1 The ability to monitor, record, and interpret single-cell electrical activity within the brain developed over a period of more than 200 years. During this time, the tenacity of a handful of motivated researchers pushed to the limits the technology of their day. Often, the desire to advance in these fields “mothered” the invention and development of new technologies and equipment. This drive for progress within the fields of electricity, neuroscience, and stereotactic neurosurgery has served to make intraoperative single-cell microrecording a routine part of many modern functional neurosurgical procedures. The work of Luigi Galvani (1737–1798) on the existence of electricity intrinsic to living organisms prompted the development of the electric battery by Alessandro Volta (1745–1827) in 1799. The history of electricity and that of the neurosciences have been closely interwoven ever since.1 Although the electric nature of neural impulses was appreciated and, indeed, for many years the isolated nerve-muscle preparation of the frog was exploited as the most sensitive indicator for short-lasting electric pulses, efforts concentrating on understanding the basic character of what would become known as the action potential were to continue another 150 years. Emil du Bois-Reymond (1818–1896) and later Ludimar Hermann (1838–1914) were able to determine that the electric activity of excitable cells occurs in the form of brief electric discharges, but the technology of the day did not lend itself to accurate or direct measurements of the impulse. Undoubtedly the lack of sensitive instruments to measure and display electrical activity was largely responsible for the “delay” by many years of this historical milestone of neurophysiology. In the early 1920s it was these very dilemmas facing neuroscience that prompted the invention of the cathode ray oscillograph2–4 and the vacuum tube amplifier5–7 that finally enabled action potentials to be characterized and accurately and directly measured.8 The British physiologist and Nobel laureate Lord Edgar Adrian Douglas, together with Yngve Zotterman, is credited as having been the first to record the electrical responses of single neurons. In their 1926 report, they described the response of single sensory end-organs to tension.9 They had reduced frog muscle to a single fiber supplied by a single neurone, stretched the muscle fiber, and attempted to record impulses. Ida Henrietta Hyde (1857–1945), a true pioneer among women scientists of the 19th century, invented the intracellular microelectrode in 1921. This milestone undoubtedly revolutionized the study of neurophysiology.10 Single cells could now be stimulated, injected, and recorded from. The utilization of microelectrodes led to a golden age of neurophysiological discoveries from the 1930s to the 1950s, when basic principles of nerve and brain function now described in textbooks were first revealed. In 1936, J. Z. Young discovered the giant axon of the squid (Loligo forbesi).11 For the first time, intracellular phenomena of an excitable cell could be studied. Two groups independently but in parallel exploited this finding. Alan Hodgkin, who had been Adrian’s student at Cambridge University, together with Andrew Huxley, used 50 μ intra- and extracellular glass electrodes filled with seawater to impale 700 μ squid giant axons and record action potentials directly across the membrane.12 Simultaneously, Kenneth Cole (who had introduced Hodgkin to the squid axon) and Howard Curtis achieved the same feat.13,14 Although this work formed the basis of the “sodium theory” of the action potential,15 there was a need to generalize beyond squid axons, to less exotic excitable cells. Judith Graham and Gilbert Ling, both working in Ralph Gerard’s laboratory at the University of Chicago in the 1940s, had investigated the biochemical maintenance of the muscle membrane potential.16,17 To facilitate this work, they had refined the design of intracellular glass electrodes to achieve tip sizes of less than 1 μ. These electrodes caused negligible membrane injury; however, they had an extremely high resistance, making them inappropriate for the recording of action potentials. When Hodgkin visited Chicago in 1948, Ling taught him how to pull and fill these microelectrodes. Back in Cambridge, together with Bill Nastuk, Hodgkin modified the technique to lower the electrode resistance and reduce distortion of the action potential. By the end of that year, they had characterized the muscle action potential.18,19 Building on work performed in the 1930s by Keffer Hartline at the University of Pennsylvania,20 work that had been limited by the lack of solid-state electronics, Ragnar Granit now used microelectrodes to make the first electrical recordings of light responses from individual cells of the vertebrate retina in 1947.21 Hartline and Granit jointly shared the Nobel Prize in physiology and medicine for this work in 1967. Intracellular glass microelectrode technology continued apace, so that by 1967, an entire international meeting held in Montreal, attended by more than 200 scientists, was dedicated to the discussion of developments in this field.22 However, for studying neural physiology in intact anesthetized animals, glass