field that have compromised the safety of the patient or operating room personnel. Two separate minor events occurred during positioning of the magnet: a ballpoint pen was extracted from a shirt pocket and a

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC suction stillette, left in the surgical field, was entrained into the magnet. As a result, peri-imaging instrument counts have been instituted and all visitors to the operating room are screened by the charge nurse and required to watch an MR safety video. Patient body weight was restricted to 115 kg because larger patients failed to gain safe entry into the magnet bore. Similarly, constraints were placed on positioning patients in the lateral or prone positions. A new, larger bore magnet (see Discussion) has been developed to address appropriately these limitations. Meningioma Intraoperative MR imaging provided quality assurance in all 12 patients (Fig. 4). In one patient, the intraoperative imaging study revealed unsuspected tumor that was then removed prior to wound closure. Inter-dissection or quality-assurance T1-weighted gadolinium-enhanced studies did not demonstrate brain enhancement even when there was surgical disruption of the pia-arachnoid. Progressive brain shift was evident in 10 cases­mild in three, moderate in six, and marked in one. This reflects the removal of relatively large tumors, and the loss of cerebrospinal fluid secondary to dissection within the subarachnoid space and/or basal cisterns.

Fig. 4. Surgical planning (upper row) and quality assurance (lower row) MR images confirming gross-total extirpation of meningioma. The images were acquired in 4.5 minutes and had a signal-to-noise ratio of 13.5. Glioma Of the eight low-grade gliomas, six were oligodendroglioma and two were astrocytoma. Seven of the eight high-grade neoplasms were glioblastoma multiforme, and one was anaplastic astrocytoma. Inter-dissection and quality assurance studies clearly demonstrated the extent of surgical resection and the degree of brain shift associated with tissue removal and cerebrospinal fluid loss (Figs. 5 and 6). Unique to those patients with either benign or malignant glioma was the observation of progressive

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC gadolinium enhancement beyond the confines of preoperative and surgical planning studies.

Fig. 5. Surgical planning (upper row) and quality assurance (lower row) axial T1-wieghted MR images with gadolium enhancement obtained in a patient with glioblastoma multiforme. Note the degree of cerebral shift following resection of the tumor and the progressively enhancing border. The images were acquired in 4.5 minutes and had a signal-to-noise ratio of 15.3.

Fig. 6. Surgical planning (upper row) and interdissection (lower row) axial T2-weighted MR images obtained in a patient with oligodendroglioma. The interdissection images

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC demonstrate the extent of residual neoplasm. Each image set was acquired in 10 minutes and had a signal-to-noise ratio of 14. Vascular Lesions The system has proven particularly useful in the surgical planning in the two patients with cavernous angioma. The small volume and frequent subcortical location of the lesions made them excellent candidates for the benefits of surgical planning images (Fig. 7). In both cases the cavernoma was readily identified and resected. In the three patients with arteriovenous malformations (AVMs), the quality assurance studies excluded the presence of intraventricular or parenchymal hemorrhage prior to the patient leaving the operating room. Because MR angiography software is in development, it was not available for these three patients.

Fig. 7. Surgical planning axial (upper row) and coronal (lower row) T1-weighted MR images revealing the relationship of the vitamin E capsule to the underlying cavernoma. Each set of images was acquired in 4.5 minutes and had a signal-to-noise ratio of 15. Miscellaneous Lesions Various other pathologies were encountered. During the resection of a small choroid plexus carcinoma related to the ependyma, the frozen section of the suspected tumor tissue showed only gliosis. Interdissection imaging was able to define well the location of the lesion and its relationship to the site at which the biopsy sample was obtained. The tumor was subsequently removed. Interdissection imaging of the patient who harbored an olfactory nerve schwannoma revealed unsuspected tumor that was readily resected (Fig. 8 upper and lower). The patient with pachymeningitis presented with headache that was associated with an extraaxial middle cranial fossa lesion. Preoperative imaging was suggestive of meningioma.

