Documentation

Blue Brain 1
Dept. OI InIormation Technology YCET`11
ABSTRACT
Touay scientists aie in ieseaich to cieate an aiticial biain that
can think, iesponu, take uecision, anu keep anything in memoiy. The
main ai m i s t o upl oad human brain into machine. So that man
can think, take decision without any eIIort. AIter the death oI the
body, the virtual brain will act as the man. So, even aIter the
death oI a person we will not loose the knowledge, intelligence,
personalities, Ieelings and memories oI that man, that can be used Ior the
development oI the human society. Technology is giowing fastei than
eveiy thing. IBN is now in ieseaich to cieate a viitual biain, calleu
"Blue biain". If possible, this woulu be the ist viitual biain
of the woilu. IBN,In paitneishipwith scientists at Switzeilanu's Ecole
Polytechnique Feueiale ue Lausanne's (EPFL) Biain anu Ninu
Institute will begin simulating the brain`s biological systems and
output the data as a working 3-dimensional model that will
recreate the high-speed electro-chemical interactions that take
place within the brain`s interior. These include cognitive
Iunctions such as language, learning, perception and memory
in addition to brain malIunction such as psychiatric disorders
like depression and autism. From there, the modeling will
expand to other regions oI the brain and, iI successIul, shed light on
the relationships between genetic, molecular and cognitive Iunctions oI
the brain.

Blue Brain 2
CHAPTER 1
INTRODUCTION
Human brain is the most valuable creation oI God. The man is
called intelligent because oI the brain. The brain translates
the inIormation delivered by the impulses, which then enables
the person to react. But we loss the knowledge oI a brain when
the body is destroyed aIter the death oI man. That knowledge
might have been used Ior the development oI the human society.
What happen iI we create a brain and up load the contents oI
natural brain into it?
1.1 Blue Brain
The name of the woilu's ist viitual biain. That means
a machine that can function as human brain. Today scientists are
in research to create an artifcial brain that can think, response,
take decision, and keep anything in memory. The main aim is to
upload human brain into machine. So that man can think, take
decision without any eIIort. AIter the death oI the body, the
virtual brain will act as the man . So, even aIter the death oI a person
we will not loose the knowledge, intelligence, personalities ,Ieelings
and memories oI that man that can be used Ior the development
oI the human society. No one has ever understood the
complexity oI human brain. It is complex than any circuitry
in the world. So, question may arise 'Is it really possible to
create a human brain? The answer is 'Yes. Because what ever man
has created today always she has followeu the natuie. When man
uoes not have a uevice calleu computei, it was a big question foi
all. Technology is giowing fastei than eveiy thing. IBN is now in
ieseaich to cieate a viitual biain, calleu "Blue biain". If possible, this
Blue Brain 3
woulu be the ist viitual biain of the woilu. With in yeais, we will
be able to scan ourselves in to the computers. Is this the
beginning oI eternal liIe?
1. 2 What i s Vi rt ual Brai n?
Virtual brain is an artifcial brain, which does not actually the
natural brain, but can act as the brain. It can think like brain, take
decisions based on the past experience, and response as the natural
brain can. It is possible by using a super computer, with a huge
amount oI storage capacity, processing power and an interIace
between the human brain and this artifcial one. Through this interIace
the data stored in the natural brain can be up loaded into the computer. So
the brain and the knowledge, intelligence oI anyone can be kept and
used Ior ever, even aIter the death oI the person
1. 3 Why we need Vi rt ual Brai n?
Today we are developed because oI our intelligence. Intelligence is
the inborn quality that can not be created. Some people have this
quality, so that they can think up to such an extent where other
can not reach. Human society is always need oI such intelligence
and such an intelligent brain to have with. But the intelligence is lost
along with the body aIter the death. The virtual brain is a solution to it.
The brain and intelligence will alive even aIter the death. We oIten Iace
diIfculties in remembering things such as people`s names, their
birthdays, and the spellings oI words, proper grammar, important
dates, history, Iacts etc... In the busy liIe every one want to be relaxed.
Can`t we use any machine to assist Ior all these? Virtual brain may be the
solution to it. What iI we upload ourselves into computer, we were simply
Blue Brain 4
aware oI a computer, or may be what iI we lived in a computer as a
program?
Blue Brain 5
1.4 How it is possible?
First, it is helpIul to describe the basic manners in which a person
may be uploaded into a computer. Raymond Kurz weil recently provided
an interesting paper on this topic. In it, he describes both invasive
and noninvasive techniques. The most promising is the use oI very small
robots, or nano bots. These robots will be small enough to travel
throughout our circulatory systems.
Traveling into the spine and brain, they will be able to monitor the
activity and structure oI our central nervous system
TheywillbeabletoprovideaninterIacewithcomputersthatisascloseasourmin
dcanbe while we still reside in our biological Iorm. Nano bots could also
careIully scan the structure oI our brain, providing a complete readout oI
the connections between each neuron. They would also record the current
state oI the brain. This inIormation, when entered into a computer, could
then continue to Iunction as us. All that is required is a computer with
large enough storage space and processing power. Is the pattern and state
oI neuron connections in our brain truly all that makes up our conscious
selves? Many people believe frmly those we posses a soul, while some
very technical people believe that quantum Iorces contribute to
our awareness. But we have to now think technically. Note, however, that
we need not know how the brain actually Iunctions, to transIer it to a
computer. We need only know the media and contents. The actual
mystery oI how we achieved consciousness in the frst place, or how we
maintain it, is a separate discussion. Really this concept appears to be
very diIfcult and complex to us. For this we have to frst know how the
human brain actually works

