Milwaukee, Wisconsin
PLACE your index finger on the bridge of your nose. Now run it across your eye, over your ear and down slightly until you get to a bump near the back of your head. If it鈥檚 a big one, you are endowed with philoprogenitiveness-the love of children and animals. At least that鈥檚 what 19th-century phrenologists believed. They thought you could divine a person鈥檚 character from bumps on their skull. Today, we dismiss this as old-fashioned quackery. But there is a new generation of phrenologists emerging-only this time they want to get inside your skull and have a look at the bumps on the surface of your brain.
Everybody鈥檚 brain is covered in wrinkles. The cerebral cortex-the layer of grey matter that forms the surface of the brain-is folded into a complex pattern of ridges and troughs like a giant fingerprint. Neuroscientists are interested in this grey matter because it contains many parts of the brain that make us human-for example, those that manipulate ideas, give us our sense of self and help us plan for the future. By studying the pattern of wrinkles and the mechanics of how and why they form, neuroscientists hope to shed light on the wiring of the brain-which areas of the cortex are strongly or weakly connected to each other, and how these connections form. They believe that they could pinpoint flaws in the brain responsible for learning difficulties and other cognitive problems, and even mental illness such as schizophrenia.
So why do we have wrinkles in the first place? What鈥檚 wrong with a flat cortex? Well, nothing-unless you want to be intelligent. For humans and other large mammals to get smart without having to lug around absurdly large heads, nature had to pull off an impressive feat of engineering. To carry out all its functions, the cerebral cortex needs to be very large and packed with interconnections. To get all that grey matter to fit, nature folds it up like a concertina during fetal development. 鈥淵ou want to get as much surface area as possible into essentially a spherical space,鈥� says Alan Evans of the Montreal Neurological Institute at McGill University. A greater surface area means more interconnections between areas of the cortex, and so more processing power.
Advertisement
The pattern of outward folds, or gyri, and inward folds, or sulci, starts with the major ones. The first signs of the large sylvian fissure, a deep inward fold that separates the frontal from the temporal lobes of the brain, can be found by the end of the fourth month of pregnancy. Most of the less prominent folds emerge after six months. By birth the sulcal-gyral pattern is complete, says Martha Shenton, director of the clinical neuroscience division of Harvard Medical School鈥檚 Laboratory of Neuroscience in Boston. So if the pattern of wrinkles contains any useful medical information, it should be detectable from birth onwards.
But this doesn鈥檛 explain what governs the formation of the pattern of wrinkles. Brain wrinkles vary a lot from one person to the next, even in identical twins. In a study published earlier this year in the journal Brain, a team from the Neuroscience Center at St Elizabeth鈥檚 Hospital in Washington DC used magnetic resonance imaging (MRI) to compare brain folding in identical and fraternal twins. They demonstrated that although brain volume seems to be strongly influenced by genes, brain folding is mainly not genetically determined.
If genes don鈥檛 dictate the folding of the brain, something else must be going on. Environmental factors and even chance may have a part to play. According to neuroscientist David Van Essen, it all comes down to a massive engineering project between nerve cells in the developing brain. Van Essen, head of the department of anatomy and neurobiology at Washington University School of Medicine in St Louis, Missouri, presented his ideas on the mechanics of cortex folding earlier this year in an article in Nature.
Van Essen studied the brains of macaques, in particular the visual cortex, which lies on the occipital lobes near the back of the brain. He noticed that the two main areas of visual processing, called V1 and V2, always lie on either side of a certain gyrus. Van Essen believes the location of the folds is dictated by the wiring of the brain. Nerve cells in the cortex are connected to each other, and to other parts of the brain, by axons which contain a certain amount of internal tension, like tiny wires. As the cortex grows, the tension along these axons prevents the brain from simply expanding evenly like a party balloon-the more axons binding two parts of the cortex, the less they can expand. 鈥淭he tension along the countless axons that run between different areas of the cortex provides a gentle tugging force,鈥� Van Essen explains, 鈥渙ne that鈥檚 strong enough to generate the folds over the course of development.鈥�
Differences in the number of connections between patches of cortex introduce a kind of competition during fetal development. As the brain expands, each patch of cortex tries to stay close to other areas of cortex to which it is connected in a game of tug-of-war. In much of the brain, connections between patches of cortex are evenly balanced, so that predicting which areas win or lose the tug- of-war-and consequently where the sulci and gyri form-is a bit of a lottery. This explains why no two brains fold in the same way.
