Monument Valley in Arizona

Application of Frameless Stereotaxy to Minimally Invasive Neurosurgery


Mark Gerber, MD
Jeffrey S. Henn, MD
Kim Manwaring, MD
Kris A. Smith, MD

Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona


The development of minimally invasive procedures in neurosurgery has been facilitated by the use of frameless stereotaxy, which enables surgeons to plan operative approaches and to navigate in real time.  Minimally invasive procedures that benefit from frameless stereotaxy include brain biopsies, placement of depth electrodes for epilepsy surgery, selective amygdalohippocampectomy, placement of Ommaya reservoirs for treating basilar meningitis, endoscopic spinal procedures, and neuroendoscopy.  This article reviews the use of frameless stereotaxy with these procedures and examines possible future developments in this technology.

Key Words: frameless stereotaxy, minimally invasive neurosurgery

The development of frameless stereotaxy in the past decade has introduced a new era in neurosurgery.  The ability to plan surgical approaches with incredible accuracy and fidelity has enabled surgeons to operate more deftly and safely on lesions than previously possible. Consequently, the rates of morbidity and mortality associated with older techniques have decreased.  This article discusses the history of minimally invasive surgery and the surgical procedures that have lent themselves to minimally invasive approaches as a result of frameless stereotaxy.

Development of Minimally Invasive Neurosurgery

As discussed elsewhere,[3] the trend toward minimally invasive surgery began only 13 years ago when Mouret performed the first laparoscopic cholecystectomy. While gynecological surgeons had already embraced laparoscopy, the advent of improved lenses, fiber optics, and light sources set the stage for minimally invasive approaches. The late 1980s, with the development of new laparoscopic devices, witnessed a rise in the popularity of laparoscopic cholecystectomy. This technological achievement kindled considerable excitement in the general surgery community as it became readily apparent that many operations could be performed through small surgical openings rather than through the large incisions used with traditional open procedures. Training courses in laparoscopy flourished well into the early 1990s and continue today.

Laparoscopic surgery offered multiple benefits. Patients experienced less postoperative pain, which shortened the length of hospitalization. In turn, the costs for patients and insurance companies were lowered. Patients also enjoyed other benefits: Fewer days were lost from work, recovery times were shorter, incisions were more cosmetically pleasing, and the amount of internal adhesions was decreased. In essence, laparoscopic surgery revolutionized the field and all parties benefited. It was only natural that endoscopes and other minimally invasive techniques have found application in almost every surgical subspecialty, including neurosurgery. In fact, minimally invasive surgery has become so prevalent that many patients demand it. Just as neurosurgery has evolved from shaven heads to minimal strip shaves, scalp incisions and craniotomies are becoming smaller and the trend is likely to continue.

In most areas of surgery, lapa roscopy and endoscopy are the key elements behind minimally invasive procedures. In neurosurgery, how ever, equally important have been advances in stereotaxy and image guidance. Neurosurgery is particularly well suited to image guidance for several reasons. First, neurological surgery is highly dependent on visual- spatial comprehension. The ability to conceptualize pathology in multiple dimensions is needed to plan and execute a surgical approach to a lesion. Secondly, image guidance systems require a fixed relationship between anatomy and external landmarks. Because the brain is essentially immobilized within the fixed rigid skull, the head meets this criteria.

Knowledge of the complex anatomy of the skull base and the brain is mandatory for neurosurgeons. The intimate relationships between intracranial structures demand extensive study and experience. The details with which present image-guidance systems depict intraoperative anatomy allow complicated lesions to be localized precisely and treated effectively. Image-guidance systems also provide a powerful tool for teaching residents the intricacies of neuroanatomy. Three-dimensional and volumetric atlases are being developed to provide insight into surgical anatomy typically only mastered after years of operative experience. Finally, the ability to plan an operative approach and to navigate in real-time minimizes damage to otherwise normal structures, lowering rates of morbidity and mortality.

