2.1 The 19^th Century

The gradual recognition that brain functions are localized led to the use of guiding devices to explore deep-seated regions of the brain in animals. In 1873 Dittmar, in Germany, described the use of a guiding device to make incisions in the medulla oblongata of rabbits. ² A decade and a half later St. Petersburg physicians conceived a cranial localization tool to be fixed to the patient’s skull to investigate function based on the concepts of phrenology-the idea that function is linked to external skull morphology. ³

2.2 The 20^th Century

Sir Victor Horsley and Robert Clarke are appropriately given credit for the development of the first stereotactic guiding device, which they described in the journal Brain in 1906. ⁴ The instrument could reliably direct a probe for the study of cerebellar physiology in cats. This first rectilinear system used an X-Y-Z axis to specify the target of a probe inserted through a holder mounted to the frame. It became the prototype for subsequent generations of stereotactic devices (▶ Fig. 2.1). Aubrey Mussen, a disciple of Clarke’s, subsequently created a device patterned after the Horsley–Clarke frame. ⁵ As reported by Phil Gildenberg, this device was potentially applicable to human brain surgery. ⁶ Clarke apparently spent time in both London and Montreal, where this device was left at the Montreal Neurological Institute. It is not clear whether any patients were treated.

Stereotactic devices reliably permit selection of a safe passage route, creation of a cranial opening, and placement of a probe into an intracranial target that has been detected by imaging. In 1918, Capt. Aubrey Ferguson published a description of the removal of intracranial bullets using a guiding device (▶ Fig. 2.2). ⁷ This report appears to be the first publication of the actual human use of such a technology – an external guiding device with a mounted instrument (an extended pituitary forceps) directed to a target seen on X-ray imaging. Considering that x-ray visualization of the body had only recently been described, this was a remarkable, pioneering, but little recognized contribution.

During the 1930’s Kirschner described the use of a cranial guiding device designed to facilitate transovale placement of a lesioning electrode to treat trigeminal neuralgia (▶ Fig. 2.3). ⁸ The 1940’s were dominated by the collaborative development of a number of stereotactic devices patterned after the original Horsley-Clarke concept. Ernst Spiegel, an Austrian neurologist who emigrated to Philadelphia to escape the Nazi Anschluss, teamed up with Henry Wycis a neurosurgeon working at Temple University to develop a practical stereotactic guiding system for use in human surgery. ^{9 , 10} This device, which they called the stereoencephalotome, was the first to use internal brain landmarks shown by encephalography (▶ Fig. 2.4). As described by Gildenberg, the first device had only translational movements of the probe or electrode. When the carrier was positioned above the target a probe could be directed to the target. In later versions, the devices were mounted to plaster casts fixed to patient’s heads. Some were rigidly fixed to the patient’s cranial vault and were simply aiming devices. Additional devices were designed to have angular adjustments of the probe trajectory in order to match the angles of lines directed to the target based on AP and lateral x-rays. The same concerns of stability, rigidity, and reproducibility of device fixation present 75 years ago remain for the current generation MRI compatible devices. Even minimal movements can distort the target position and the probe trajectory. Spiegel and Wycis focused on movement and refractory behavioral disorders so as to minimize the surgical risk involved in its treatment. In the 1940s and 1950s the widely performed operation of frontal leucotomy was done by the more invasive craniotomy until Freeman began the practice of transorbital lobotomy. ^{11 , 12} Thousands of patients underwent such procedures in an era devoid of psychotherapeutic drugs.

Lars Leksell had been trained as a neurophysiologist and was largely responsible in the 1930’s for the descriptions of the Gamma motor postural tone system. His collaborator, Ragnar Granit, later shared the 1967 Nobel Prize in Physiology or Medicine with George Wald and Keffer Hartline for his work on vision and retinal cones. Leksell went on to be concerned with what he considered to the poor neurosurgical outcomes of patients who underwent brain surgery in his home country of Sweden. Yet, his clinical mentor, Herbert Olivecrona, was the acknowledged Northern European master of neurosurgery, with a large experience in brain tumors and vascular malformations. Leksell, however, was convinced that less invasive and more accurate methods were needed to allow surgeons to reach deep-seated brain sites.

In 1947 Leksell traveled to Philadelphia for fellowship training with Spiegel and Wycis. When he returned to Stockholm, he submitted his landmark paper in which he described the prototype of a rectilinear coordinate system that bears his name (▶ Fig. 2.5). ¹³ The targeting was accomplished by direct imaging, when, for example, a calcified tumor capsule could be seen by skull x-ray. Otherwise, encephalography, with lateral and AP X-ray imaging, identified stable landmarks such as the anterior and posterior commissures. Targets in the internal capsule, globus pallidus or specific nuclei of the thalamus were found relative to these reliable points. Leksell’s concept was that these devices should be simple enough that even a neurosurgeon could master their use!