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC Fig. 8. Imaging studies obtained in a patient with olfactory nerve schwannoma. Upper: Surgical planning axial T1-weighted MR images without (upper row) and with (lower row) gadolium enhancement. Each image set was acquired in 4.5 minutes and had a signal-to-noise ratio of 16.1. Lower: Interdissection axial T1-weighted MR images without (upper row) and with (lower row) gadolium enhancement. Unsuspected, residual schwannoma is evident (arrow). Anterior to the enhancing tissue is prolapsed periorbital fat. Also, note the reflected scalp anterior to the orbit. The quality assurance studies demonstrated complete resection of all miscellaneous lesions except for residual dural enhancement in the patient with pachymeningitis. DISCUSSION The results of this study demonstrate that high-quality MR images can be obtained in a reasonable length of time in the operating room. Furthermore, these studies can be accomplished without compromising or altering traditional neurosurgical, nursing, or anesthetic techniques. We are also able to perform electrophysiological studies, such as electrocorticography, which are not currently feasible when using existing intraoperative MR technology. To date, our experience with this intraoperative MR imaging system has focused largely on tumor

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC surgery, with resection of benign and malignant lesions in 40 cases. Five vascular malformations (three cavernomas; two AVMs) have been surgically treated, with complete removal in all cases. It was believed that operative management was significantly altered in 11 patients, resulting in minimalist surgery but maximizing technical outcomes. Placement of craniotomy was altered in many patients by up to 2 cm following application of vitamin E capsules on the scalp prior to surgical-planning imaging studies. Craniotomies also tended to be smaller than usual, which is the case with frameless stereotaxy.[4,16,31,41] Unlike the latter studies, however, we were able to update our surgical navigation ad-lib. This proved to be of particular help in accounting for the brain shifts associated with positioning and dural opening. Of occasional importance was the observation of significant tumor growth in the weeks following diagnosis and scheduling of surgery; it is not unusual in the Canadian healthcare system for patients with a "stable" brain tumor to have their surgery delayed for a significant period of time. A significant number of patients were noted to have residual tumor on intraoperative imaging, resulting in reexploration to complete the resection where feasible. It is possible that aggressive cytoreductive surgery in malignant tumors improves patient survival,[9,32,43] but our sample size and short follow-up period prevent any formal interpretation at this stage. However, total removal of benign lesions may prevent the need for reoperation or adjuvant radiosurgery. It is also very reassuring to be able to perform MR imaging that excludes the presence of hematoma in the operative bed or beyond prior to closure, particularly in the case of those patients with AVM. Intraoperative assessment of the results of malignant glioma resection margins offers[39] useful information regarding residual tumor volume and also provides intriguing information about the temporal profile of their contrast-enhancement characteristics. It is our experience that aggressive, extracapsular resection of meningioma does not result in peritumoral, parenchymal penetration of gadolinium, despite pial disruption. However, it is evident, intraoperatively, that a crescent of gadolinium advances beyond the preoperatively defined contrast limits of malignant glioma. Biopsy samples of this newly enhancing zone have thus far been histologically tumor positive, despite gross-total resections conforming to the preoperative tumor contour. It is possible that tumor manipulation provokes release of angiogenically active molecules unique to high-grade astrocytoma and/or that tumor-infiltrated parenchyma possesses uniquely susceptible blood-brain barrier characteristics. Investigators have differentiated between immediate and delayed enhancement at the resection border, and they hypothesize that the former is a consequence of cautery.[39] Closer characterization of this phenomenon will shed more light on the veritable industry of research regarding postoperative MR imaging for neurooncology. In one study it was noted that postsurgical enhancement, which was at its maximum intensity from postoperative days 5 to 14 in 53% of patients, could potentially masquerade as tumor.[15] Additionally, methemoglobin corrupted assessment of residual tumor in 35% of cases.[15] This image interference could be minimized were intraoperative MR imaging used. Our experience also allows us to refute the assertion that gadolinium-enhanced MR imaging during postoperative Days 1 to 3 "avoids surgery-induced contrast enhancement and minimized interpretative difficulties."[1] Of great interest is the report in which 64% of epilepsy patients developed enhancement of nonneoplastic resection bed within 17 hours of surgery.[17] Perhaps we should exploit the intraoperative enlargement of enhancement in malignant gliomas by administering chemotherapeutic agents during surgery.[24] An alternative application in the clinical realm includes the use of MR angiography in assessing vessel resection in AVM surgery. The use of MR angiography at completion of carotid endarterectomy could evaluate vessel patency, and diffusion-weighted imaging could be used to assess for any ischemic changes. Multinuclear MR spectroscopy could also be used in the assessment of cerebral