Minimally Invasive Image-Guided Neurosurgical Procedures

In the early 1990s, several MR image-based guidance systems became available commercially. Most of these systems used an articulated arm that relayed its position in space to a computer workstation. These early systems demonstrated the feasibility and utility of image guidance. Unfortunately, they were limited by the often cumbersome articulating arm. This limitation was the impetus to design frameless guidance systems using light emitting diode (LED) arrays as a means of triangulating objects in space. Microscopes with LED arrays have also been integrated into guidance systems.[7] The computer workstation interprets the microscope’s focal point as the point of interest and generates orthogonal views based on this information. In this case, the image guidance system is useful in that it helps to plan the surgical approach and to insure optimal lesion resection. Frameless stereotaxy has also affected the way in which frame-based stereotactic procedures are performed. At our institution, almost all brain tumors (supra- and infratentorial), arteriovenous malformations, and cavernous malformations are resected using frameless stereotaxy.

Stereotactic Biopsies

Stereotactic brain biopsies are common neurosurgical procedures. Classically, stereotactic biopsies are performed using a stereotactic frame such as the Leksell, Brown-Roberts-Wells (BRW), or Cosman-Roberts-Wells (CRW) frame, which is attached to the patient’s head. The patient is brought to the operating room (OR) where they are anesthetized, and the frame is affixed to their head. Next, the surgical team transports the patient to obtain magnetic resonance (MR) or computed tomography (CT) images with the reference frame in place. The patient is then returned to the OR, where the procedure is performed. Frame-based fiducials and target points are entered into a computer that calculates the frame settings. Entry points and trajectory can also be calculated.

Stereotactic brain biopsies can be performed using frameless stereotaxy, which provides several advantages. MR imaging or CT can be obtained any time before the procedure. This flexibility is convenient for scheduling and avoids issues related to transporting the patient after the frame has been placed. The image guidance workstations can be configured to view the trajectory on which a biopsy needle will travel. The surgeon can minimize the risk of injuring an artery or vein by selecting an entry point over a gyrus rather than a sulcus, and, theoretically, decrease the risk of bleeding during the procedure. The ability to change biopsy targets in the OR is yet another benefit of frameless image guidance. If nondiagnostic tissue is obtained, the target can easily be readjusted without reimaging or arithmetic calculations. This advantage is lacking during a conventional stereotactic biopsy.

One potential drawback to using frameless stereotaxy for biopsy is the lack of a rigid frame to serve as a structural support for the biopsy needle. Although the risk is small, freehand imaged-guided advancement of the biopsy needle can increase movement of the needle and decrease accuracy. To address this issue, various fixtures that attach to the OR table have been designed. In most cases, the fixture is a flexible arm that can be locked into position once a target trajectory has been selected. The distance to the target is then determined from the guidance system, and a movable stop is set to the appropriate depth on the biopsy needle. Other solutions include disposable plastic guides that screw into the entry holes, again providing rigidity to the biopsy needle.

Others have gone even further when using frameless stereotaxy for stereotactic biopsy. Moriarty et al.[8] recently published their results of frameless stereotactic brain biopsy using intraoperative MR imaging. Although some surgeons might argue that the use of intraoperative MR imaging for stereotactic biopsy is excessive, their results were impressive. Their overall mean error in target acquisition was only 0.2 mm. Of their 68 patients, 66 (97.1%) of the biopsies yielded diagnostic tissue. Only two patients (2.9%) developed an intraparenchymal hemorrhage as a complication from the procedure.

