A Short History of
Robert Levy, MD
Adapted with permission from:
Patrick J. Kelly, MD
"Introduction and Historical Aspects"
Philadelphia: W.B. Saunders Company (1991)
Philip L. Gildenberg, MD, PhD "Stereotactic Surgery: Present and Past" Stereotactic Neurosurgery (Editor: M. Peter Heilbrun) Baltimore: Williams and Wilkins (1988)
DEVELOPMENT OF EXPERIMENTAL STEREOTACTIC SYSTEMS (1873-1932)
The first recorded use of guided probes in the neurophysiology laboratory was in 1873, when Dittmar in Ludwig's laboratory applied a uniquely designed guiding device to the medulla oblongata. Zernov, a Russian anatomist, subsequently developed the encephalometer, and arc-based guiding device based on polar coordinates that was designed for anatomic operations on the human brain. This device was actually used clinically on at least three occasions.
The most definitive description of the principles and device for stereotaxis is usually credited to Robert Henry Clarke and Victory Horsley, in their detailing and design of an apparatus to study cerebellar function in the monkey. In 1906 they wrote that "by this means every cubic millimeter of the brain could be studied and recorded." A more complete description of the stereotactic instrument, atlas, and methods was reported in their classic paper of 1908. The Horsley-Clarke Device was based on the reproducibility of the relationships between landmarks on the skull (external auditory canals, inferior orbital rims, midline) and anatomical structures within the brain of the experimental animal. The cranial fixation points established the baselines of a three-dimensional Cartesian stereotactic coordinate system.
See Fig. 1 - Horsley and Clarke's original animal stereotactic apparatus
Clarke suggested to Sir Victor Horsley that the stereotactic method could be useful in human neurosurgery. However, Horsley scoffed at this idea and reported this ended their well-known association.. Nevertheless, Clarke submitted a patent application for a human stereotactic instrument in 1912. Extension of the technique for use in patients was fraught with great difficulty, however, due to the great variability between skull landmarks and cerebral structures in the human.
Aubrey Mussen, a physiologist from the Montreal Neurological Institute, commissioned a London instrument maker to construct a human stereotactic frame in 1918. This instrument was largely a modification of the original Horsley-Clarke apparatus. It attached to the patient's head by ear bars, which were to be inserted into the external auditory canals, and a clamp fixed to the infraorbital ridge. Mussen also developed a human stereotactic atlas based on cranial landmarks that was similar to Clarke's animal stereotactic atlas. However, Mussen's stereotactic instrument was never used clinically as he was never able to convince a neurosurgeon to use the device.
It was not until 1932 that the Horsley-Clarke apparatus was copied from illustration in Horsley and Clarke's article and rebuilt at Northwestern University Medical School. Ranson and Ingram then used this device in their classic studies on the reticular formation, midbrain, and hypothalamus. Shortly thereafter, the Horsley-Clarke apparatus was again reproduced at the University of Chicago. Sugar and Gerard then used it in a series of experiments that systematically studied electrical activity in the brains of cats and the effects of anoxia on these brain potentials. Following this, the Horsley-Clarke apparatus was duplicated in many other medical school laboratories, and modifications of the original were developed for a variety of neurophysologic and anatomic investigations.
EARLY HUMAN STEREOTACTIC SYSTEMS IN THE UNITED STATES (1933 - 1960)
Although the Horsley-Clarke coordinate system, based on cranial landmarks, was reasonably accurate for the reproducible localization of subcortical structures in small animals, anatomic variability between individual specimens troubled many early investigators including Clarke himself. Too much spatial variability existed between individual human brains to make reliance on bony reference planes reproducible and safe. Cranial landmarks did, however, reliably indicate the position of structures known to lie reasonably close to that landmark. Therefore in 1933, Kirschner was able to develop a stereotactic instrument for thermal coagulation of the human gasserian ganglion to treat trigeminal neuralgia. His method used the foramen ovale as a reference structure from which the location of the gasserian ganglion could be inferred.