Blue Brain 6
CHAPTER 2
WORKING OF NATURAL BRAIN
2.1 Getting to know more about Human Brain

The brain essentially serves as the body`s inIormation
processing centre. It receives signals Irom sensory neurons (nerve cell
bodies and their axons and dendrites)in the central and peripheral
nervous systems, and in response it generates and sends new
signals that instruct the corresponding parts oI the body to move
or react in some way. It also integrates signals received Irom the body
with signals Irom adjacent areas oI the brain, giving rise to perception
and consciousness. The brain weighs about 1,500 grams (3 pounds)
and constitutes about 2 percent oI total body weight. It consist s
oI three major divisions;
1 The massive paired hemispheres oI the cerebrum
2 The brainstem, consisting oI the thalamus, hypothalamus,
epithalamus, subtha-lamus, midbrain, pons, and medulla
oblongata
3 The cerebellum.
The human ability to Ieel, interpret and even see is controlled, in
computer like calculations, by the magical nervous system. The
nervous system is quite like magic because we can`t see it, but
its working through electric impulses through your body. One oI
the worlds most 'intricately organized electron mechanisms is
the nervous system. Not even engineers have come close
to making circuit boards and computers as delicate and precise as
the nervous system. To understand this system, one has to know
the three simple Iunctions that it puts into action; sensory input,
integration &motor output.
Blue Brain 7