Placement of Depth Electrodes

Frameless stereotaxy is becoming increasingly useful in epilepsy surgery. In the past, placing depth electrodes was similar to placing a stereotactic biopsy needle in that a frame was attached to the patient’s head and transport was required. Placement of a depth electrode, however, is more forgiving than stereotactic biopsy because precise targets are less critical. Setting the stereotactic frame over each target involves a significant amount of intraoperative adjustment and manipulation of the frame, and considerable time is required for preoperative planning and data entry. Depth electrodes can be placed using a frameless technique similar to that of frameless stereotactic brain biopsy. Again, flexible guide arms can be used to stabilize the drill and the electrode as it is advanced. The use of frameless stereotaxy for the placement of depth electrodes can decrease operative time and potential risks to the patient as well.

Selective Amygdalohippocampectomy

Surgery for intractable epilepsy caused by mesial temporal sclerosis has also improved as a result of frameless stereotaxy. In the past, patients who met the criteria for surgery typically underwent temporal lobectomy, which required a frontotemporal craniotomy and retraction of the temporalis muscle. Although usually well tolerated, this procedure sometimes led to a bony defect and the possibility of atrophy of the temporalis muscle. In addition to the cosmetic issues of surgery, patients also experienced discomfort after surgery while chewing. Neurological deficits, including visual field defects, memory impairment, and language deficits, have also been attributed to this approach.

Several surgeons have published their experience with selective amygdalohippocampectomy as an alternative to anterior temporal lobectomy. Proponents believe that the lateral aspect of the temporal lobe, which is typically normal on pathologic evaluation, does not need to be resected. Furthermore, some evidence suggests that resection of the lateral temporal structures may be responsible for postoperative neurologic deficits and that only the mesial temporal structures harbor the epileptogenic focus. Wieser and Yasargil[10] described a pterional transsylvian approach to the mesial temporal structures, which improved outcomes.

To date at our institution, 68 patients with intractable temporal lobe epilepsy secondary to temporal mesial sclerosis have undergone a selective amygdalohippocampectomy through a trans-sulcal approach usually between the middle and inferior temporal gyri (Smith K, unpublished data, 2001). Because sulcal anatomy varies among individuals, frameless stereotactic guidance allows the small incision and craniotomy to be placed in the optimal location to allow complete hippocampal resection through a key-hole approach. The image guidance and trajectory views are critical in gaining access to the temporal horn of the lateral ventricle, which allows visualization of the amygdala and hippocampus. The extent of resection is also demonstrated intraoperatively with the neuronavigational system. Neuropsychiatric evaluations performed after this minimally invasive procedure have documented no new language impairments. Seizure control in these patients is also encouraging.


Situated in the desert southwest, our institution frequently treats patients with Coccidioides mycosis, a potentially deadly fungus endemic to our geographic area. The fungus can cause valley fever, central nervous system complications (including basilar meningitis), or both. Treatment for the meningitis involves placing an Ommaya reservoir to facilitate the administration of intrathecal amphotericin. Because the base of the brain is affected the most, the Ommaya reservoir is preferentially placed in the posterior fossa. Medication can then be instilled into the basilar cisterns. With image-guided stereotaxy, a tiny retrosigmoid craniostomy allows accurate placement of the catheter with minimal risk and maximal benefit. This technique has been used to place numerous Ommaya reservoirs with gratifying results. The reservoir itself is placed against the bone in the posterior occipital/mastoid region and is hardly noticeable compared to an Ommaya placed in the frontal region.

Spinal Surgery

The ability of frameless imaging to permit minimally invasive spinal surgery is another evolving area in neurosurgery. Several systems that allow image-guided spinal surgery are available. Because the bone is usually of more interest than the soft tissues in the spine, most of these devices are based on CT. The accuracy of registration and interference with the surgeon’s ability to maneuver in the operative field are two issues that still require improvement. Surgeons and device manufacturers are developing systems that will allow spinal instrumentation such as pedicle screws and rods to be placed percutaneously. Presently, placing spinal instrumentation requires large incisions and significant retraction of the posterior spinal musculature. Postoperative pain, lengthy operative times, and blood loss are complications that may be decreased with minimally invasive techniques. Certainly, many hurdles exist before minimally invasive techniques can be successfully applied to this type of surgery.

