Neurology. Plain X-Rays. Computed tomography

Following the neuroanatomical discoveries of Englishman Thomas Willis in the seventeenth century and Scotsman Alexander Monro ‘‘Secundus’’ in the eighteenth, and the neurophysiological work of Frenchmen Pierre Flourens, Guillaume Duchenne, and Jean-Martin Charcot in the nineteenth, knowledge of the brain and nervous system advanced significantly.

But rapid progress in the clinical neurological sciences occurred only after discovery of x-rays by the German Wilhelm Roentgen in 1895 and subsequent developments made it possible to create images of living brain and nerve tissue. The diagnostic science of neurological imaging is broadly called neuroradiology.

In chronological order of development, six general neuroradiological techniques have dominated: plain x-ray films, pneumography, radiopaque myelography, cerebral angiography, computed tomography (CT), and magnetic resonance imaging (MRI). The development of the first four techniques is generally complete, but the fifth and sixth are still being refined. Other techniques of neurological examination include echoencepha- lography, electroencephalography, various ultrasound applications, and removal of cerebrospinal fluid (CSF) from subarachnoid space. The era from plain films to cerebral angiography took roughly 75 years, from 1898 to 1973.

Plain X-Rays. Plain x-ray films, available after 1896, were first used extensively for skull pictures in the Spanish- American War. Austrian Artur Schuller, the acknowledged father of neuroradiology, investigated their application to intracranial diagnosis and published his groundbreaking work in 1912. Swede Erik Lysholm refined the photographic method in the 1920s and 1930s. His manual of ‘‘skull tables’’ allowed precise identification of intracranial anatomical relations. Even as recently as the 1960s, physicians preferred plain films for some kinds of neurological diagnosis.

Pneumography/ventriculography. Pneumography, invented by American Walter Dandy in 1918, is the x-ray photography of the skull (pneumoencephalography and pneumoventriculography) or spine (pneumomyelography) after air has been introduced. Dandy observed that abdominal surgical x-rays showed gas as black against the gray of soft tissues and the white of bones, and he had read American William Henry Luckett’s surgical case report of air from the sinuses entering a fractured skull and showing the ventricles and subarachnoid space in an x-ray.

He discovered that if air was introduced through a lumbar puncture needle, a pneumoencephalogram could be produced. His pneumographic procedure was dangerous in cases of brain tumors or papilledema because the increased pressure could herniate the brain stem. For such cases, making holes in the skull and putting the needles directly into the ventricles (ventriculography) was safer.

Formaldehyde, first used as an anatomical fixative in 1893, enabled the pioneer work of Swiss psychiatrist Adolf Meyer in the postmortem confirmation of neuroradiological diagnoses. Meyer discovered that the precise anatomical relations of the brain can be preserved for inspection if formaldehyde is injected into the CSF and allowed to stand 24 hours before autopsy.

By this method, Meyer learned that what often killed brain cancer patients was not the tumors themselves but the herniation of brain tissue through the tentorial notch or into the foramen magnum. Meyer described sideward and downward shifts of the brain tissue in 1920; American Arthur Ecker described upward shifts in 1948. Transtentorial herniation became a common topic of neurosurgical research in the 1950s.

By 1934 Dandy had refined his diagnostic procedure into a standard routine. His patients at Johns Hopkins University would first undergo a general and neurological exam by a surgical resident, a neuro-ophthalmological exam by Frank B. Walsh, and an otological exam by Benjamin M. Volk. The next day Dandy himself would perform ventriculography and make his judgment on the basis of all these reports. Errors were often made, and the mortality rate was high, even for benign intracranial tumors; but at the time this method was the best available in America.

Ventriculographic technique was refined to a higher level in Sweden in the 1930s, primarily by Lysholm’s team directed by Herbert Olivecrona. Pneumography, especially pneumoencephalography, was improved from the 1920s through the 1950s by Germans Otfrid Foerster and Erich Fischer-Brfigge, Americans Leo M. Davidoff and Cornelius G. Dyke, Briton Henry Head, Australian E. Graeme Robertson, and Italian Giovanni Ruggiero.

Radiopaque myelography. Radiopaque myelography, x-rays of the spine with a variety of contrast media, was begun in 1921 by Frenchmen Jean Athanase Sicard and Jacques Forestier. Since then, the quest for safer, more effective, more easily applied contrast media is a major part of the history of neuroradiology.

Cerebral angiography. In 1926 or 1927 Portuguese physician and statesman Antonio Caetano de Egas Moniz invented cerebral angiography, x-rays of the skull after introducing a contrast medium into both carotid arteries. Egas Moniz mounted his camera on the ‘‘radio-carousel’’ invented by his colleague Jose; Pereira Caldas to get a large series of angiograms in rapid succession. His earliest contrast media for cerebral angiography were very dangerous substances such as strontium iodide. A colloidal thorium dioxide solution, Thorotrast, was commonly used after 1929 because it provided better contrast and was somewhat safer.

Figure 3. Cerebral angiogram made by Arthur D. Ecker, M.D., in 1949

Egas Moniz’s method discouraged patients and physicians alike because it required incisions to expose both carotid arteries and left unsightly scars on the neck. The development in 1936 of percutaneous carotid injections of Thorotrast alleviated that difficulty. But Thorotrast is radioactive, and if any of the injected solution fell outside the carotid artery it would cause proliferation of neck tissue with disastrous results. In the 1940s some organic iodides, such as Perabrodil, were found to be safer than Thorotrast. Yet even into the 1950s some physicians argued in favor of Thorotrast and against percutaneous carotid injections.

The U.S. lagged behind Europe, especially Sweden, in the acceptance and development of cerebral angiography. In 1951 Ecker, who had translated Egas Moniz’s method for his own use in the 1930s, published the first American monograph on the topic. The Anglophone world to its disadvantage neglected the German work of Fischer-Brügge and Hermann Coenen in the early 1950s.

Computed tomography. Mathematical equations may sit idle long after their discovery until someone finds practical uses for them. Such was the case with the equations necessary for the development of computed tomography (CT) scans, the greatest single advance in radiology since x-rays. Austrian Johann Radon discovered equations for determining plane functions from line integrals in 1917, but South African physicist Allan Macleod Cormack only learned of them in 1972, nine years after he independently solved that problem and successfully applied it to computerized composite imaging.

When British physicist Godfrey Newbold Hounsfield led the team of radiologists that built the first clinical CT scanner in 1971, he was unaware of either Radon or Cormack. Many improvements of CT have appeared since the 1970s, including positron emission tomography (PET) and single photon emission computed tomography (SPECT).

Magnetic resonance imaging. In the 1990s magnetic resonance imaging (MRI) and PET enabled ultrasophisticated, computer- driven brain mapping projects, such as that directed by Arthur Toga at the University of California at Los Angeles. MRI is safer than PET because it does not expose patients to radiation.

MRI was originally called nuclear magnetic resonance (NMR) because it resonates hydrogen nuclei in the body, but in 1983 the American College of Radiology voted to change the name to avoid misunderstandings within the burgeoning ‘‘No Nukes’’ movement. A refinement of MRI, magnetoencephalography (MEG), measures brain activity directly rather than inferring it from relative oxygen levels, electrical impulses, and other intracranial data.