| |
|
|
|
|
|
|
|
|
| |
|
|||||||
|
|
||||||||
| |
||||||||
|
Fluoroscopically-Assisted Stereotactic Laser Localization of Intracranial Lesions: Technical Note. Michael K. Landi, M.D., Walter Grand M.D., Department of Neurosurgery, State University of New York at Buffalo. Introduction: A new method of localizing intracranial lesions for biopsy or resection was developed to provide an alternative to conventional frame-base stereotactic and dedicated frameless image-guided surgery systems. This method utilizes computed tomography (CT) for preoperative intracranial localization and the Dual Radiation Targeting System (MINRAD INC., Orchard Park, NY), a laser targeting system, for intraoperative fluoroscopically-guided localization. The DRTS™ Platform attaches to a standard C-Arm fluoroscope (fig.1a). The principal features of the DRTS™ are illustrated in Fig 1b). The system identifies a line of x-ray radiation from the x-ray source to the image intensifier and positions a visible laser beam (approximately 690 mm wavelength) collinear with that line. A cross hair target symbol displayed on the x-ray image video monitor represents a point on the line of radiation defined by the laser beam. During fluoroscopic imaging, the DRTS™; target symbol is positioned over the x-ray image of a deep tissue structure, and the x-ray is turned off. The laser beam indicates the surface point of entry and angel of approach to the target. The system has a +/- 1 mm accuracy at the center of the laser spot as measured on an optical bench. Materials and Methods: For this preliminary study, six intracranial tumors that could be approached along a trajectory imaged by CT (i.e., target lesion and surface point of entry were on the same CT image slice) were selected for DRTS™ laser localization (fig. 2). The tumors ranged in size from 1.5-3.0 cm; five patients with supratentorial lesions including three meningiomas, two glioblastomas, and one posterior fossa metastasis in one patient.
On the day of surgery, a preoperative localizing cranial CT scan was obtained using the following procedure: 1) Acquire thin slice axial CT image through the tumor; 2) Select CT target slice (the axial image that contained the tumor center); 3) Return CT table to the coordinates of the target slice; 4) Use CT image marker light to position two radiopaque fiducials (2 mm diameter each) on the patient's head. This marker projects a laser light line across the patient's head, indicating the plane of the axial section for image acquisition. Fiducials are placed on this line, one at the desired surface point of entry (site of craniotomy, determined by the surgeon) and one at the opposite side of the head; 5) Repeat thin-slice CT through the target slice sections were obtained to demonstrate the relationship of the tumor and the two fiducials. Fiducials are in correct position if a line drawn between them intersects the desired intracranial target (Fig. 2); 6) After satisfactory fiducial placement, target depth was determined by measuring the distance from the outer table of the skull to the target using the CT console cursor along the fiducial line; 7) The final fiducial skin sites were marked with an indelible marker in the event a fiducial became dislodged. After the CT localization procedure, the patients were transported to the preanesthesia holding area or back to their rooms until surgery. In the operating room, after the induction of general anesthesia, the patient's head was secured in either a Mayfield skull clamp (Codman Inc., Raynham, MA) or a Sugita head holder (Mizuho Ikakogyo Co., LTD, Tokyo, Japan). The DRTS™ was used with a standard C-Arm fluoroscope (OEC 9400 or OEC 9600, OEC Medical Systems, Salt Lake City, UT) to align the fiducials such that they were superimposed on one another at the center of the cross hair target symbol (figs. 3 and 4); this relationship positioned the laser beam collinear with the line through the intended target defined by the two fiducials on computed tomography.