Laparoscopes and thoracoscopes have already also become more common in spinal surgery although experience with linking these devices with stereotactic image guidance is lacking. In these cases, the scope is introduced through small, puncture-type incisions. Navigation with the scope depends upon the surgeon’s view of and familiarity with the anatomy. Anterior lumbar interbody fusion, sympathectomies, thoracic discectomies, and resection of thoracic schwannomas are procedures commonly performed with these devices at our institution. It is only a matter of time before frameless stereotaxy is merged with these thoracoscopic approaches to improve navigation and guide resection.

Neuroendoscopy and Image Guidance

Neuroendoscopy and ventriculoscopy in and of themselves provide an exciting means of performing minimally invasive surgery, and a role for image guidance in this area is beginning to emerge. The concept of neuroendoscopy is not unlike that underlying larger laparoscopes and thoracoscopes. The primary difference, aside from the size, is that neuroendoscopes navigate in a liquid medium (cerebrospinal fluid) instead of in air.

At our institution, neuroendoscopy has many neurosurgical applications. Some of the more common endoscopic procedures include placement of ventricular catheters, removal of colloid cysts, fenestration of arachnoid cysts, third ventriculostomies, aqueductal stenting and plasty, and biopsy of intraventricular tumors.[7,9]

At least one company has developed an image-guidance system that allows both endoscopic and image-guided views to be displayed side-by-side in real-time. This powerful graphic capability enables the surgeon to compare the endoscopic view with the guidance system’s interpretation of the location of the endoscope.[4,5] This capability is particularly useful because the endoscope is limited to visualizing only surface features. Furthermore, endoscopic views are significantly degraded when blood obscures the field. Stereotactic neuro navigation systems provide the endoscope with “x-ray vision,” allowing it to “see” through structures and to navigate in suboptimal conditions. The increasing speed of microprocessors has permitted the complex arithmetic calculations needed for both orthogonal images and three-dimensional graphical rendering to be updated in real time. Layering algorithms provides the ability to enhance the virtual image of structures such as bone, blood vessels, and tumors with color. In general, these systems are robust and easy to use.

The use of endoscopes in neurosurgery continues to expand. The generally accepted standard of care for the treatment of deep-seated intracranial hemorrhages is nonsurgical follow up except when a hemorrhage exerts excessive mass effect. Several groups have reported stereotactic drainage of intracranial hematomas after they have liquified. Others have described the placement of catheters into the hemorrhage followed by instillation of thrombolytic agents through the catheter to improve the speed with which the clot is broken up and drained. Auer et al.[1] randomized 100 patients with spontaneous intracranial hematomas to endoscopic clot evacuation or observation. In the surgical group, the rates of morbidity and mortality were significantly reduced in younger patients. In another study, 13 patients with intracerebral hematomas underwent intraoperative MR image-guided evacuation with a cutting suction cannula.[2] The clot was evacuated completely in eight (62%) patients, and 75 to 90% of the clot was evacuated in another four (31%). Clot evacuation was subtotal in only one patient. Although it is still somewhat unclear if these patients benefited from hematoma evacuation, this study demonstrates the utility of frameless image-guided minimally invasive approaches to deep lesions and opens the possibility for similar procedures such as tumor resection.

The Future of Frameless Stereotaxy and Minimally Invasive Surgery

Considering the incredible technological advances made since the introduction of CT in the mid1970s, one can anticipate that perhaps the best of image-guided and minimally invasive surgery is yet to come. Likewise, minimally invasive endovascular techniques, such as the coils and stents used by interventional radiologists and neurosurgeons, are changing the practice of cerebrovascular surgery.

