Exploring Frameless Stereotactic Image Guided Surgery

MARCH 1999, VOL 69, NO 3 Tessman

Exploring Frameless Stereotactic Image Guided Surgery

I

space of the cranium. At this time, the term stereotacrus-based on the Greek word stereo, meaning three dimensional, and the Latin word tactus, meaning touch-was introduced.’ Framed stereotaxy. The first attempt to develop this method was the construction of a stereotactic frame in the early 1 9 0 0 ~ This . ~ technique-which was not actually tested on a human until the 1940s-worked by mounting the stereotactic frame to a patient’s head using the exterior auditory canals and the inferior orbital rims as landmarks.’ With these landmarks present on plain x-rays, neurosurgeons could calculate desired intracranial locations. This technique ultimately was abandoned because of the inability to calculate BACKGROUND accurate adjustments when a patient’s brain tissue had The idea of navigating three dimensionally in been shifted because of lesions: the skull has been around since the beginning of the Modern advances. Stereotactic surgery was century. The objective was to develop a method to not practiced for many decades until the 1980s, define a specific point within the three-dimensional when advances occurred in computed tomography (CT) and magnetic resonance imaging (MRI).5 These radiA B S T R A C T ographic modalities provided the Frameless stereotactic image guided surgery (IGS)-a fairly new ability to see abnormal shifting of modality originally used in cranial neurosurgical procedures-has the brain and determine a desired been found to have several advantages over traditional framed meth- three-dimensional location in the ods of three-dimensional navigation. Frameless stereotactic IGS is cranium. Stereotactic neurosurproving useful in neurosurgical procedures by allowing screws to be gery quickly became standard placed in the spine more quickly and accurately than with traditional practice for assisting in lesion methods, As IGS becomes standard clinical practice for certain spinal resections and collecting deep procedures-overlapping both neurosurgery and orthopedic special- brain tissue for biopsy. This was a ties-new ideas for surgical application of the technology are devel- big step for neurosurgery, providoping. Currently, frameless stereotactic IGS is being investigated for ing surgeons with greater confiuse in hardware placement in orthopedic procedures and for endo- dence and patients with better scopic navigation in otorhinolaryngologic sinus procedures. outcomes and quicker recoveries. Frameless stereotactic IGS is expected to gain application approval in Frameless stereotaxy. With both orthopedic and otorhinolaryngologic specialties from the US continued advances in computer Food and Drug Administration in the near future. AORN J 69 (March technology, the idea of frameless navigation developed. The first 1999) 498-51 2. mage guided surgery (IGS) is a broad term referring to a number of radiographic modalities used in anatomic localization procedures. One of the most recent modalities is a technology called frameless stereotaxy-a term originating from the technology’s initial use in cranial neurosurgical procedures. Frameless stereotactic IGS provides a new way to navigate three dimensionally without having to mount a frame to the patient’s head. Many manufacturers produce this equipment, which is seen widely throughout neurosurgical ORs. Now this technology is broadening, and new applications in other areas and specialties are being implemented.

CHARLES TESSMAN, RN

498 AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman

To use IGS, landmarks

Advantages for spinal procedures include reduction and possible elimination of intraoperative radiation exposure; improvement of hardware placement accuracy, providing better patient outcomes; easier methods for surgical planning (eg, determining screw length and width for a particular vertebra preoperatively); reduction in surgical procedure time by making hardware placement decisions more efficient.

must be established on the patient’s skin that can

be seen by MRI or CT.

SYSTEM COMPONENTS

neurosurgical procedure using a frameless technique was performed in the early 1 9 8 0 ~ Since .~ then, many frameless systems have been manufactured, and clinical trials for cranial surgery by the US Food and Drug Administration (FDA) have been completed and approved. The FDA next explored and approved a new application of the frameless technique in spinal surgery, which enabled surgeons not only to navigate to specific points but also to determine trajectories in the anatomy. This proved to be extremely useful for determining angle, length, and width of hardware placement. It is now common practice across the United States to use frameless stereotaxy for both cranial and spinal neurosurgical procedures. FRAMELESS STEREOTAXY ADVANTAGES

Frameless stereotaxy in cranial and spinal procedures has many advantages over traditional methods for surgeons and, more importantly, for patients. Advantages for cranial procedures include real-time interaction with MRI images, allowing more precise surgical planning; smaller incisions and craniotomies; less invasive excisions, which may improve patient outcomes; less restriction of surgical access because of lack of stereotactic frame; less patient discomfort from having to wear a stereotactic frame to and from the CT or MRI unit and OR; less preparation time than framed stereotaxy because the frame does not need to be applied to the patient and calculations do not need to be made in the CT or MRI unit; easier use with pediatric patients, as less cooperation is needed.’