After fluoroscopic fiducial alignment was confirmed on the video monitor, the x-ray was turned off. The laser beam was used to indicate the surface point of entry and angle of approach to the target (Fig. 5). A burr hole was made at the position on the skull indicated by the laser beam. A ventricular catheter, aligned with the laser beam, was place to the depth measured during CT localization. The C-Arm was then removed from the operative field. A craniotomy was made around the catheter. Dissection was performed along the catheter to the tumor. Results: The tumors in all six patients were successfully localized and resected during the DRTS™. Two patients required one of the fiducials to be repositioned to intersect the tumor at a point closer to the center of the tumor. For the first two patients, it was technically difficult to align the fiducials due to physical interference from the headrest-support mechanism while positioning the C-Arm. Repositioning of these patients was required for alignment. Anticipation of this simplified subsequent cases. The Sugita headholder afforded the least x-ray obstruction to alignment of the fiducials and facilitated the greatest operative field. Radiolucent headholders were not required for any procedure and were not felt necessary for the application of this method. Discussion: Stereotactic targeting systems have achieved a high-level of sophistication over the last decade with advancements in imaging software and hardware technology (2, 3, 8). With these advancements, a proliferation of stereotactic systems have become commercially available to surgeons at prices ranging from $60,000-$450,000. Although highly accurate, these systems can be time consuming to set up and are frequently more complex than is necessary for the average neurosurgical localization procedure (3,8,9). Stereotactic neurosurgery, whether frameless or frame-based, requires the acquisition of image data points (image space) followed by the algebraic transformation of the image data set into a Cartesian coordinate system (stereotactic space) that will reference a point within the patient (patient space). Frame-based systems use a head-mounted frame to establish the reference for stereotactic space. Frameless systems utilize anatomic landmarks or an array of synthetic fiducials to develop a spatial reference. The accuracy of conventional frame-based systems is dependent on several well-established variables (1, 6). 1) The ability to localize the anatomical target in image space. This relies on the resolution of the imaging modality, slice thickness, and pixel dimensions. 2) The ability to establish a relationship between the image space and the stereotactic space. This is directly affected by geometric disturbances inherent in the imaging modality (principally magnetic resonance), the ability to resolve the fiducial reference points, and the method (manual vs. computerized) used to localize the fiducials or define their coordinates in image space. 3) The ability to reach the desired anatomical target once it has been transformed into stereotactic space. This is directly related to the mechanical stability of the frame and the accuracy of the calibration of it components (i.e., stereotactic coordinates may have +/- mm accuracy, however, mechanical error introduced by the frame system may add significantly to the error) (5, 6). The accuracy of the frameless systems is also affected by the above-mentioned sources of error, however, fiducial localization and image space conversion into sterotactic space becomes much more complex. Additional sources of error manifest with fiducial movement during image space acquisition, after acquisition or during the procedure, and with patient movement during image acquisition (head tilt, rotation, and axial shift). For tumor localization and craniotomy planning, information typically needed is the angle of approach and depth to a lesion from a selected surface entry point. Conventional frame-based systems provide this with high accuracy but they must be attached to the patient's head prior to CT or MR imaging and surgery, which frequently requires patient sedation and sometimes general anesthesia before attachment in the radiology suite (1,2, 4-6). Also, their size and physical features can restrict the operative field. Frameless stereotactic systems have been shown to have comparable accuracy to the standard frame-based systems (3,6,9) with the obvious advantages of an ambulatory patient in radiology, an unrestricted operative field, and intraoperative intracranial position information. Their disadvantages are the higher cost, the need for a trained technician, lengthy set-up time, and failure due to technical and/or operator errors. To provide an alternative to conventional stereotactic targeting systems, we developed a simple, inexpensive, frameless, highly accurate method for frequently encountered localization needs. Tumor Selection: The DRTS™ technique was useful in localizing selected tumors at a variety of intracranial locations. Although only tumors that were approachable along trajectories coplanar with axial CT sections were evaluated, the technique could be extrapolated to other CT orientations, such as coronal slices or non-conventional positioning, or used with magnetic resonance imaging. Limitations: Limitations of the technique included manipulation of the C-Arm fluoroscope in some trajectories due to the configuration of the operating table and headholder. In some instances, it was difficult to make the fine adjustments necessary for fiducial alignment with the older model C-Arm. Late model fluoroscopes provided the greatest maneuverability and highest resolution of the fiducials. Anticipation of these factors simplified later cases and facilitated rapid localization. Localizations were limited to targets approachable from a surface point of entry coplanar with the intended target represented by a single CT slice. Sources of Error: The accuracy of fluoroscopically-assisted laser targeting with the DRTS™ is +/-1mm on bench testing. The translation of this accuracy to actual clinical applications has yet to be established. Potential sources of error include those for frame-based and frameless systems. A distinguishing feature of the DRTS™ method is that conversion from image space to stereotactic space is unnecessary. The fiducial-fiducial line defines image space, patient space, and stereostactic space. Advantages: Compared to standard frame-based sterotactic techniques, the DRTS™ localization method eliminated the need for preoperative head frame placement, sedation, and intubation in radiology and the potentially hazardous transport requirements of patients to the operating room. Compared to the typical frameless stereotactic techniques, patients underwent a simplified preoperative localization imaging study with fiducial placement; downloading of imaging data was not required nor was a trained frameless technician necessary. Fiducial alignment is comparable to fiducial registration. Compared to both techniques, there is decreased set-up time, and decreased cost and complexity. Summary: Fluoroscopically-assisted laser targeting is an inexpensive, simple, alternative method for defining a surface point of entry and angle of approach to an intracranial target. The DRTS™ laser localization technique is useful for targeting needs such as endoscope or catheter placement, biopsy or tumor localization. Careful preoperative surgical planning with anticipation of limitations of the C-Arm and headholder configuration can facilitate rapid intracranial localization. Further studies are needed to compare the accuracy of this method to other stereotactic techniques. Disclosure: Landi is the inventor of the Dual Radiation Targeting System (DRTS™), manufactured by MINRAD INC. (Orchard Park, NY). Both authors are shareholders in MINRAD INC. and are actively involved in developing applications of this product. REFERENCES
Acknowledgement We thank Paul H Dressel for the preparation of figures and illustrations. (This abstract was accepted for the Congress of Neurological Surgeons 1998 Annual Meeting, October 3-8, 1998, Seattle, Washington. ) |