However, certain heavily connected areas always win and stay close together, and other weakly connected areas always lose, producing gyri or sulci common to all brains. For example, the prominent gyrus between V1 and V2 always forms because they have so many axon interconnections, and resist being separated. Sulci lie between areas of cortex that have few connections, such as the auditory and somatosensory areas on the sylvian fissure.
So according to Van Essen, the whole cortex is held together under tension. This idea makes sense because it keeps the distance between heavily connected areas short, just as a computer designer tries to keep the wiring short between the circuits and chips that process most information. But can the wrinkles tell the new phrenologists something about the wiring of the brain? 鈥淵es,鈥� says Van Essen, 鈥渂ut with a giant caveat. You can look at a pattern of folding and know that it is a consequence of the connections and the size of areas in the cortex, but there is an enormous number of connectivity patterns and arrangements of areas that could give the same folding pattern.鈥� So although a wrinkle might indicate particular connections, you can鈥檛 say for sure which ones.
Before neuroscientists can try a phrenological examination of the living brain, they need a way of 鈥渇eeling鈥� the wrinkles. Several groups are now working on ways to map the surface of the cortex. Researchers have developed software that can take an image of the cortex from a MRI scan and either unroll it into a flat sheet, inflate it like a balloon or build a three-dimensional image.
In the flat sheet method, the end result looks like a topographical map of the Earth with the sulcal-gyral folds shown as contour lines, like mountains and valleys. The 鈥渂alloon鈥� brain map inflates the cortex to several times its normal size, and highlights sulci and gyri using patches of colour. Both techniques can be used to compare brains directly-vital for spotting differences in wrinkle patterns.
Jumping over folds
Producing a flattened or inflated cortex sounds simple, but first it has to be isolated from a MRI scan of the whole brain. A MRI scan is made up of millimetre-thick slices of the brain, subdivided into tiny rectangular blocks called voxels. A special algorithm selects only the voxels of the cortex by following the trail of grey matter through the layers. But until recently, the brain mappers kept running into the same problem-in young, healthy brains, the gyri are so plump and rounded that they nearly touch. Instead of following the contours of the cortex surface around the inside of the sulcal fold, the algorithm jumped across to the next gyrus. So many hidden pockets of cortical surface lying within the sulcal folds were missed. To make a high-resolution map of the brain, 鈥測ou鈥檝e got to get down to the bottom of the sulcus,鈥� says Evans. 鈥淚f you can鈥檛 do that, then what you have is junk.鈥�
But in March, programmers at the University of Iowa鈥檚 Schizophrenia Clinical Research Center, headed by neuroscientist Nancy Andreasen, found a solution. Ted Cizadlo, Greg Harris and Dan Heckel created a new algorithm that can tell the difference between the grey matter of the cortex, the white matter beneath it, and cerebrospinal fluid (CSF), a liquid that bathes the entire brain. As it steps from voxel to voxel, their algorithm checks the contents of each for grey and white matter and CSF. When the algorithm reaches a voxel made of CSF as well as grey matter, it knows it has reached the edge of the brain-or one of the gaps between gyri.
The programmers have built the new algorithm into software that characterises the cortex in its three-dimensional form. The program divides the cortical surface into 100 000 tiny areas and looks at the thickness of the cortex under each, as well as how the surface curves. By examining these data, Andreasen and her colleagues hope to find meaningful patterns in the topography of the brain.