Although today several manufacturers produce different frameless stereotactic systems, the purpose and functions of the systems basically are the same. The main variations in the systems involve software and tools. Most systems have two separate pieces that can be rolled into the OR (ie, a cart, which houses the components, and a camera, which is mounted on a separate boom) (Figure 1). All frameless stereotactic systems include the following components: a high-power computer that has the ability to instantly store and process MRI or CT images to provide real-time feedback; a high-resolution monitor that shows the necessary detail of an MRI or CT image; user-friendly software that does all calculations and provides an easily understood layout of images; tools with light-emitting diodes (LEDs) that the surgeon can use-both sterile and nonsterile-to interact with the computer (Figure 2); a two- or three-headed optical camera to track tools; a network connection or tape or disk drive to receive images from the MRI or CT unit; a tracking device with LEDs to account for patient or bed movement during the surgical procedure; and a backup power supply or surge protector to avoid damage to the computer if a power failure or surge occurs. REOUIREMENTS

With advances in technology and the competition among manufacturers, requirements for OR nurses and surgeons to make these systems work are few. Landmarks for registration. The main requirement of using a frameless system is establishing fiducials (ie, landmarks) on the patient’s skin that can be seen on the MRI or CT scan. This process, 500

AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman 9

known as registration, allows the surgeon to tell the system the exact location of the patient's landmarks during the procedure. Registration can be performed by either of two methods. Registration for cranial procedures. The first method mainly is used for procedures on a part of the head and must be performed before the MRI or CT scan is performed. In this method, trained personnel (eg, surgeon, radiologist, physician assistant, nurse) first apply an ink mark to the skin using a deep staining dye (eg, costellani paint, methylene blue). The radiographic marker-which can be seen by either MRI or CT scan-then is placed directly above the ink mark. As long as the ink mark is still present at the time of surgery, the same point can be seen on both the patient and the images. Unless these types of marks are applied as tattoos (ie, permanently), this method must be performed the same week as the surgery, and strict instructions must be given to the patient and family members to not wash off the marks. Registration for spinal or orthopedic procedures. The second method of registration is used for spinal or orthopedic procedures and is performed during surgery. In this method, finding fiducial points is accomplished differently and can be more difficult. Here the surgeon identifies rigid points directly on the bone that can be seen on the MRI or CT images and during the procedure. For example, using this method, the surgeon can register a vertebra using the lateral tips of the right and left transverse process and the posterior tip of the spinous process. MRI or CT sequence. The right MRI or CT sequence for the frameless software also is needed to use the stereotactic technology. All manufacturers have different requirements; however, most typically need images to be acquired in a specific format. The critical element involves how thick the image slices are and that they remain at that thickness throughout the scanning process. This becomes an issue when dealing with CT images because of radiation exposure to the patient. The thinner the slices, the better the image resolution-and, in turn, the more radiation exposure to the patient. The radiologist and the surgeon determine the CT slice thickness and the scan parameters, which should be established as a standard protocol for each patient. Patient compliance. The last requirement for frameless technology is patient compliance. This is minimal compared to traditional stereotactic methods. A patient undergoing a surgical procedure of

.

Figure 1 Most image guided surgery systems have two separate pieces; the cart holds the components, and the camera is mounted on a separate boom. This system features a two-headed optical camera. (Photo courtesy of Picker International, Inc, Cleveland)

503 AORN JOURNAL

Figure 2 A basic Y probe with four lightemitting diodes, which are calibrated to the six-inch tip. As the probe is moved, the optical camera tracks the tip's three-dimensional position.