microelectrodes had certain intrinsic disadvantages. These included the fragility of the electrode, the damage done by impaling the cell, and the movement associated with arterial and respiratory pulsations. Thus, for recording in intact animals, attention had returned to using extracellular, nonglass microelectrodes, which had been used for the study of isolated excitable cells for many years. Stephen Kuffler’s pioneering work on the synapse, for example, had used glass-insulated platinum or silver electrodes, previously described by Eccles23 for localized extracellular recordings from single skeletal muscle fibers.24–26 These, too, had disadvantages, such as “multiunit” recording, making it difficult to differentiate single-unit spikes above the neural “noise.” Other, more technical considerations included stiffness, type of insulation, and ease of manufacture. The success of David Hubel and Torsten Wiesel, students and colleagues of Kuffler, in elucidating the function of the visual cortex in the 1950s,27 was dependent on their development of the tungsten microelectrode for recording from single cortical units.28 Techniques for mechanically advancing microelectrodes29,30 and for recording from single and multiple31–33 subcortical units in moving animals during learning, motor performances, and varying states of sleep and arousal were described.34–36 Electrodes manufactured from stainless steel,37 silver coated with platinum,38 elgiloy (a cobalt-chromium-nickel alloy),39 iridium, and platinum-iridium40 were described, each with its own characteristics. Varnish, glass, polyethylene, tygon, formvar, thermobond M-472, parylene, and other materials have been used as insulation. Today the neurophysiologist has all of these options commercially available at modest cost, such that the desired tip size, resistance, and other characteristics can be selected.41 Although the concept of anatomical localization of cortical function had been discussed for many years, the first unequivocal proof came with Paul Broca’s 1861 report associating articulate speech with the frontal lobe.42 Electrical stimulation of the cortex could have occurred as early as 1800 with Volta’s description of the electrolytic battery43; however, primitive cortical mapping did not occur until 1870.44 Eduard Hitzig and Gustav Fritsch, whose interest had been sparked by their earlier observations in the 1860s of head-injured soldiers,45 exposed the cortex of a dog brain to galvanic currents and observed for muscle twitches. These experiments were repeated in monkeys by David Ferrier in 187546 and by Sir Victor Horsley in the 1880s.47,48 Together, these observations provided the first evidence of a motor homunculus. At about the same time, Roberts Bartholow, a professor of medicine from Cincinnati, Ohio, stimulated the cortex of a patient by the name of Mary Rafferty.49 The patient reported contralateral tingling sensations, and Bartholow observed contralateral movements in response to the cortical stimulation behind and in front of the rolandic fissure, respectively. Bartholow was severely criticized for conducting these experiments on a human being, even though he explained that they were done with her approval and that the patient’s brain had been exposed by a rodent ulcer and abscess. Richard Caton, who had been a medical student with Ferrier in Edinburgh, was probably the first, in 1875, to record the spontaneous electrical activity of the brain.50 His research was based on Ferrier’s descriptions of the effects of stimulation and ablation of discrete areas of the cortex. By applying electrodes to the scalp and directly to the brain surface in rabbits and monkeys, he noted that the electrical changes taking place in the brain, measured with a sensitive galvanometer, varied in location with the specific peripheral stimuli he was using. This research was undoubtedly the groundwork for evoked potentials51 and for electroencephalography, and was acknowledged as such by Hans Berger in 1929.52 In 1892, Ransom stimulated the brain of an awake epileptic patient with a pair of electrodes introduced through the scalp and a previous trephine of the skull.53 He obtained both motor and sensory responses. The following year, Charles Dana reported a similar observation in a patient with chorea.54 It is likely, however, that Horsley had already utilized intraoperative cortical stimulation for localizing foci during epilepsy surgery as early as 1886.47 In 1908, Harvey Cushing observed a “sensory fit” while operating on a subcortical cystic tumor in an awake patient. He subsequently reported two other patients in whom he had intraoperatively “mapped” the sensory cortex using electrical stimulation.55 Otrid Foerster (1873–1941) and Wilder Penfield (1891–1976) confirmed these findings some 25 years later by systematically mapping sensory and motor responses to direct cortical stimulation in awake patients undergoing surgery for epilepsy56–59 These and many other illustrious scientists, neurologists, and neurosurgeons contributed to building our understanding of the motor and sensory cortices.60 However, it was not until the era of microrecording that precise cortical and subcortical localization of function and a clearer concept of cortical physiology