Variability in the spatial relationship between subcortical structures and cranial landmarks was probably the most important reason that human stereotaxis was not practical until intracerebral reference points could be visualized and correlated to subcortical target structures. Ventriculography had been developed by Walter Dandy in the 1920s. Nevertheless, it was not until 25 years later that Spiegel and Wycis considered using positive contract ventriculography and the pineal body to localize intracranial targets. The first stereotactic instrument used routinely in human subcortical surgery was described by these investigators in 1946. Their original stereoencephalatome is now in the Smithsonian Institution.
The Spiegel and Wycis design was centered on a plaster cap that was fitted to each individual patient. A head ring was suspended from the plaster cap, and an electrode carrier was mounted to the head ring. Their unique contribution was the idea of relating anatomical targets to landmarks within the brain itself - hence the name "stereoencephalotomy". The landmarks that were the basis for Spiegel and Wycis' original human stereotactic atlas were the pineal gland and the foramen of Monro, visualized by preoperative or intraoperative pneumoencephalograms. With the later advent of positive contrast agents for ventriculography, the anterior and posterior commissures and the connecting intercommissural line became the most commonly used internal cerebral landmarks.
See Fig. 2 - Two views of the original Spiegel and Wycis Model I instrument adapted to the human head from the Horsley-Clarke stereoencephalotome used in the laboratory on small animals. The instrument was custom fitted to each patient and head in place with a plaster cast.
The first human stereotactic instrument was developed for the coagulation of the dorsal median nucleus of the thalamus in patients with severe psychiatric illness in order to provide a less traumatic alternative to the more destructive frontal lobotomies that were popular at that time. However, in this original publication, Spiegel and Wycis proposed other applications for stereotactic technique beyond psychosurgery: they also suggested the use of stereotaxis for interruption of pain pathways, for movement disorders, and for the "withdrawal of fluid from pathological cavities, cystic tumors."
In order to localized subcortical structures in the treatment of pain and movement disorders, Spiegel and Wycis devised a human stereotactic atlas that was based on ventricular system landmarks. This consisted of a series of photographed coronal brain slices that had been cut at constant intervals in relationship to the posterior commissure and the midline. These coronal slices were photographed with a millimeter reference grid around the borders of each coronal section. Utilizing the reference grid, the surgeon could simply measure height and laterality coordinates of a subcortical target structure identified on a particular coronal section having a known distance from the posterior commissure. Coordinates for many target structures could be derived by this means, and selective neuroablative procedures became practical. These basic studies provided the localization methods necessary to proceed with pallidoanstoomies in the treatment of movement disorders and messencephalotomy for chronic pain, which were reported by Spiegel and Wycis in the 1950s and subsequently the stereotactic aspiration of cystic tumors and their treatment by the stereotactic instillation of radioactive phosphorus-32.
EUROPEAN STEREOTACTIC SYSTEMS (1949 - 1960)
Encouraged by Spiegel and Wycis, Lars Leksell in Stockholm developed his own stereotactic instrument in 1949. This instrument was different in principle from that of Spiegel and Wycis and utilized a new concept: the arc-quadrant. Leksell's device consisted of a fixation device that was fixed to the patient's skull and a movable arc-quadrant that was attached to the fixation device. The arc-quadrant was moved so that the arc canter was positioned at the desired intracranial target point. Subsequently, Leksell's instrument underwent modification: the support structure was changed to a cuboidal frame but the arc-quadrant principle remained the same. As had Spiegel and Wycis, Leksell also described the use of ventriculography for delineating subcortical targets.
See Fig. 3 - First version of Leksell stereotactic instrument.