-

MARCH 1999, VOL 69, NO 3 Tessman

Functional capacity of stereotactic software has become extremely powerful due to the growth in technology.

the head must have fiducial ink marks applied before the MRI or CT scan begins. This requires approximately 30 minutes of preparation time. For this, the patient must be well informed of the purpose of the ink marks and understand that the marks must stay on until the time of the procedure. The patient also must understand the importance of remaining completely still throughout the entire scan, as any recorded motion can decrease the accuracy of the frameless system’s ability to localize. TECHNICAL FUNCTIONS

The functional capacity of these stereotactic software programs has become extremely powerful because of the competitive growth in technology in the past few years. The programs also have become more user friendly-a benefit for busy OR staff members. Most of the interaction with the system is performed preoperatively and just before the first use of the device. Before the procedure begins, the radiologist or circulating nurse spends a few minutes setting up the system, including placing the equipment in the room so that the monitor is easily visible, plugging in the camera, and manipulating the images that radiology personnel previously transferred to the system. At this point, each manufacturer has different software functions to manipulate the images. Regardless of the software used, the images must be oriented correctly, and the fiducial points for registration must be stored. Registration. Storing the image fiducials for registration is done by either fiiding the ink marks on the patient’s skin-which were placed before the MRI or CT scan was performed-r finding distinct marks on the anatomy. After the marks are found, the surgeon assigns numbers to these points and inputs their locations in the system. These points then are ready to

be “picked” on the patient by touching a tool or probe that has LEDs to the exact spot on the patient as the stored points are identified on the image. This process allows the system to know where the patient is located and can be performed at one of two points of the procedure. If the fiducial marks are on the patient’s skin, such as in a craniotomy procedure, the marks should be picked on the patient after the patient is completely positioned and before sterile scrubbing or draping occurs. If the fiducial marks are on the anatomy, such as in spinal surgery, the marks should be picked during the procedure after the particular anatomical points are exposed. The system automatically will do the math and calculate a registration error in millimeters. This error is related directly to how close the location stored in the image is to where it was picked on the patient. Typically, the user will not be able to match up the two sets of points exactly. For most surgical procedures, if the volume of points is off by more than 2 mm, the registration should be redone to lower the error. A registration error under 2 mm is considered highly accurate. Variations in software can influence the accuracy of the registration process. Some systems are easier to register and, therefore, seem to have better accuracy. After accurate registration is achieved, it is equally important that it is maintained. To do this, almost all frameless stereotactic systems include a tracking device. On some systems a tracking device is an FDA requirement. This device typically is a sort of block, star, or arch that has LEDs attached to it. The device is mounted to the part of the anatomy being registered and will account for any movement that takes place. For example, in a craniotomy procedure, the tracking device often is mounted to the neurosurgical threepoint fixation headrest. This will track any movement of the patient or camera as long as the tracking device moves with relation to the head. The tracking device and the registered anatomy must move together or registration can be lost. If this happens, reregistration is required. To do this, the surgeon would have had to establish backup points on the skull during the initial registration that could be used as landmarks. Alternatively, the surgeon would have to identify new anatomical landmarks on the head that could be seen by the CT or MRI scan. Neither of these methods of recovery has as much accuracy as the initial registration.