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC ischemia[22,38,40] and central nervous system neoplasms.[3,23,34,35] Indeed, it may be possible with proton MR spectroscopy to differentiate normal from neoplastic tissue. The technical specifications of our system are such that they address many of the concerns published by various investigators on intraoperative imaging. Steinmeier, et al.,[37] have noted that a minimum of 1 hour can be added to a case if the Magnetom Open imager is used, and in the case of pituitary adenoma, T1-weighted and turbo spin­echo image sequences alone required 30 minutes in total acquisition time. In our series the time acquisition of T1- weighted images (with gadolinium) is 4.5 minutes. Total time after the scalp is covered until surgery recommences averages, in our experience, 27 minutes. However, calibration comprises approximately 18 minutes of total elapsed time. New software that is being introduced in February 1999 will dramatically reduce calibration time to approximately 1 minute. Currently, T2-weighted imaging requires 10 minutes. However, new fast spin­echo pulse sequences will decrease the time to 4 minutes. Additionally, fluid-attenuated inversion recovery sequences will minimize the bright water signals of irrigation saline. Although we are able to move the magnet into the operative field within 90 seconds, its 62-cm bore does not allow facile placement of patients in the prone or lateral position. We have therefore constructed a 92-cm bore (72 cm with gradient inserts) magnet, which is only 1.3 m in length. The opening has also been bevelled to accommodate larger patients and to maximize safety access. Gradient field strengths have been increased from 15 to 25 mT/m, which allows the use of echo planar imaging. These changes permit the operating room to have the full range of modern MR imaging capabilities. Macunias has written, "Less convincing are the speculations by the authors regarding intraoperative functional MR imaging or thermosensitive imaging of lesions.[37] These applications require stronger magnetic fields than are presently available." Our full-strength magnet has a signal/noise ratio fully capable of these functions, which shall be introduced in 1999. At 1.5 tesla, MR angiography will also be of a superior quality and may be useful in vascular surgery. Although the resolution would obviously not reveal thalamostriate or basilar artery perforators, diffusion-weighted imaging could rapidly reveal brain injury secondary to their occlusion and allow repositioning of the aneurysm clip prior to wound closure. The unique two-part RF coil is highly independent of loading. It is relatively large and can be placed over the entire head and surgical field, thereby providing images that are of high quality and readily reproducible. The coil is being modified to accommodate different head positions, and it will be able to move in a vertical plane. In addition, a three-pin head holder has been developed that will be placed in the bottom half of the RF coil. The current operating room table is versatile but is hydraulic based, which requires a fairly high level of maintenance. We have thus designed, and are in the process of constructing, an operating room table that is moved by small electric motors, much like that used in conventional tables. We are able to report that stringent but simple strategies of safety have prevented any adverse MR imaging events. All surgical personnel, including maintenance and cleaning, are required to study a safety video. Wooden doors are closed around the magnet when it is not in use, encouraging free operating room access during the operative part of the procedure. All instruments that must remain in the operative field at the time of imaging are MR imaging compatible, while others are merely moved beyond the 5-Gauss line. To date, no patient has experienced any cardiorespiratory compromise during imaging. Routine anesthetic techniques with invasive monitoring are performed, and the semitransparent RF tent allows visualization of the patient.

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC In their report, Steinmeier, et al.,[37] conclude that: "the time needed for intraoperative neuronavigation-image updating limits the procedure to selected cases." Lunsford[27] wonders, "Why should we compromise the surgery to obtain the imaging?" Our contention, however, is that the highfield, mobile system we are using offers substantive benefits to neurosurgery. It is quite possible that within the next decade increasingly sophisticated intraoperative MR imaging will become the standard of care for which surgeons may strive. Acknowledgments The authors thank Innovative Magnetic Resonance Imaging Systems Inc. of Canada, Magnex Scientific Ltd., of the UK, Surrey Magnetic Imaging Systems Ltd., of the UK, the National Research Council of Canada, Institute for Biodiagnostics, and the neurosurgical staff and operating room personnel at the University of Calgary, Calgary, Alberta, Canada, for their help in developing and implementing this intraoperative MR system.

