Byte by Byte, a Map of the Brain

Monday, July 2, 2001 - Arthur W. Toga is feeding the human brain to the
Reality Monster, one digital slice at a time.

This thing devouring the brain in bytes is a graphics supercomputer known
more formally as a Silicon Graphics Onyx2 Reality Monster, which is
designed to handle unusually large three-dimensional visual displays of
information.

In its automated carousels of 80-gigabyte storage drives, the computer
cradles the digital essence of the most complex object in the known
universe: the human brain.

In fact, it holds detailed anatomical data files on more than 2,000 human
brains, with data on 3,000 more waiting to be added.

The data set for just one of these brains is the electronic equivalent of a
million books.

By the end of the year, Toga at UCLA's Laboratory of Neuro Imaging and Dr.
John Mazziotta, director of UCLA's brain-mapping division, expect to have
7,000 brains from seven countries scanned into this electronic organ bank.

When complete, this database of anatomy and behavior will help guide
researchers and physicians in a wide range of applications, such as
studying brain development in children, probing the mysteries of mental
illness and determining the proper course of treatment for patients with
neurological disorders.

Sitting in a hushed corner of a fourth-floor computer room at UCLA, the
Reality Monster defines a frontier of computer science, advanced graphics
research and the effort to understand the almost infinite variety of the
human brain.

At a time when all science is in some way computer science, researchers
have harnessed this artificial thinking machine to take the measure of the
mind.

It is the center of a $14-million federal project underway at eight
laboratories to create an interactive electronic atlas that will accurately
capture all that can be calculated about the brains of the world's population.

The researchers are developing the mathematical formulas, data management
tools and graphic techniques needed to navigate this ever-changing terrain
reliably.

An additional $30 million is being spent around the world - from Japan to
Finland - to gather the brain data, along with DNA samples and a detailed
biomedical dossier on each person being studied, including information
about race, ethnic origins, habits, diet, education and occupation.

Already, researchers are beginning to catch glimpses of patterns that may
hold true across large groups of people.

Preliminary studies using the mapping techniques suggest there are:

* Distinctive differences in brain structure between men and women with
schizophrenia, highlighting the possibility of gender-related patterns of
mental illness.

* Telltale patterns of gray matter loss in the brains of people with
Alzheimer's disease, suggesting the possibility of diagnosis long before
any changes in behavior or memory are noticeable.

* Continual changes from childhood through adolescence in the distribution
of gray matter in the brain, offering evidence of developmental changes
later in life than previously assumed.

The mapping effort is part of the federal Human Brain project, an
initiative sponsored by the National Institute of Mental Health and 15
other government organizations.

Scientists hope that the brain-mapping project ultimately will bring order
to a cascade of startling findings about the brain in recent years based on
new imaging techniques and computerized measuring tools.

Virtually every aspect of neuroscience today relies in some way on
sophisticated digital images that give researchers a way to visualize
enormous quantities of data about the organ of thought.

Some of these imaging techniques employ mildly radioactive tracers to
detect how the brain uses energy.

Others take advantage of the magnetic properties of oxygenated blood to
make images of mental activity.

Computerized arrays of electrodes trace magnetic and electrical signals
associated with mental processes.

Using them, researchers have declared that they can see the complex neural
patterns of joy, fear, anger and romantic love.

They have pointed to a place in the brain where memories dwell and contend
they can discern the difference between true and false recollections.

In subtle patterns of light and darkness projected on their computer
screens, researchers are convinced they can detect signs of degenerative
disease in the brain many years before symptoms develop.

They can see how men and women use their brains differently, tell
left-handers from right-handers, the old from the young, and how humankind
differs from other primates - all by scanning the brain in different ways.

At this level of brain studies, however, nobody really knows how to match
images from one brain to another reliably or to accurately merge data
gathered with different techniques.

No laboratory, therefore, can duplicate the work of another consistently
enough to confirm the discoveries of such brain-imaging studies.

Nor is it easy to relate findings based on the study of a few individuals
to the population at large.

With so much new data in so many incompatible forms, it is virtually
impossible for any single scientist to maintain an integrated view of the
brain.

"We have to understand what we are looking at before we can understand what
it means," Toga said.

All people may be created equal, but no two brains are alike, despite their
overall anatomical similarity.

Each brain changes throughout a lifetime as it is altered by the effects of
experience and aging.
Even the simplest mental activities, such as watching a moving dot, can
involve slightly different areas in different people's brains.

And no one knows yet how much human brains around the world differ from
each other in size, structure and function.

"Studying the brain is kind of a mess because we don't have a way to
communicate where we are in it," Mazziotta said.

"Unlike the Earth, which has a relatively fixed geometry from day to day,
there is no single unique physical representation of the human brain,
because everybody's brain is different in terms of structure and function.
We don't know by how much they are different," Mazziotta said.

"This is a project born of frustration, no question about it."

So far, the researchers have focused primarily on three-dimensional
anatomical data collected with MRI scanners.

But they have also started collecting information about behavior,
biochemistry and brain function.

To analyze their growing store of brain images, Toga and Mazziotta's
research group first developed a standard reference grid that could be used
like latitude and longitude to precisely locate the most minute anatomical
features of the brain.

They then created a formula that can stretch or adjust the shape of one
brain to match another on a standard form without introducing any
inaccuracies or distortions.

"We can do this over the entire cortex. We can do this lobe by lobe. We can
transform multiple brains into the same space," Toga said.

In this way they can compare a few dozen brains or a few thousand.

So far, the Reality Monster has consumed nearly 100 terabytes of
anatomical, structural and behavioral data in what is rapidly becoming one
of the world's largest databases on the human brain - a raw volume of
information approximately equal to a library of 100 million tomes.

Yet it barely scratches the surface of the complexity of the organ that
animates thought.

"I think you can see this as a first step in terms of trying to create a
system that enables you to detect differences in brains," Toga said.


By ROBERT LEE HOTZ, Times Science Writer
Copyright 2001 Los Angeles Times
http://www.latimes.com/news/science/science/20010702/t000054720.html

janet paterson: an akinetic rigid subtype, albeit perky, parky .
pd: 54/41/37 cd: 54/44/43 tel: 613 256 8340 email: [log in to unmask] .
snail mail: 375 Country Street, Apt 301, Almonte, Ontario, Canada, K0A 1A0 .
a new voice: the nnnewsletter: http://groups.yahoo.com/group/janet313/ .
a new voice: the wwweb site: http://www.geocities.com/janet313/ .

----------------------------------------------------------------------
To sign-off Parkinsn send a message to: mailto:[log in to unmask]
In the body of the message put: signoff parkinsn