Read an Excerpt
Looking Inside the Brain
The Power of Neuroimaging
By Denis Le Bihan, Teresa Lavender Fagan PRINCETON UNIVERSITY PRESS
Copyright © 2012 Odile Jacob
All rights reserved.
ISBN: 978-1-4008-5216-1
CHAPTER 1
ELEMENTARY PARTICLES
Before setting out on our discovery of the brain, we must first understand what makes it such a unique organ (figure 1.1). The brain weighs on average close to 1.5 kilograms (3 pounds, for males, a bit less for females) and is made up of two hemispheres, the left and the right, which play somewhat different roles. It is part of a larger whole, the encephalon, which also includes the brainstem, through which neurons travel to communicate with the spinal cord (some neurons that start at the brain and end at the base of the spinal cord are more than a meter long ...). Although relatively small in size, the brainstem is crucial because it gathers together vital smaller elements that regulate our life (sleeping, breathing, heartbeats). This is no doubt why it is very protected and difficult to access, because if it is damaged, the result is often fatal. Alongside the brainstem is the cerebellum (the "little brain," which in fact has little in common with the brain). One of its noticeable roles is to smooth out and coordinate our movements; thanks to the cerebellum we are able to walk straight or play the piano.
The human brain is impressive, perhaps less in its size (the brain of an elephant, whose memory is legendary, is larger, at 4 to 5 kilograms [11 pounds]), than in its complexity: the human brain is made up of a large number of folds and bumps, called sulci and gyri, that are not as developed in other animal species, including the great apes. It is to the French surgeon, Paul Broca, that we owe the fundamental discovery that the two hemispheres are not functionally identical, and that the brain is an organ that, although apparently homogeneous, is made up of regions that have different functional specificities, which is not the case with other such organs (the cells of the liver all do the same work regardless of their location in the organ). This discovery opened the door to modern cerebral physiology and is at the very heart of the goal of neuroimaging: more than just producing images, neuroimaging involves the creation of maps that show the "natural geography" of these regions, and, more important, how they are implicated in the sensory-motor or cognitive functions that they underlie.
Broca's Discoveries
Paul Broca was a surgeon at the Bicêtre hospital (reference 1.1). In 1861 he had a patient, Monsieur Leborgne, who became unintentionally famous. In the hospital M. Leborgne was nicknamed "tan-tan" because he responded to questions (what's your name? what day is it? and so on) with only one syllable, "tan," which in general he repeated twice. This patient had what today is known as aphasia, a sort of "mutism," a disorder affecting the production of language, although he suffered from no paralysis of the "bucco-phonetic" muscles used in speech. After the patient died on April 17 of the same year, Broca did an autopsy and dissection of his brain. He found a lesion on it and drew two major hypotheses from this (figure 1.2), which the next day he presented to the Paris Anthropological Society: the patient's functional disorder must have been owing to the specific localization of the lesion in the brain. That lesion was toward the front, at the base of the third circumvolution of the frontal lobe, in the left hemisphere. Had the lesion been elsewhere, farther back, or in the right hemisphere, M. Leborgne would have perhaps had other symptoms, but he would not have been aphasic. This hypothesis was quickly tested on other patients. Thus was born the principle, which today has been well verified, of a direct link between cerebral localization and function, each cerebral region being associated with a specific function (motor functions, vision, hearing, language, and so on), and this all fitting together on different scales like a set of nesting Russian dolls.
The region that was affected in M. Leborgne's case is today known as Broca's area. Though the role of Broca's area in the production of language is beyond any doubt, we now know that many other cerebral areas (forming a network) are important to language. And inversely, Broca's area is also implicated in other functions. But it remains true that the postulate of regions of the brain being associated with localized functioning is completely established today.
Broca's second great discovery was that the two hemispheres each have their own areas of specialization: language is mainly seated in the left hemisphere (the twenty or so aphasic patients of Broca all had lesions on the left). Until Broca, the two hemispheres were assumed to have identical functions, like our two kidneys or two lungs, which have exactly the same role. For the first time, then, it appeared that the two cerebral hemispheres were not identical, not functionally interchangeable. In around 85 percent of us, the functions tied to language are predominately located in the left hemisphere; for the others, they are in the right hemisphere, and sometimes the dominance is less clear, with the two hemispheres clearly participating to the same degree in the production of language. At what moment, during the millions of years of evolution, did the lateralization of the human brain appear? Is this predisposition of the left hemisphere for language of genetic origin? Is it present in the brain of the fetus and the baby before they begin to speak? Is it linked to manual dexterity? Studies in neuroimaging are beginning to provide rudimentary answers to those questions.