Leksell was a restless and often unsatisfied inventor and he needed his device to be optimized to the last little screw. As imaging evolved from x-rays and encephalography, Leksell evaluated the use of ultrasound, computed tomography, and magnetic resonance imaging. The frame needed to be redesigned in order to accommodate these imaging changes so as to maintain reliability and imaging compatibility. New devices included the standard frame of the 1950s, the D frame of the 1970s introduced after the early development of CT, and eventually the G frame which was redesigned to facilitate MRI targeting for both open as well as closed (radiosurgical) stereotactic procedures (▶ Fig. 2.6). The target is positioned at the center of two arcs. The arcs allowed movement of the probe on the head from side to side and anterior to posterior while the target and probe tip remained at the X, Y and Z intersection. This is the arc-centered principle of localization. The angular adjustments of the probe trajectory optimized the trajectory so that the target was always reached when the probe was advanced the distance of the radius of the arc (19 cm).

Other devices were developed at many U.S. and European centers during the 1950s. These included in the United States the Todd-Wells device that translated the target to the intersection of the arcs and the Richert-Mundinger system (Freiburg, Germany), which used polar coordinates to place the target at the center of a base coordinate head ring (▶ Fig. 2.7). A phantom simulator was created to adjust the arc system, or target bow, so that the arc would facilitate placement of the probe at the target. The arc system was then transferred to the patient’s base ring secured to the head, and the procedure began after appropriate burr hole placement. Tailarach and various students working in France used his apparatus to deliver multiple electrodes to targets selected for epilepsy surgery or implantation of radioactive isotopes for tumor management.

The eras of the 1950s and 1960s saw considerable stereotactic activity fostered by an immense interest in surgical options for behavioral, movement, and epileptic disorders. Multiple individual stereotactic devices were constructed depending on the interest of the stereotactic surgeon and then used at other centers when their students moved to new sites. For example, Irving Cooper in New York performed thousands of ablative procedures using his device. Edward Hitchcock, working in Birmingham, England developed a base square device that could be used for functional as well as morphological surgery.

In 1977 Russ Brown described the use of the N localizer to define targets in computed tomography stereotactic space. ^{14 , 15} Subsequent collaboration with manufacturer Trent Wells and neurosurgeon Ted Roberts led to the commercial sale and use of the Brown-Roberts-Wells (BRW) device (▶ Fig. 2.8). The target coordinates were determined using a computer system relative to a base ring after determination of the target in stereotactic space. The probe was adjusted using four angular measurements to reproduce a preoperatively chosen trajectory. Using similar principles, the team of John Perry, Arthur Rosenbaum, and Dade Lunsford developed a similar CT compatible guiding device under the auspices of Pfizer Pharmaceuticals, which at the time had a significant interest in imaging technologies (▶ Fig. 2.9). ¹⁶

Eric Cosman modified the BRW frame to facilitate imaging compatible functional and tumor surgery. The frame was then marketed as the Cosman-Roberts-Wells (CRW) device. Pat Kelly modified the original Todd Wells device to create his Kelly Stereotactic system. It incorporated CT imaging with laser and radiofrequency ablative technologies. His stereotactic operating suites at the Mayo Clinic and later at New York University became the most advanced neurosurgical operating rooms developed of the era. They combined advanced targeting imaging with precision CO2 lasers to vaporize deep-seated brain tumors. His functional neurosurgical practice also used radiofrequency ablation to create thalamic lesions for movement disorders.

In the 1980s modifications of these original devices continued. Lauri Laitinen in Finland described the Laitinen Stereoguide, consisting of an oval shaped base ring with rigid skull fixation and an attached arc to house the probe (▶ Fig. 2.10). ¹⁷ Various adapters were used to define the target based on the imaging modality selected – plain X-rays, MRI, or CT. ¹⁸

Most stereotactic devices were made from radiopaque metals. These required reengineering as imaging evolved to the use of CT and MRI. With CT scanning, dense metal could obscure the target or trajectory. With MRI, magnetic susceptibility artifacts became an issue. In response, a variety of aiming devices made of plastic or other imaging neutral compositions emerged in the 1980s. For example, Arun-Angelo Patil designed a device made of composite material that had a base platform and the now-standard detachable arc delivery system. ¹⁹

During this interval from 1977-2000 most neurosurgical training sites instructed residents on the safe and appropriate use of frame based technologies. The American Society of Testing and Manufacturing (ASTM) required that such devices be able to reliably place a probe at the target with an accuracy of +/−1 mm. Device accuracy however is dependent on the accuracy and reliability of the imaging used to define the target. By using digital subtraction angiography with a 1024 × 1024 grid, theoretical mechanical accuracy can approach 0.1 mm. Of course, this disregarded brain or target-shift during open stereotactic procedures. Using a 512x 512 grid during CT imaging, accuracy approaches 0.5 mm. For most MRI units, the grid size is 256 × 256, such that accuracies of 1 mm are the maximum that can be obtained. For MRI based stereotactic surgery it is critical that magnetic susceptibility artifacts not distort the target position and imaging. The magnets must be properly shimmed and maintained, and regular verification studies with phantoms must be done to verify continued stereotactic compatibility.