Jean Talairach in Paris was aware of the work of Spiegel and Wycis. Independently, however he also developed a novel stereotactic instrument which was reported in 1949. Hacaen, with Talairach and others, also reported in 1949 the clinical use of stereotactic thalamotomy and anterior capsulotomy in the treatment of thalamic pain and mental disease. The Talairach system consisted of a basic unit that was fixed to the skull and a double grid that attached to that base unit. Collimated radiographs obtained with the grid in place demonstrated superimposition of the holes in the grid over a positive contrast ventriculogram. Talairach also introduced the concept of teleradiography. In a teleradiographic system, tightly collimated x-ray beams and long x-ray tube-to-patient distances are employed in order to reduce magnification distortion of the x-ray images. Thus precise measurement could be made directly from radiographs obtained during the procedure. In addition, the Talairach frame was unique in that it could be removed and replaced on the patient in precisely the same position. Thus radiographs, ventriculograms, and later arteriograms obtained at the procedure could be used to guide subsequent operations.
Following the original descriptions of stereotactic instruments by Spiegel and Wycis, Leksell, and Talairach, others worldwide saw the merits in stereotactic guidance. Most developed their own instruments. In Germany, Riechert and Wolff devised a device similar to that described by Kirschner. This instrument utilized an aiming bow that attached to a circular base ring fixed to the patient's head. The aiming bow could be transposed to a phantom base ring on which the surgeon would set up his target coordinates on a simulator. The base ring of the original instrument was simply strapped onto the patient's head. Subsequent modifications incorporated skull fixation for the base ring. This instrument, with modifications, became known as the Riechert-Mundinger system and has been used in many functional and tumor stereotactic applications.
In Geneva, Monnier devised a stereotactic frame similar, in part, to Talairach's system, and radiologic and electrophysiologic methods for intracranial localization for thalamotomy in patients with chronic pain. Guiot and Brion also developed an instrument specific for stereotactic surgery in movement disorders. This instrument and that modified by Gillingham attached to the midline of the patient's skull and was used to direct a probe through an occipital burr hole through the thalamus to the globus palidus.
TYPES OF STEREOTACTIC FRAME SYSTEMS
Simple Orthogonal Systems
Several approaches have been employed to reach a defined target point in space. The earliest and most straightforward is that in which the probe is directed perpendicular to a square base unit fixed to the skull. These systems provide three degrees of freedom by means of a carriage that moved orthogonally along the base plate or along a bar attached parallel to the base plate of the instrument. Attached to the carriage was a second track that extended across the head frame perpendicularly. The probe holder on this track could be moved from right to left. The height coordinate (superior-inferior) was provided by the probe holder, which inserted the probe to a specific depth.
This orthogonal system was the basis for the Spiegel and Wycis apparatus (as well as the Horsley-Clarke device); in the original model, the electrode carrier was held vertical and moved by a translational system in two dimensions, with a microdrive to advance the electrode in the third. Later models (six Spiegel-Wycis models were eventual developed) incorporated precision lockable hinges to adjust angels in the anteroposterior and lateral planes as well. The Talairach frame is another example of a simple orthogonal system: probes are inserted to a measured depth through coaxial holes in a grid that is mounted on a base plate orthogonal to the x or y axes.
See Fig. 4 - The four basic types of stereotactic apparatus. A. Translation system; B. Arc system; C. Burr hole mounted; D. Interlocking arcs.
Simple orthogonal approaches to subcortical targets were not felt to be desirable in certain human surgical procedures. For instance, a straight orthogonal approach to the target point in ventrolateral (VL) thalamotomy may traverse the motor strip. Therefore, newer instruments included an angular adjustment of the carriages or probe holder so that the probe could be directed at an angle to the three axes of the stereotactic frame. This provided more flexibility in the selection of probe trajectories but created another problem: the mathematics required to calculate the position of the probe were more complicated. It was necessary to incorporate trigonometric functions into the calculations.
Burr Hole-Mounted Systems A burr hole-mounted system provides a limited range of possible intracranial target points with a fixed entry point. For pallidotomy and thalamotomy, these devices were usually fixed into a coronal burr hole. Simple in design, they provided two angular degrees of freedom (angle from the horizontal plane, angel from the vertical plane) and a depth adjustment. The surgeon could place the burr hole over nonessential brain tissue and utilize the instrument to direct the probe to the target point from the fixed entry point at the burr hole.