were revealed. Early pioneers in this field were Vernon Mountcastle, David Hubel, and Torsten Wiesel. Building on work by Adrian61 utilizing evoked potentials to map the contralateral cortical sensory homunculus in cats, rabbits, and monkeys, Mountcastle used semimicroelectrodes (insulated steel electrode, 0.4 mm shaft, 40 μ tip) and the same evoked potentials technique to explore the sensory representation within the thalamus of the cat and the monkey.62,63 Subsequently, using microrecording in cats, he reported on the vertically orientated columns of sensory function in the postcentral gyrus.64 Discrete cortical columns in the visual cortex were described by Hubel and Wiesel in the cat and monkey.27 Many other aspects of cortical function have since been elucidated with the aid of microrecording and microelectrode stimulation.60 Hans Berger made the first human electroencephalogram (EEG) recording in 1924 using scalp electrodes in patients with cranial bone deficits. However, in 1931, concerned about the limitations of such recordings for purposes of localization of deep foci, he began introducing intracerebral “depth” electrodes for subcortical recordings.65 Subsequently, many groups introduced electrodes deep within the brain tissue, some for localizing epileptic foci,66–80 others for exploring possible therapeutic avenues in psychotic patients,81–85 and others for localizing other pathology, such as tumors.86 Various materials, designs, and sizes of electrode were used.78 Silver and copper were especially popular; however, when these were found to provoke inflammatory reactions and necrosis in the brains of cats,87 they were abandoned in favor of stainless steel. Single needles, double needles, concentric electrodes, and multielectrode needles were but some of a large number of different designs.78 Although the gauge of the wires used as depth electrodes steadily decreased, very small wires were fragile and associated with high impedances. The 40 (77.5 μ) or 42 (62. 5 μ) gauge size became the optimum. Introducing electrodes into neurosurgical patients undergoing ventriculography for a variety of reasons, Williams compared EEG recordings from subcortical structures, probably including the thalamus, with cortical recordings.88 He concluded that “with the methods so far available, exploration of the electrical activity of the basal grey matter has no immediate application to the routine investigation of organic brain disorder.” In 1947, human stereotactic surgery, primarily for the management of “extrapyramidal” motor disorders, began. Pioneering this work were Spiegel and Wycis in the United States, Leksell in Sweden, Riechert and Hassler in Germany, Talairach in France, and Narabayashi in Japan. Although it is clear that in their earliest human stereotactic procedures for extrapyramidal disorders and convulsive disorders, Spiegel and Wycis made electrical recordings, these observations were only published later.89,90 They used a grounded copper Faraday cage to shield the patient’s head in the electrically noisy operating room of those days. Electrodes were thin silver wires with ball tips introduced in an insulating polyethylene sheath. Stimulation and recording were routinely used in an attempt to both identify the physiological location of the electrode and to replicate laboratory studies to study the pathophysiology of the diseases being treated.91 However, depth recordings were similar to the scalp EEGs that were simultaneously recorded, and no single units could be identified with this technology. Wetzel and Snider published the first report, in 1958, using electrical recordings during movement disorder surgery with the stated purpose of physiologically refining location.92 Their techniques were somewhat crude even by their own-day standards, and it is doubtful whether their experience really helped them refine lesion location. They described having used “thin stainless-steel Steinman pins or nichrome wire” for electrodes, which did not represent any advance over the stainless steel wire that had already been introduced some years previously for EEG work.78 It is unclear whether any of their electrodes were actually introduced into the deep nuclei of the brain, and indeed their published recordings were identical to EEG traces. Evoked potential recording within subcortical structures was more successful. Jouvet, in a heterogeneous group of neurosurgical patients, had succeeded in recording visual and somasthetic evoked potentials.93,94 In 1960, Ervin and Mark95 reported on a series of patients in whom they had performed a thalamotomy for terminal head and neck pain. For the first time, they made use of evoked potentials to verify the position of their electrode within the sensory thalamus. The electrodes used were bipolar macroelectrodes 1.6 mm in diameter. Following the experience of Heath,82 they knew they could safely implant electrodes for weeks or months at a time. They exploited this time to stimulate and record from the sensory thalamus prior to making a destructive lesion. Their setup allowed them to record “neural noise” but no single units. A significant step forward was taken by Denise Albe-Fessard and her colleagues in 1961.96 They described using low-impedance, concentric