Following Broca's discovery there was considerable progress, because this concept had opened a true breach in our understanding of the functioning of the brain. For more than a century neurologists (and in particular those of the French school at the beginning of the twentieth century, with Pierre Marie, Jules Déjerine, Joseph Babinksi, and many others) learned a lot from their patients with cerebral lesions. They had only to closely observe the patients and their functional disorders, then "recover" their brains after their deaths to establish a link between the localization of the lesion and the functional deficit. We owe a debt of gratitude to these neurologists, excellent brain sleuths, who, from their careful and detailed observation of sometimes tiny neurological signs (a small anomaly in an eye movement, or subtle cognitive disorders brought to light through complex tests), were able to establish the localization of the lesion within a few centimeters (even a few millimeters in the brainstem). Even if the nature of the problem (for example, the blockage of a small cerebral artery by a blood clot) could sometimes be suggested by symptoms, this rarely resulted in the patient being cured, as therapeutic treatments at that time were quite limited. This conceptual approach in which functional deficits were associated with cerebral localizations nonetheless had its limits. First, brains had to be retrieved and dissected-and not all patients died! And those that did were not systematically autopsied. In addition, the localization of the lesions was not scientifically controlled in advance but were the work of Mother Nature. Whereas some regions were very often afflicted, others almost never were, and their functional role remained unexplored. This is explained in part by the fact that many lesions are of vascular origin, and that some vessels are more exposed or more fragile than others.
The Birth of Modern Neuroimaging
A radically different approach was taken as neurosurgery began to advance in the 1950s, in particular within the school of the Canadian neurosurgeon Wilder Penfield. During surgical interventions on the brain patients were awakened during the operation (the brain, even though it is the primary nerve center, is not sensitive to pain). By touching or electrically stimulating a cerebral region, the patient could report directly on his or her sensations, such as: "my thumb is asleep." Personalized functional maps of the brain could thus be drawn for each patient to pinpoint the regions to be avoided during the removal of a cancerous tumor, or the source of epileptic seizures, in order to preserve the patient's motor or language functions. This approach for the first time enabled Broca's theories to be proven "positively" (the expression of functions), whereas up until then the process was "negative" (a deficit of functions in the patients with cerebral lesions), by eliciting expressions of the functional content of healthy cerebral regions.
It was within this context that in the 1970s modern, computerized neuroimaging emerged, a true revolution that forever changed our approach to the brain. Until then neurologists had only cranial radiography at their disposal. X-rays, discovered by Wilhelm Röntgen in 1895, at best enabled the creation of shadowgraphs of the skull upon radiographic film, the shadows being roughly classed into four types depending on their intensity: bone and calcified structures (very opaque in x-rays); water and tissues containing water; fat (not very dense in x-rays); and air (transparent). X-rays of the skull thus showed only fractures of the bone, the appearance of abnormal blood vessels, and sometimes invasive tumors on the skull, occasionally calcified tumors, but not much more. One could go further by injecting a liquid containing iodine (opaque to x-rays) in the veins or arteries of the patient to display the vascular tree. Beyond direct afflictions of those vessels ("aneurysm": swelling of a fragment of the artery; "angioma": an abnormal proliferation of vessels; shrinking from an atheromatous plaque; blockage from a clot; and so on), one could guess at the presence of otherwise invisible lesions, such as tumors, from the fact that they displaced the normal vascular architecture1 (figure 1.3). One could also "make opaque" the ventricular cavities, open spaces in the center of the brain that contain cerebrospinal fluid (CSF), by injecting an iodized liquid or a bubble of air (gas encephalography) into the spinal cord at the lower back. By making the patient assume acrobatic positions (to the point of standing on his or her head), the bubble of air would travel to the cerebral ventricles, then into a specific corner of those ventricles. Besides the discomfort of the method and the horrible headaches that often followed, here too, only lesions that displaced or molded the ventricular cavities could be seen. Everything else remained discouragingly invisible.
On the functional side, neurologists were able to record the electrical activity emitted by the brain (electroencephalography, or EEG). Indeed, nerve impulses that enable neurons and brain cells to communicate among each other depend on the movements of ions (atoms that have lost their electrical neutrality) such as sodium, potassium, or calcium. The movements of these charged particles represent little electrical currents that create localized electrical and magnetic fields that can be detected and recorded at a distance using electrodes placed on the scalp. But the signals, whose localization remained very uncertain, above all enabled the detection of electrical occurrences that were abnormal in their intensity, as in the case of epileptic seizures, true cerebral "electrical storms," or in their absence in the case of tumors or a localized affliction of cerebral tissue.