In practice, these instruments were usually threaded into the burr hole and AP and lateral radiographs were obtained. The surgeon would determine the position of a target point on a positive contrast ventriculogram. On the radiography, a line would be drawn between the target point and the center of the stereotactic probe holder seated in the burr hole through the radiographic image of a protractor on the instrument. The lateral and frontal angular settings required to reach the target were simply read off of the radiographic image of the protractor. Theoretically, the probe depth could be measured off the radiograph, and corrected for radiographic magnification and the angular settings of the instrument. However, the trigonometric manipulations required to execute this task were relatively complex when performed manually. More commonly, therefore, the depth settings on this instruments were actually determined intraoperatively by trial and error, checking each probe depth adjustment with multiple radiographs.
The major disadvantage of these systems lay in the fact that probe position in the depths of the brain was dependent on the angular settings of the burr hole-mounted apparatus. A fraction of a degree in error could result in a probe being millimeters in error when advanced to its depth. Because of this inaccuracy, burr hole-mounted systems were never really embraced by serious stereotacticians but were popular during the time when many neurosurgeons with little stereotactic background were doing a large number of thalamotomies for Parkinson's disease.
Newer modifications of this old concept have appeared with the advent of CT scanning. A recent modification of this system mounts the fulcrum more securely to a base plate attached to the skull, and a phantom is used to aim the device. It is used primarily for CT-stereotactic surgery.
A more accurate and mathematically simple solution for accessing a target from nonorthogonal trajectories was provided by the arc-quadrant stereotactic frame. This concept was originally described by Leksell in 1949. In principle, probes are directed perpendicular to the tangent of an arc (which rotates about the vertical axis) and a quadrant (which rotates about the horizontal axis). The probe, directed to a depth equal to the radius of the sphere defined by the arc-quadrant, will always arrive at the center or focal point of that sphere.
The Todd-Wells apparatus also is an arc system, but it is designed reciprocally so that the arc system is fixed and the patient's had is moved in a controlled fashion to align it with the target point. The Riechert system, and its later modifications with Mundinger, is likewise an arc system, but instead of the target being at the center of the arc, the electrode carrier is offset, and it is desirable to use a phantom (or later a computer) to adjust the apparatus. The phantom mechanically simulates the placement of the arc system on the head and the coordinates of the target.
Another solution that provides mathematically uncomplicated yet unlimited probe trajectories to a target point is the arc-phantom concept. This concept was introduced by Clarke in 1920, used by Kirschner in 1933, and also described by Riechert and Wolff. An aiming bow attaches to the headring, which is fixed to the patient's skull, and can be transferred to a similar ring that contains a simulated target. In this system, the phantom target is moved on the simulator to the x, y, and z coordinates calculated from radiographs, positive contrast ventriculography, or imaging studies. After adjusting the probe holder on the aiming bow so that the probe touches the desired target on the phantom, the transferable aiming bow is moved from the phantom base ring to the base ring on the patient. The probe is then lowered to the determined depth in order to reach the target point deep in the patient's brain.
In the arc-phantom system the probe trajectory is independent of the arc. Lateral targets, which are difficult to reach with arc-quadrant systems, may be attainable with the arc-phantom system. However, four angles must be set correctly for the probe to arrive at the desired target. Small errors in any of these angular settings can result in significant errors in the position of the probe tip.
The Brown-Robbers-Wells apparatus, a recent modification of the arc-phantom system, was designed primarily for CT and MRI-stereotactic surgery and consists of interlocking arcs. Because of the complexity of adjusting the individual arcs to define a specific trajectory, it is necessary to use a computer to define the coordinates of the target point and the adjustment necessary to reach the target.