References 1. Albert FK, Forsting M, Sartor K, et al: Early postoperative magnetic resonance imaging after resection of malignant glioma: objective evaluation of residual tumor and its influence on regrowth and prognosis. Neurosurgery 34:45­61 1994 2. Alexander E III, Moriarty TM, Kikinis R, et al: The present and future role of intraoperative MRI in neurosurgical procedures. Stereotact Funct Neurosurg 68:10­17, 1997 3. Arnold DL, Emrich JF, Shoubridge EA, et al: Characterization of astrocytomas, meningiomas, and pituitary adenomas by phosphorus magnetic resonance spectroscopy. J Neurosurg 74:447­453, 1991 4. Barnett GH, Kormos DW, Steiner CP, et al: Intraoperative localization using an armless, frameless stereotactic wand. Technical note. J Neurosurg 78:510­514, 1993 5. Barrow DL, Boyer KL, Joseph GJ: Intraoperative angiography in the management of neurovascular disorders. Neurosurgery 30:153­159, 1992 6. Black PM, Moriarty T, Alexander E III, et al: Development and implementation of intraoperative magnetic resonance imaging and its neurosurgical applications. Neurosurgery 41:831­845, 1997 7. Broca P: Nouvelle observation d'aphémie produite par une lesion de la moitié postérieure des deuxième et troisième circonvolutions frontales gauches. Bull Soc Anat Paris 6:398­407, 1861 8. Chandler WF, Knake JE, McGillicuddy JE, et al: Intraoperative use of real-time ultrasonography in neurosurgery. J Neurosurg 57:157­163, 1982 9. Ciric I, Ammirati M, Vick N, et al: Supratentorial gliomas: surgical considerations and immediate postoperative results. Gross total resection versus partial resection. Neurosurgery 21:21­26, 1987 10. Dandy WE: Roentgenography of the brain after the injection of air into the spinal canal. Ann Surg 70:397­403, 1919 11. Darius J: Radiography of the living brain. Nature 308:225, 1984 (Letter) 12. Derdeyn CP, Moran CJ, Cross DT, et al: Intraoperative digital subtraction angiography: a review of

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC 112 consecutive examinations. AJNR 16:307­318, 1995 13. Dohrmann GJ, Rubin JM: Dynamic intraoperative imaging and instrumentation of brain and spinal cord using ultrasound. Neurol Clin 3:425­437, 1985 14. Dorward NL, Alberti O, Velani B, et al: Postimaging brain distortion: magnitude, correlates, and impact on neuronavigation. J Neurosurg 88:656­662, 1998 15. Forsyth PA, Petrov E, Mahallati H, et al: Prospective study of postoperative magnetic resonance imaging in patients with malignant gliomas. J Clin Oncol 15:2076­2081, 1997 16. Golfinos JG, Fitzpatrick BC, Smith LR, et al: Clinical use of a frameless stereotactic arm: results of 325 cases. J Neurosurg 83:197­205, 1995 17. Henegar MM, Moran CJ, Silbergeld DL: Early postoperative magnetic resonance imaging following nonneoplastic cortical resection. J Neurosurg 84:174­179, 1996 18. Hoult DI, Deslauriers R: A high-sensitivity, high-B1 homogeneity probe for quantitation of metabolites. Magn Reson Med 16:411­417, 1990 19. Hoult DI, Saunders JK, Sutherland GR, Roberts FA: Surgical procedure with magnetic resonance imaging. U.S. PATENT # 5,735,278, 1998 20. Hwang F, Hoult DI: Automatic probe tuning and matching. Magn Reson Med 39:214­222, 1998 21. Hwang F, Hoult DI: Tuning and matching an impedance (automatic matching and tuning RF Coil for the intraoperative). PATENT PENDING: International Patent Application No. PCT/CA98/00345. U.S. Patent Application No. 835900, 1999 22. Kaibara T, Sutherland GR, Colbourne F, et al: Hypothermia: depression of tricarboxylic acid cycle flux and evidence for pentose phosphate shunt upregulation. J Neurosurg 90:339­347, 1999 23. Kaibara T, Tyson RL, Sutherland GR: Human cerebral neoplasms studied using MR spectroscopy: a review. Biochem Cell Biol 76:477­486, 1998 24. Knauth M, Wirtz CR, Tronnier VM, et al: [Intraoperative magnetic resonance tomography for control of extent of neurosurgical operations.] Radiologe 38:218­224, 1998 (Ger) 25. Lundsford L, Kondziolka D, Leksell D: The Leksell stereotactic system, in Tasker RR, Gildenberg PL (eds): Functional and Stereotactic Neurosurgery. New York: McGraw-Hill, 1998, pp 51­64 26. Lunsford LD: A dedicated CT system for the stereotactic operating room. Appl Neurophysiol 45:374­378, 1982 27. Lunsford LD: Comment on Steinmeier R, Fahlbusch R, Ganslandt O, et al: Intraoperative magnetic resonance imaging with the magnetom open scanner: concepts, neurosurgical indications, and procedures: a preliminary report. Neurosurgery 43:739­748, 1998 28. Lunsford LD, Kondziolka D, Bissonette DJ: Intraoperative imaging of the brain. Stereotact Funct Neurosurg 66:58­64, 1996 29. Lunsford LD, Parrish R, Albright L: Intraoperative imaging with a therapeutic computed