Broken brains
They have already begun their first phrenological examinations-on the brains of people with schizophrenia. The illness afflicts 1 per cent of Western populations. Schizophrenics can suffer hallucinations, paranoid delusions and other debilitating symptoms, but despite years of study there is no consensus on what causes it. The Iowa team has now scanned the brains of more than 100 schizophrenics and an equal number of non-schizophrenic people. They hope to discover if there is a structural defect at the root of schizophrenia.
Andreasen is one of the architects of the 鈥渂roken brain鈥� model, which proposes that flaws in the brain itself cause some mental illnesses. She says she started looking at brain wrinkles to find out whether schizophrenia is neurodevelopmental-that is, arising when something goes wrong in the growth and wiring of the brain.
If this is the case, Andreasen suspects that the defect for schizophrenia is located in the frontal lobes, where she and other scientists have found tissue loss in schizophrenics. Her team has already found a difference between the shape of the sulcal and gyral curves of schizophrenics and non-schizophrenics. The schizophrenic patients have more tightly curved gyri and looser, wider sulci. Andreasen is now looking to see if these differences are more pronounced in specific areas.
Subtle differences
Finding tell-tale wrinkle patterns will mean comparing the brains of schizophrenics and non-schizophrenics in great detail. 鈥淭he differences have to be subtle,鈥� says Shenton, who specialises in schizophrenia. 鈥淚f there were huge differences we would have found them already.鈥� Nevertheless, Shenton and MRI specialist Ron Kikinis, also at Harvard Medical School, are finding further evidence to suggest that in some patients schizophrenia might reveal itself in wrinkle patterns.
In a study published in 1994, Shenton and Kikinis compared the folding pattern of the temporal lobes of 15 schizophrenics with those of people without the illness. They found that the orientation of the sulci in this region tended to be more vertical in the brains of schizophrenics. Shenton hopes the new highly sensitive MRI techniques will show up more differences.
If schizophrenics do show a consistent abnormality in the folding of their brains, it could provide a way of diagnosing the disease before onset. 鈥淭he sulcal-gyral pattern shouldn鈥檛 change,鈥� Shenton says. 鈥淚f this is some sort of indicator, you could use it to evaluate children at risk.鈥� Eventually, Kikinis adds, discovering the cause of the illness 鈥渃ould lead to the possibility of doing something to prevent it鈥�.
In Montreal, Evans is part of a team of researchers at McGill headed by neuroscientist Noor Kabani that is hoping to correlate wrinkle patterns with a mental deficit called familial language impairment. People with this condition have trouble mastering abstract grammatical rules, such as forming the past tense. As the symptoms are similar across several languages, including English, French, Japanese, Greek and Inuktitut, the language of the Inuit people, some researchers suspect the disorder stems from a flaw in the brain. The team is targeting the areas of the cortex already known to process language-Broca鈥檚 area on the frontal lobe and Wernicke鈥檚 area on the temporal lobe.
Evans and Kabani have already studied 10 families with the language disorder. MRI scans of sufferers鈥� brains show unusual wrinkles. 鈥淭he sulci are deep and wide compared to the gyri which are thin and small,鈥� says Kabani. She is now developing an automatic technique for measuring the ratio of surface areas of sulci to gyri directly from MRI scans.
For the moment, however, the ideas and techniques of a new phrenology will have to wait for the evidence to catch up. 鈥淲e are a long way from knowing whether this might be useful for diagnosis,鈥� cautions Daniel O鈥橪eary, a researcher in Andreasen鈥檚 laboratory. 鈥淚t鈥檚 like trying to make sense of a fingerprint, to get a feel for what all those little bumps mean. And it may turn out that they don鈥檛 mean anything.鈥� But if schizophrenia and a rare language disorder can be highlighted by an unusual pattern of wrinkles, other conditions will come under scrutiny. And maybe then it will be time to crack open the symbological skill used by Victorian phrenologists and draw a new map on the exposed wrinkles of the brain.