In order to effectively us a stereotactic apparatus, especially for functional stereotactic surgery, it is necessary to know the relationship between the landmarks and any given anatomical target and to know how much these measurements vary in a patient population. Horsley and Clarke included information about a monkey atlas in their historical publication. Likewise, Spiegel and Wycis developed a human stereotactic atlas, which they published in 1952, based on the same principle, that is a series of precise brain sections as measured intervals containing a grid system to relate the coordinates of any point to the intracerebral landmarks. This was followed soon after by a number of human stereotactic atlases. Schaltenbrand and Bailey's atlas contained transparent pages on which the anatomic nuclei were drawn, overlying stained sections of the brain, with particular emphasis on the area around the thalamus. The Talairach atlas contains information about the location of blood vessels and was designed with an accent on epilepsy surgery. Later atlases by Andrew and Watkins and Van Buren and Borke contained greatly enlarged drawings defining the relationships of subnuclei, particularly those of the thalamus. Afshar's atlas concerns the brainstem and cerebellar nuclei, and Tasker's atlas contains much physiological as well as anatomical information for localization.
See Fig. 5 - Each page in a stereotactic atlas represents a brain slice taken at a measured position within the brain.
The use of computers in the past 10 years has greatly affected the development of stereotactic atlases. Computer-based stereotactic atlases presently consist of graphic illustrations derived from published atlases. The computer atlas representation may be purposefully distorted to fit the measurement of the individual patient's landmarks or nuclei, as demonstrated on x-ray, CT or MRI scan, or magnified and manipulated in the operating room to provide information and views helpful to the surgeon. It may provide a repository for information obtained in the operating room by physiological testing or stimulation, or may be correlated with results.
THE "STEREOTACTIC SOCIETY"
The increasing activity in the field required a forum to disseminate information about new procedure, techniques, and results to a growing number of scientist-neurosurgeons interested in stereotactic neurosurgery. Confina Neurolgica, founded in 1938 and edited by Spiegel, became the major source of information in the field of stereotactic surgery. In 1975, the title was changed to Applied Neurophysiology, when Gildenberg became editor. The subtitle, Journal of Stereotactic and Functional Neurosurgery, was added in 1985 and in 1989, the title was shortened to the Journal of Stereotactic and Functional Neurosurgery. Acta Neurochirurgica has become the major European publication concerned with stereotaxis. In it are published the proceedings of the meetings of the European Society for Stereotactic and Functional Neurosurgery.
The increasing interest in this field led ot the formation of a new society, which was established in conjunction with the First International Symposium on Stereoencephalotomy held in Philadelphia in October 1961. The International Society for Research in Stereoencephalotomy held its Section International Symposium partly in Copenhagen and partly in Vienna in 1965. The Third International Symposium was held in Madrid in 1967, the Fourth in New York in 1969, and the Fifth in Freiburg in 1970.
The Sixth International Symposium held in Tokyo in 1973 marked a turning point. There was considerable discussion as to whether the acceptable spelling should be "stereotactic" or "steroetaxic". "Stereo-" is from the Greek root meaning "three-dimensional", and it was agreed to be appropriate. By majority vote, "stereotactic", combining the Latin root "to touch" rather than "steroetaxic" from the Greek root for an "arrangement" was accepted as the official spelling, since surgery involves introducing a probe to the target rather than merely defining the relationships.
At the same meeting, it was also agreed to change the name from International Society for Research in Stereoencephalotomy to the World Society for Stereotactic and Functional Neurosurgery, indicating that the members were interested in all aspects of functional neurosurgery, i.e., surgery designed to change the function of the nervous system, in addition to just stereotactic techniques, thereby including epilepsy surgery and the new field of chronic stimulation of the spinal cord. Shortly thereafter, the various branches were renamed as the American, European, and Japanese Societies for Stereotactic and Functional Neurosurgery.
The Seventh meeting was held in Sao Paulo in 1977, the Eighth in 1981 in Zurich, the Ninth in 1985 in Toronto, and the Tenth meeting, representing 40 years of progress, in Tokyo in 1989. In addition, the American and European Societies meet in alternate years in which there is no World Society meeting, and the Japanese Society meets annually. The proceedings of all these meetings, with the exception of the independent European Society meetings, are published in Confina Neurologica or Applied Neurophysiology.
The AANS Archives would like to thank the following individuals and corporations for their assistance in the development and completion of this brochure:
Philip L. Gildenberg, MD, PhD