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC tomographic scanner. Neurosurgery 15:559­561, 1984 30. Lunsford LD, Rosenbaum AE, Perry J: Stereotactic surgery using the "therapeutic" CT scanner. Surg Neurol 18:116­122, 1982 31. Maciunas RJ, Galloway RL Jr, Fitzpatrick JM, et al: A universal system for interactive image-directed neurosurgery. Stereotact Funct Neurosurg 58:108­113, 1992 32. Nitta T, Sato K: Prognostic implications of the extent of surgical resection in patients with intracranial malignant gliomas. Cancer 75:2727­2731, 1995 33. Parkinson D, Legal J, Holloway AF, et al: A new combined neurosurgical headholder and cassette changer for intraoperative serial angiography. Technical note. J Neurosurg 48:1038­1041, 1978 34. Peeling J, Sutherland GR: High-resolution 1H NMR spectroscopy studies of extracts of human cerebral neoplasms. Magn Reson Med 24:123­136, 1992 35. Preul MC, Caramanos Z, Collins DL, et al: Accurate, noninvasive diagnosis of human brain tumors by using proton magnetic resonance spectroscopy. Nat Med 2:323­325, 1996 36. Roberts DW, Hartov A, Kennedy FE, et al: Intraoperative brain shift and deformation: a quantitative analysis of cortical displacement in 28 cases. Neurosurgery 43:749­760, 1998 37. Steinmeier R, Fahlbusch R, Ganslandt O, et al: Intraoperative magnetic resonance imaging with the magnetom open scanner: concepts, neurosurgical indications, and procedures: a preliminary report. Neurosurgery 43:739­748, 1998 38. Sutherland GR, Peeling J, Sutherland E, et al: Forebrain ischemia in diabetic and nondiabetic BB rats studied with 31P magnetic resonance spectroscopy. Diabetes 41:1328­1334, 1992 39. Tronnier VM, Kunze S: Comment on Steinmeier R, Fahlbusch R, Ganslandt, O, et al: Intraoperative magnetic resonance imaging with the magnetom open scanner: concepts, neurosurgical indications, and procedures: a preliminary report. Neurosurgery 43:739­747, 1998 40. Tyson RL, Sutherland GR, Peeling J: 23Na nuclear magnetic resonance spectral changes during and after forebrain ischemia in hypoglycemic, normoglycemic, and hyperglycemic rats. Stroke 27:957­964, 1996 41. Watanabe E, Mayanagi Y, Kosugi Y, et al: Open surgery assisted by the neuronavigator, a stereotactic, articulated, sensitive arm. Neurosurgery 28:792­799, 1991 42. White DN: The early development of neurosonology: I. Echoencephalography in adults. Ultrasound Med Biol 18:115­165, 1992 43. Winger MJ, Macdonald DR, Cairncross JG: Supratentorial anaplastic gliomas in adults. The prognostic importance of extent of resection and prior low-grade glioma. J Neurosurg 71:487­493, 1989

Manuscript received January 18, 1999. Accepted in final form February 15, 1999.

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC Address reprint requests to: Garnette R. Sutherland, M.D., Division of Neurosurgery, University of Calgary, 1403 29th Street N.W., Calgary, Alberta, T2N 2T9, Canada. email: [email protected].

Unauthenticated | Downloaded 09/26/21 01:51 AM UTC