As with any new technological tool, limitations ultimately will be identified. Most frameless stereotactic neurosurgical navigation systems are based on preoperatively acquired images and are therefore unable to compensate for anatomic distortion that occurs during surgery from brain shift and edema. To alleviate this problem, intraoperative magnetic resonance scanners have been developed. These systems allow the patient’s head to be reregistered to a set of MR images acquired during surgery, thereby incorporating morphologic changes that occur intraoperatively. Advocates of intraoperative MR imaging have argued that updated image sets facilitate the resection of brain tumors because abnormal tissue can be identified before closing the case. Epilepsy surgery might also be improved because remaining tissue, such as residual ectopic grey matter, might be resected more completely.

The power of computers to generate detailed three-dimensional renderings will continue to improve. Yet another area of investigation includes the use of head-mounted displays (HMD) or heads-up displays (HUD). This technology, which was developed for the military, is now being used in the aerospace, automotive, entertainment, and medical industries. With HMDs, each eye sees slightly different images creating a stereoscopic effect. Levy et al.[6] reported the use of HMDs by the first assistant and ancillary OR personnel. Although the image seen in the HMD was not as good as that seen under the operating microscope, the first assistant did benefit from the stereoscopic view. Continued improvements in color liquid crystal displays and high-definition television will spark further improvements in this technology. Voice-activated control of such systems is a possibility as is the integration of HMDs to stereotactic systems so that the position of the surgeon’s eyes and head is monitored in a fashion similar to that used to track the operating microscope. Another possibility is the inclusion of picture-in-a-picture technology in HMDs that would enable computer-generated three-dimensional images to be displayed adjacent to the surgeon’s field of view, providing simultaneous projection of the surgical field and stereotactic images.

The role of image guidance in minimally invasive surgery will continue to evolve. The merging of CT, functional MR imaging, ultrasonography, and MR spectroscopy into image-guided systems will allow neurosurgeons to treat exceedingly complex lesions through smaller incisions. Advances in the application of minimally invasive techniques to lasers, ultrasonic aspirators, and even aneurysm clip appliers will continue. It is only a matter of time before surgical robotic arms are coupled with stereotactic image-guided systems. This is an exciting and unique time in medicine as technology increasingly improves our diagnostic and interventional capabilities.


  1. Auer LM, Deinsberger W, Niederkorn K, et al: Endoscopic surgery versus medical treatment for spontaneous intracerebral hematoma: A randomized study. J Neurosurg 70:530-535, 1989
  2. Bernays RL, Kollias SS, Romanowski B, et al: Near-real-time guidance using intraoperative magnetic resonance imaging for radical evacuation of hypertensive hematomas in the basal ganglia. Neurosurgery 47:1081-1090, 2000
  3. Gadacz TR: Update on laparoscopic cholecystectomy, including a clinical pathway. Surg Clin North Am 80:1127-1145, 2000
  4. Hopf NJ, Grunert P, Darabi K, et al: Frameless neuronavigation applied to endoscopic neurosurgery. Minim Invasive Neurosurg 42:187-193, 1999
  5. Hopf NJ, Perneczky A: Endoscopic neurosurgery and endoscope-assisted microneurosurgery for the treatment of intracranial cysts. Neurosurgery 43:1330-1337, 1998
  6. Levy ML, Chen JCT, Moffitt K, et al: Stereoscopic head-mounted display incorporated into microsurgical procedures: Technical note. Neurosurgery 43:392-396, 1998
  7. Manwaring K, Crone K: Neuroendoscopy. New York: Mary Ann Liebert, Inc., 1992
  8. Moriarty TM, Quinones-Hinojosa A, Larson PS, et al: Frameless stereotactic neurosurgery using intraoperative magnetic resonance imaging: Stereotactic brain biopsy. Neurosurgery 47:1138-1146, 2000
  9. Oi S, Shibata M, Tominaga J, et al: Efficacy of neuroendoscopic procedures in minimally invasive preferential management of pineal region tumors: A prospective study. J Neurosurg 93:245-253, 2000
  10. Wieser HG, Yasargil MG: Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy. Surg Neurol 17:445-457, 1982