The First Revolution: The X-Ray CT Scanner
The lives and comfort of these patients (and their neurologists) improved dramatically with the appearance of the x-ray computed tomography (CT) scanner in 1972 (figure 1.4), thanks to the English engineer Godfrey Hounsfield, who, along with Allan McLeold Cormak, received the Nobel Prize in Medicine or Physiology in 1979. This date marks a turning point: the introduction of computers into radiology (reference 1.2). The brain would finally become visible without having to open the skull. First, classic radiographic film was replaced by x-raycapturing sensors linked to a computer. Thus, instead of the four nuanced shadows detectable by the eye of the radiologist, the computer, through the sensors, could see hundreds. And so nuances between shadows produced by a healthy brain and those coming from lesions, or even different structures of the brain, could then be detected. Above all, instead of projecting a single dimension of shadow, the brain could be scanned under dozens, even hundreds, of different angles. This was made possible by the extreme sensitivity of the sensors and the digitization created by the computer: the dose of x-rays needed for a projection was infinitely reduced compared to classic radiography, which allowed the number of scans to be increased while keeping radiation at a low level.
Cormak showed that it was possible to combine these multiple projections, adding them to the computer's memory, to reconstruct, point by point, the entire picture of what was shown by the x-rays of the skull and its precious contents. The last step was to transform the virtual image into a real image by projecting it onto a screen (this general principle, moreover, was later used in future imaging methods that did not use x-rays). In CT scanning, scanning by x-rays is carried out on a plane (by turning the x-ray scanner around the head of the patient); the image obtained is thus that of a "slice," perpendicular to the axis of the head. In that slice (figure 1.4), one sees cutaneous structures, the bone of the skull with the tiny details of its internal structure, such as the external and internal bony parts of the skull, but above all, and for the first time, the interior of the skull, that is, the brain, the intracerebral ventricles, of course, and any lesions that might be found there—without an autopsy, without dissection, without pain or injury. The patient just lies down for a few minutes in the scanner and his or her brain is virtually dissected into slices by an x-ray beam.
The revolution is complete, and not only on a technological level. Radiologists, used to simple shadows, see much more than they did with their own eyes: and so they must learn to regulate the levels of contrast in the image in order to enhance a given structure. One speaks of "windows" of contrast, for it is indeed a matter of seeing, with the human eye, only a small bit of the landscape seen by the computer, by choosing the position and the width of the windows. Depending on that choice, the appearance of images can vary enormously, some windows allowing the details of the bones of the skull to appear more clearly, others the cerebral structures. Most important, at the beginning practitioners had to learn to think in terms of slices and rethink the entire three-dimensional (3D) anatomy of the brain learned during long years in anatomy classes and during dissections. Atlases of "tomodensitometry" (another name for computerized x-ray scanning) were created. For neurologists it was also a great surprise, a dream come true, but also from time to time disillusioning when they sometimes discovered that the localization of the lesion revealed from an in-depth analysis of the symptoms of their patient was not at all the same as the one revealed by the scanner ...
With x-ray CT scanning it became possible, in the treatment of a living patient presenting neurological symptoms, and thus when there was still time to prescribe a treatment, to reveal the presence of a lesion, localize it, and understand its impact on neighboring functional regions, and even sometimes to specify its nature depending on how it was shown in the image. One could also again inject an iodized liquid or another contrasting agent to accentuate the distinction between healthy and abnormal tissue. Such an agent, opaque to x-rays, is distributed in the vascular network, even into the smallest capillaries. A lesion with many vessels, such as a tumor, is thus clearly apparent (figure 1.4). One could also follow the evolution of a lesion in time, in particular to monitor its progression or shrinking following treatment. But the eye of the scanner sometimes sees too much, lesions that one doesn't know what to do with, that one can't explain, fortuitous discoveries during research into another lesion. And some visible lesions, alas, remain without precise diagnosis and remain "nonidentifiable objects," and above all without treatment.
In not much more than a dozen years the CT scanner began to appear in hospitals, and in another dozen it began to be systematically integrated into the healthcare system. Anecdotally, when I was a medical resident in 1980, it had been strongly recommended that we not speak of the CT scanner in our reports, so we would not be called doctors of science fiction! And yet, at about the same time that the CT scanner appeared, another revolution was brewing, that of magnetic resonance imaging, or MRI, the premises of which were published in 1973 in the journal Nature by the American chemist Paul Lauterbur (reference 1.3). It took this extraordinary technology even longer to penetrate the medical world, as some renowned scientists "didn't believe it." Lauterbur didn't receive the Nobel Prize in Medicine, by the way, until 2003, along with the English physicist Peter Mansfield.
(Continues...)
Excerpted from Looking Inside the Brain by Denis Le Bihan, Teresa Lavender Fagan. Copyright © 2012 Odile Jacob. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