504 AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman

It is important for OR staff members to be taught the basic functions of this system. Most complications in the functioning of frameless technology occur because someone distorts the delicate relationship between the camera, tracking device, and patient’s position by bumping the tracker or using it as an armrest. Cranial application. The main function of frameless stereotaxy in cranial neurosurgery is to provide anatomic localization during surgical planning and intraoperative intervention. After the patient’s position is registered, the surgeon simply can place the frameless probe on the patient’s head and see the location of the fiducial marks on the images. This is extremely Figure 3 Frameless stereotaxy can be used to approach and resect an valuable in deciding the Precise intraventricular tumor. The top right view lets surgeons interact with an area for a smaller incision and actual MRI three-dimensional reconstruction of the head, brain, and tumor. bone flap and for determining a safe approach to avoid crucial brain tissue and large vessels in the targeted area Most frameless stereotaxy software is capable of (Figure 3). manipulating images so that they adjust to the angle Lesion resection. During procedures such as of the frameless probe. With this capability, the surlesion resection, frameless stereotaxy can provide geon can put the frameless probe tip to the desired navigation to help determine the lesion’s border. This entrance point and angle it to create a vector to safecan be extremely useful if the patient has had previ- ly obtain the desired tissue. This application is ous surgeries in the same affected area. In these extremely helpful in procedures in which the tumor cases, it often can be hard to tell the difference does not show up on a CT scan (eg, low-grade between tumor, scar, and brain tissue. By using gliomas, cavernous angiomas, arteriovenous malframeless stereotactic IGS and MRI with contrast, formations, mesial temporal sclerosis). The biopsy this differentiation can be more easily seen. then can be performed using a contrast MRI and Although this type of application can help the neurosurgical three-point fixation headrest. make crucial decisions, the technology can be lim- Although this method has the benefit of eliminating ited depending on which area of the brain is being the stereotactic frame, most surgeons prefer the examined and if diuretics such as mannitol or framed method because of the infancy and learning furosemide have been administered to decrease curve associated with IGS technology. intracranial pressure. These medications work to Spinal application. As mentioned, frameless “shrink” the brain, which causes the brain to shift in stereotaxy application for spinal surgery evolved the direction of gravity. As IGS cannot account for from its use in cranial procedures. The main function this type of shift, it no longer is accurate for naviga- of IGS in spinal procedures has been for screw placetion within the brain. For procedures that require ment during spinal fusion. It also has been useful for these diuretics, IGS can be helpful in the plan and resection of some spinal tumors (eg, swannomas, approach, but not the resection. meningiomas, neurofibromas, vascular lesions, Bruin biopsies. A more recent cranial applica- hemangiomas, metastasis), in which the navigation tion of frameless stereotaxy is for brain biopsies. function is the same as in cranial procedures. 505 AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman

Achieving accurate registration The biggest issue in using IGS in the spine is the fact that the spine is segmented. When a CT scan is performed on a part of the spine, the position of the spine in the CT unit will be different from its position while on the OR bed. In fact, the patient’s lumbar spine typically is scanned in supine position and then placed in prone for the surgical procedure. A surgeon can-

Figure 4 * (top) The upper two views indicate where the probe tip is located on the anatomy. The lower two indicate the screw width, length, and trajectory needed. (right) Using this, the pedicle screws can be placed accurately in the lumbar spine.

not, therefore, register the IGS system to one vertebra and assume the next vertebra-whether superior or inferior-is in the same place as when the initial scan was performed. So, to accurately set screws, each vertebra that receives hardware must be registered separately. For example, in an L-4 to L-5 fusion, the surgeon starts with L-4 and registers the left and right transverse and spinous process of that vertebra. As a result, the computer knows exactly where L-4 is, and screws can be accurately set. After these holes are drilled, the surgeon moves on to the next vertebra. Attaching the tracking device to the spine before registration can prevent inaccurate registration because of respiratory or OR bed movement. Setting the screws. With accurate registration, IGS systems can help set certain screws (eg, pedicle screws) more accurately and faster than with a fluoroscope or x-ray. After the vertebra is registered and the surgeon has visually confirmed its accuracy, the surgeon places the tip of the probe on the desired entrance point of a screw. Then the surgeon simply changes the angle of the probe to find the ideal trajectory. This trajectory is shown as a line extending into the vertebra that matches the screw’s length and width in relation to the actual size of the anatomy (Figure 4).Some software has the capability to plan the exact screw length and width before the procedure begins. One of the major advantages of IGS over other modalities is its ability to show this real-time information in more than one image view. When using a fluoroscope, the surgeon typically will ask for several views. To do this, the fluoroscopy equipment must be rotated to make sure an ideal trajectory is chosen. This can be both time consuming and inconvenient. 508

AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessnian

ADVANCED APPLICATIONS

With the continuing growth in computer technology, new applications are being explored to make surgical procedures faster and better. The current focus for IGS advancement is in orthopedic and otorhinolaryngologic procedures.

