From stereotactic frames to image-guided surgery
- B. Lombard
In the history of medicine, the advent of a new technology has often heralded a leap forward, either by enhancing the physician’s perceptual acuity, revealing new diagnoses, or by placing a new tool in the physician’s hand, improving the act of treatment. Thus it was, for example, with the microscope, electrocardiograph and X-rays, each of which gave birth to major medical specialties.
Computer-assisted surgery emerged from the convergence of new needs and new means, but can also be seen as the fulfillment of one of surgery’s oldest dreams: a means of guidance and spatial location that would improve surgical accuracy. It thus seems logical to consider it as the prolongation of stereotactic surgery, which was developed well before the computer age: the first stereotactic frames appeared at the outset of the last century.
The beginnings of stereotactic frame surgery
In 1905 Clarke and Horsley 1
1. Sir Victor Horsley (1857–1916), considered as the first neurosurgeon in Great Britain [ ]. Robert Henry Clarke (1850–1926) was an English mathematician, whose work sought to apply geometry to the study of the brain.
developed the first modern stereotactic frame, and in 1908 published a method of rectilinear topography applied to the Rhesus macaque brain ( Figure 1.1 ). Its complexity foretold the general problem of referencing and registration which are dealt with in the next chapter. What they understood was that there is no systematic relation between the anatomy of the skull and the location of the brain structures it contains, and that the only way of knowing where to position an instrument to reach a given brain structure is to draw up a 3-dimensional map [ ]. Their work included designing a dedicated saw to obtain slices of frozen rat and monkey brain with fixed thickness (2 mm), and mapping their structures onto a Cartesian space – something that CT-scanners would later achieve completely non-invasively. Once the anatomic map was complete, the frame enabled lesion-inducing electric current to be applied to different points within the brain via insulated guided needles [ ].
The Clarke-Horsley frame was precise enough for neurophysiological studies, but was too complex to set up and too invasive to be of clinical use in the field of neurosurgery that was emerging with the work of Cushing.
The Canadian neurophysiologist Mussen modified the Clarke-Horsley frame in 1918, making it less invasive, set only on the external auditory canals and inferior orbital edges. In 1947, Spiegel came back to Clarke and Horsley’s work, seeking an alternative to the appalling frontal lobotomy practiced by Egas Moniz [ ], which left thousands of American children lifeless idiots for no other reason than that they were considered troublesome. Mussen delivered electric current to induce lesions in the medial nucleus and thalamus, using a slightly modified version of the stereotactic frame. His results were not much better than Moniz’s 2
2. Just 6 years after being awarded one of the most controversial Nobel prizes ever, Moniz was killed in a revenge attack by one of his patients.
, but did lead to less aggressive, less hemorrhagic and less lethal forms of psychosurgery.In Paris in 1949, Talairach published his renowned stereotactic atlas, based on serial brain dissection meticulously mapped onto 3D Cartesian coordinates; more than half a century later, it is still a reference, sometimes integrated in present-day neurosurgical navigation systems.
Cranial and brain radiology
All of the above, however, lacked the immense contribution of medical imaging, and was founded purely on statistical analysis of human brain anatomy from postmortem dissections. In 1914, Cushing was probably the first to realize the considerable interest of craniofacial radiography. To determine a strategy for approaching the sella turcica for hypophysectomy, he developed a method of measurement on radiographs at precise incidences [ ]. Soon after, he left for France, to serve on the battlefields of the Great War, where he operated on many cranial and brain injuries. Marie Curie’s radiology vans, taking X-ray equipment to front-line hospitals, created a new need for means of combining X-ray and 3D location. This led to the development of the Hirtz compass and of more or less complex and more or less effective systems for localizing shrapnel inside the head ( Figure 1.2 ).
Cushing’s method was gradually refined, with a multitude of indices, landmarks and ratios to guide radiologists, who were evolving into neuroradiologists, while neurologists were turning into neurosurgeons, in a distinct specialization.
The idea of using a computer to pilot an X-ray generator dates back to the early 1960s with the Californian engineer and neurologist Oldendorf ( Figure 1.3 ), who published a report on the inadequacies of traditional imaging of the brain as compared to exploration of other organs, highlighting the risks inherent to iodine or gas-based encephalography and brain angiography 3
3. William H. Olendorf (1925–1992). Despite his seminal contribution to medical imaging, he was not awarded the Nobel prize which he should have shared with Hounsfield and Cormack, because he was seen as not a fundamental but an applied scientist, being a neurologist, according to Nobel committee’s tradition.
. He suggested using a computer to pilot precise rotation of a radio-tomograph and record the data. By way of demonstration, in 1961 he constructed a system at his home which produced a rudimentary tomodensitometric image by means of numerical calculation, using his children’s train-set to get a linear displacement of a phantom and the motor of a record-player to rotate a source of gamma-rays [ ], easier to use than X-rays, thereby demonstrating the feasibility of creating an image. However, the manufacturers he contacted with his idea failed to see the potential; moreover, computers at that time were rather rudimentary [ ].
Not until 1972 did the tenacity of the inventive self-taught British engineer Godfrey Hounsfield ( Figure 1.4 and 1.5 ) result in the first transverse slice of a patient’s brain, acquired in just 22 minutes on one of the most powerful computers of the time: the ICL 1905, manufactured by EMI (who also produced the Beatles). Nobody was much interested in what Hounsfield was up to, except James Ambrose, a neuroradiologist in London, who loathed having to inflict on his patients pneumo-encephalographies, which were horribly painful and stressful, requiring a voluminous sample of cerebrospinal fluid to be extracted before the patient was seated with the head rotated down for 15 minutes – all that to obtain images that were often uninterpretable. Within 3 years, the Hounsfield CT-scanner had conquered the whole world [ ].