bipolar microelectrodes with a 30 to 50 μ tip to record from and differentiate the various thalamic nuclei and the internal capsule. They advanced the electrode through the diencephalon along an oblique posterior-anterior approach used by Gerard Guiot et al,97 in whose department these studies were conducted. Initially, they too succeeded in recording field potentials related to somatosensory stimulation. The neural noise that they observed with what would now be considered to have been semimicroelectrodes allowed them to propose98 a contralateral somatotopic thalamic arrangement similar to that described in animals by Mountcastle and Henneman 10 years previously62,63 and by their own group in cats and monkeys.99,100 Subsequent work by Albe-Fessard’s group at Hôpital Foch in Paris,98,101–107 by Gaze and colleagues in Britain,108 and by Hardy’s group in Canada,109–111 now using true microelectrodes, provided more precise anatomophysiological details and drew attention to the presence of rhythmic cellular discharges, more or less synchronous to the parkinsonian tremor, in the therapeutic ventrobasal target area, so-called tremor cells. Microrecording was thus also instrumental in effecting change. In the early days of surgery for Parkinson’s disease (PD), most focus was on the relief of tremor. Hassler, whose classification of the thalamus is still popular, had, by 1965, performed almost 2000 surgeries for PD, but he still believed that the Vop nucleus was the best target to ablate for the relief of tremor.112 It was not until after the evidence presented by Guiot’s group from their intraoperative microrecording98 showing most tremorigenic cells to lie somewhat more posteriorly, in the Vim, that Hassler reconsidered.113 Bertrand and his coworkers114 in Montreal used a high-impedance curved tungsten microelectrode, which would protrude obliquely from a side opening of the introduced sheath, and a low frontal trajectory. They identified different sensory modality–sensitive cells and contributed greatly to the understanding of the somatotopic and stratified sensory organization at the transition between the motor and sensory thalamus.114 By the late 1960s, intraoperative microrecording had become a routine part of stereotactic movement disorder surgery. The human thalamus had been explored with microelectrodes in hundreds, possibly thousands, of patients, and good correlation between the anatomy and physiology had been realized.115 Marg116 described a refined design of Hubel’s tungsten electrode for use in human microrecording made to be more flexible and resistant to wear. However, the introduction of L-dopa in 1967 had such an impact on the management of PD that stereotactic surgery for PD ground to a halt in all but a few centers around the world117–120 for almost 20 years. In those centers that sustained functional neurosurgery programs for movement disorders and for other indications such as pain, microrecording research continued. Several events brought about a dramatic change. The limitations and complications of L-dopa therapy, specifically the drug-induced dyskinesias, became appreciated. In the late 1980s, Laitinen and colleagues reexplored the pallidal locale of Leksell as a therapeutic target for the alleviation of parkinsonian symptoms.121,122 Their findings sparked worldwide interest in pallidotomy. New microrecording-based research in primate models of PD focused interest on the subthalamic nucleus as a potential therapeutic target.123–126 Magnetic resonance (MR) imaging and surgical computerized navigation enabled more accurate stereotactic targeting. This rebirth of movement disorder surgery has been accompanied by renewed interest in the utility of microrecording. Dedicated microrecording units with purpose-designed computer programs have been marketed by several manufacturers (Chapter 3), making microrecording more user-friendly for those without a background in microelectronics. Controversies over the necessity of microrecording as a targeting tool have been debated,127 contrasting the risks of microelectrode recording (MER) against its advantages and attempting to compare the outcomes of patients operated with and without MER. Attempts to collect microrecording data into a reference database and set an “industry standard” for recording techniques have been made. Integration of optical tracking neuronavigation and patient-specific atlases have been described to allow for “real-time” correlation of anatomical and physiological information during surgery. Although no historical review of a developing science might be considered complete without some speculation concerning the future, this is dealt with elsewhere in this book. Whether microrecording remains the gold standard for physiological targeting will depend on the progress of noninvasive functional imaging. Meanwhile, it remains an elegant and accurate clinical and research tool.
History of Electrophysiological Recording for Functional Neurosurgery
ZVI ISRAEL AND MICHAEL SCHULDER
Electricity and Neuroscience
The Action Potential
Development of the Microelectrode
Microelectrode Design
Mapping the Brain
Introducing Electrodes into the Human Brain
Recording and Stimulating during Movement Disorder Surgery
REFERENCES