Otorhinolaryngology. Currently only a few IGS manufacturers customize hardware and software for otorhinolaryngologic procedures. The main interest is in providing additional information for endoscopic sinus surgery and helping reduce the morbidity and mortality associated with the current surgical methods.’ This is accomplished by setting up the IGS to track the endoscope tip as to its location in the sinus and project its location onto a CT image. This type of real-time feedback, along with the videotape from the endoscope, has many advantages. First, it allows the surgeon to interact with easily seen pathology on the preoperative CT image. Second, it provides residents, medical students, and other OR personnel with a better appreciation of the sinus anatomy. Finally, it helps speed up surgical procedures and reduce major surgical compli~ations.~ Orthopedics. The expected growth of the technology for orthopedic procedures is aimed at assisting difficult screw placement to secure fractures or hardware. Presently, the department of orthopedic surgery at the University of New Mexico Hospital, Albuquerque, has performed 12 procedures using the IGS equipment to assist screw placement. Eleven of the procedures were open reduction and internal fixation (ORIF) of complex pelvic acetabulum fractures, and one procedure was a femur fracture. For the acetabular procedures, IGS was found to be a reliable method in setting screws just adjacent to the joint space. Surgeons also found IGS

Figure 5 (top) Image guided surgery is used to acquire a trajectory to place a 130 mm x 6.5 mm lag screw through a fractured iliac crest. (left) In the postoperative anterioposterior x-ray, the lag screw is easily seen.

reliable in determining screw length and ideal trajectories and keeping the screws out of the joint space. Surgery time and radiation exposure were reduced in some procedures, especially in fractures requiring long lag screws (eg, > 100 mm) intended to bring two pieces of the pelvis together. For the femoral procedure, IGS was found to be equally helpful in placing screws to secure a femoral rod. In femoral procedures, IGS’s goal is to reduce surgery

509 AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman

With IGS, the process took minutes; with standard

fluoroscopy, it could have

taken more than one hour.

time and radiation exposure and increase the accuracy in putting screws through the distal holes in long bone roads after they have been inserted into the bone. Currently, there are several methods of doing this, but IGS may become the most accurate and expedient technique. CASE STUDY

Mr T was a 51-year-old male who fractured his pelvis while skiing. He was airlifted to the University of New Mexico Hospital, where he was seen for a left iliac crest and complex acetabular fracture. When admitted, he received a CT scan of the entire left side of his pelvis, in the proper protocol to be compatible with the IGS equipment. Three days after admission, Mr T was taken to the OR for O R F of the left acetabulum. After obtaining exposure to the fracture sights, the surgeon registered the IGS system. The surgeon obtained fiducial points from the bone by touching the probe to three easily seen points (ie, anterior superior iliac spine, two sharp points on the fracture lines). Before registering, a star tracking device was mounted on a stabilizing pin that already was screwed into the iliac crest. This would maintain registration if table or camera movement occurred. Registration was accomplished in several seconds with less than 1 mm of error. The surgeon then placed the probe on a known point to visually confirm the low reported error. Several separate pieces of bone needed to be reduced before setting the hardware because of the severity of the fracture. This would alter the anatomy in relation to the preoperative CT scan and, thus, affect accuracy. To fix this problem, each separate piece of bone was registered separately. This worked exactly as it would for spinal procedures,

registering each vertebra separately. This process would take a little more time initially; however, surgical time would later be reduced because of the added information. The IGS system was used to find the length, width, and angle for several screws placed adjacent to the joint space to secure braces. The surgeon used fluoroscopy to verify screw positioning, as this type of IGS application still was under FDA investigation. In addition to placing the screws to secure braces, the surgeon placed two long lag screws to secure fractured pieces, using IGS to determine the angle. Screws were placed in the iliac crest without any assistance from IGS imaging. The most beneficial timesaver came from using the IGS to help drive a 130 mm x 6.5 mm lag screw from the top of the iliac crest into the posterior column of the acetabulum (Figure 5). This length of screw is difficult to keep in the bone through the entire iliac crest. With IGS, this process was accomplished in minutes, whereas with standard fluoroscopy, it could have taken more than one hour, exposing the patient to that much more radiation. LIMITATIONS

Although IGS has proved to be a useful tool in many types of surgery, the technology has its limitations. Maintaining registration. The biggest limiting factor is in maintaining registration. Often the accuracy of IGS is altered by OR staff members rnanipulating the tracking equipment or the exposed anatomy. If the tracking device moves separately from the anatomy, loss of registration will occur. This is recoverable if the surgeon still has fiducial landmarks from which to reregister. If no landmarks are available, the procedure must be performed without assistance from IGS, in which case the equipment can be packed up to save space in the OR. Altering anatomy. Another limitation is created if the surgeon alters anatomy before registration. This often is seen in spinal procedures when the surgeon clips off anatomical landmarks such as the spinous and transverse process in visualization. In doing this, the points on the CT scan no longer match the anatomy. Anatomy also can be altered and useful landmarks can be destroyed during spinal decompression procedures; therefore, holes for implants must be drilled before decompression is performed. Altering the anatomy can be prevented by teaching OR staff members how IGS works and ensuring that surgeons 510

AORN JOURNAL

MARCH 1999, VOL 69, NO 3 Tessman

understand the technology’s requirements. Navigation after brain shift. As with most new technologies, the hardware and software themselves can be limiting. One of the biggest limitations of IGS has existed since the technology’s first applications in cranial surgery and involves its ability to adjust to brain shift during the procedure. All frameless stereotactic systems for cranial procedures register anatomic location by landmarks on the skin or skull. If the brain has shifted within the skull because of lesion resection or diuretics, navigation within the brain tissue can be difficult. This is especially noticed in procedures involving the temporal lobes. Currently, there is no good fix to this problem, but ideas are being explored. These might include a sort of LED strip that is applied to the brain tissue, an integration with ultrasound to localize the shifted brain surface, or the ability to obtain new MRI or CT images during the procedure. To date, there are a handful of sites throughout the country in which neurosurgeons are performing craniotomies within MRI suites, gaining the ability to obtain fresh images to register the frameless stereotactic equipment. Although this can be ideal, the average cost of such setups is approximately $4 million. NOTES 1. D E Bullard, B S Nashold,

“Evolution of principles of stereotactic neurosurgery,” Neurosurgery1 Clinics of North America 6 (January 1995) 27-41. 2. Bullard, Nashold, “Evolution

of principles of stereotactic neurosurgery,” 27-41; D League, “Interactic, image-guided, stereotactic neurosurgery systems,” AORN Journal 61 (February 1995) 360-370. 3. Ibid. 4. Ibid. 5. J B Anon, “Computer-aided

SUMMARY

Frameless stereotactic IGS systems have been providing valuable information in cranial and spinal procedures during the past few years. This technology is developing fast enough that it will soon be standard in ORs everywhere. It also will become more common to see surgeons in other specialties such as orthopedics and otorhinolaryngology using this technology on a regular basis as the FDA grants approval for expansion into these areas. With this type of recognition and the advances in computers and imaging, IGS will continue to provide beneficial information to make surgery safer, faster, and easier. A Charles Tessman, RN, BSN, is a regional managerfor image guided surgery at DePuy AcroMed, Cleveland. At the time this article was written, he was the director of image guided surgery at the Veterans Affairs Medical Center, Albuquerque. The opinions or assertions contained in this article are the private views of the author and are not to be construed as oficial or as reflecting the views of the US VeteransAffairs Administration.

endoscopic sinus surgery,” Laryngoscope 108 (July 1998) 949961; J Gybels, P Suetens, “Image guided surgery,” VerhandelingenKoninklijke Acadernie voor Geneeskunde van Belgie 59 (January 1997) 35-37; M W Vannier, J L Marsh, “Three-dimen-

computer integrated surgical technique for percutaneous fixation of transverse acetabular fractures,” Lecture Notes in Computer Science 1205 (1997) 565-572; P J Kelly,

“Volumetric stereotactic surgical resection of intra-axial brain mass lesions,” Mayo Clinic Proceedings

sional imaging, surgical planning, and image-guided therapy,”

63 (1988) 1186-1198. 7. League, “Interactic, image-

Radiologic Clinics of North America 34 (May 1996) 545-563.

guided, stereotactic neurosurgery systems,” 360-370. 8. Anon, “Computer-aided endoscopic sinus surgery,” 949-961.

6. Anon, “Computer-aided endoscopic sinus surgery,” 949-961; D M Kahler, R Zura, “Evaluation of a

512 AORN JOURNAL

9. Ibid.