2. 1. 1 Sensory Input
When our eyes see something or our hands touch a warm
surIace, the sensory cel l s, al so known as Neur ons, send
a message st rai ght t o your br ai n. Thi s act i on oI getting
inIormation Irom your surrounding environment is called sensory
input because we are putting things in your brain by way oI your
senses.
2. 1. 2 Int egr at i on
Integration is best known as the interpretation oI things we
have Ielt, tasted, and touched with our sensory cells, also known
as neurons, into responses that the body recognizes. This process
is all accomplished in the brain where many, many neurons work
together to understand the environment.
2. 1. 3 Mot or Out put
Once our brain has interpreted all that we have learned, either by
touching, tasting, or using any other sense, then our brain sends
a message through neurons to eIIecter cells, muscle or gland
cells, which actually work to perIorm our requests and act upon
our environment How we see, hear, Ieel , & smel l ?
2 . 2 . 1 N o s e
Once t he smel l oI Iood has r eached your nose, whi ch
i s l i ned wi t h hai r s, i t travels to an olIactory bulb, a set oI sensory
nerves. The nerve impulses travel through the olIactory tract, around,
in a circular way, the thalamus, and fnally to the smell sensory
cortex oI our brain, located between our eye and ear, where it is
interpreted to be understood and memorized by the body.
2 . 2 . 2 E y e
Blue Brain 8
Seeing is one oI the most pleasing senses oI the nervous system.
This cherished action primarily conducted by the lens, which
magnifes a seen image, vitreous disc, which bends and rotates an
image against the retina, which translates the image and l i ght by
a set oI cel l s. The r et i na i s at t he back oI t he e ye bal l
wher e r ods and cones structure along with other cells and tissues
covert the image into nerve impulses which are transmitted along the
optic nerve to the brain where it is kept Ior memory.
2.2 How we see, hear, feel, & smell?
2.2.1 Nose
Once t he smel l oI Iood has r eached your nose, whi ch
i s l i ned wi t h hai r s, i t travels to an olIactory bulb, a set oI sensory
nerves. The nerve impulses travel through the olIactory tract, around,
in a circular way, the thalamus, and fnally to the smell sensory
cortex oI our brain, located between our eye and ear, where it is
interpreted to be understood and memorized by the body.
2.2.2 Eye
Seeing is one oI the most pleasing senses oI the nervous system.
This cherished action primarily conducted by the lens, which
magnifes a seen image, vitreous disc, which bends and rotates an
image against the retina, which translates the image and l i ght by
a set oI cel l s. The r et i na i s at t he back oI t he e ye bal l
wher e r ods and cones structure along with other cells and tissues
covert the image into nerve impulses which are transmitted along the
optic nerve to the brain where it is kept Ior memory.
2.2.3 Tongue
A set oI microscopic buds on the tongue divide everything we eat
and drink in to Iour kinds oI taste: bitter, sour, salty, and sweet. These
buds have taste pores, which convert the taste into a nerve impulse
Blue Brain 9
and send the impulse to the brain by a sensory nerve fber. Upon
receiving the message, our brain classifes the diIIerent kinds oI taste.
This is how we can reIer the taste oI one kind oI Iood to another.
2.2.4 Ear
Once the sound or sound wave has entered the drum, it goes to a
large structure called the cochlea. In this snail like structure, the sound
waves are divided into pitches. The vibrations oI the pitches in the
cochlea are measured by the Corti. This organ transmits the vibration
inIormation to a nerve, which sends it to the brain Ior interpretation and
memory.
Blue Brain 10
CHAPTER 3
HOW THE BLUE BRAIN PRO1ECT WILL WORK?
3.1 Goals & Objectives
The Blue Brain Project is the frst comprehensive attempt
to reverse engineer the mammalian brain, in order to understand
brain Iunction and dysIunction through detailed simulations. The
mission in undertaking The Blue Brain Project is to gather all
existing knowledge oI the brain, accelerate the global research
eIIort oI reverse engineering the structure and Iunction oI the
components oI the brain, and to build a complete
theoretical Iramework that can orchestrate
the reconstruction oI the brain oI mammals and man Irom the
genetic to the whole brain levels, into computer models Ior
simulation, visualization and automatic knowledge archiving
by 2015. Biologically accurate computer models oI mammalian and
human brains could provide a new Ioundation Ior understanding
Iunctions and malIunctions oI the brain and Ior a new generation
oI inIormation-based, customized medicine.
3. 2Archi tect ure of Bl ue Gene
Blue Gene/L is built using system-on-a-chip technology in which
all Iunctions oI a node (except Ior main memory) are integrated onto a
single application specifc integrated circuit (ASIC). This ASIC includes
2 PowerPC 440 cores running at 700MHz. Associated with each core is
a 64-bit 'double foating point unit (FPU) that can operate
in single instruction, multiple data (SIMD) mode. Each (single) FPU can
execute up to 2 'multiply-adds per cycle, which means that the peak
perIormance oI the chip is 8 foating point operations per cycle (4 under
normal conditions, with nouse oI SIMD mode). This leads
Blue Brain 11
to a peak perIormance oI 5.6 billion foating point operations per second
(gigaFLOPS or GFLOPS) per chip or node, or 2.8 GFLOPS.
In non SIMD mode. The two CPUs (central processing units) can
be used in 'co-processor mode (resulting in one CPU and 512 MB RAM
(random access memory)Ior computation, the other CPU being used Ior
processing the I/O (input/output) oI the main CPU) or in 'virtual node
mode (in which both CPUs with 256 MB each are used Ior
computation). So, the aggregate perIormance oI a processor card in
virtual node mode is: 2 x node 2 x 2.8 GFLOPS 5.6 GFLOPS, and its
peak perIormance(optimal use oI double FPU) is: 2 x 5.6 GFLOPS 11.2
GFLOPS. A rack (1,024 nodes 2,048 CPUs) thereIore has 2.8 tera
FLOPS or TFLOPS, and a peak oI 5.6 TFLOPS. The Blue Brain Projects
Blue Gene is a 4-rack system that has 4,096 nodes, equal to8,192 CPUs,
with a peak perIormance oI 22.4 TFLOPS. A 64-rack machine should
provide 180 TFLOPS, or 360 TFLOPS at peak perIormance.
3.3 Modelling the Microcircuit
The scheme shows the minimal essential building blocks required
to reconstruct a neural microcircuit. Microcircuits are
composed oI neurons and synapticconnections. To model neurons, the thr
ee-dimensional morphology, ion channel composition, and distributions
and electrical properties oI the diIIerent types oI neuronare required, as
well as the total numbers oI neurons in the microcircuit and the relative
proportions oI the diIIerent types oI neuron. To model synaptic
connections, the physiological and pharmacological properties oI the
diIIerent types oI synapse that connect any two types oI neuron are
required, in addition to statistics on which part oI the axonal arborization
is used (pre synaptic innervation pattern) to contact which regions oI the
target neuron (postsynaptic innervations pattern), how many synapses are
involved in Iorming connections, and the connectivity statistics between
Blue Brain 12
any two types oI neuron. Neurons receive inputs Irom thousands
oI other neurons, which are intricately mapped onto diIIerent branches oI
highly complex dendritic trees and require tens oI thousands
oI compartments to accurately represent them. There is
thereIore a minimal size oI a microcircuit and a minimal complexity oI a
neuron`s morphology that can Iully sustain a neuron
.
A massive increase in computational power is required to make
this quantum leap an increase that is provided by
IBM`sBlue Gene supercomputer. By exploiting the computing power
oI Blue Gene, the Blue Brain Project1 aims to build accurate models oI
the mammalian brain Irom frst principle. The frst phase oI the project
is to build a cellular-level (as opposed to a genetic- or molecular-level)
model oI a 2-weekold rat so mato sensory neo cortex corresponding to the
dimensions oI a neocortical column (NCC) as defned by the dendritic
arborizations oI the layer 5 pyramidal neurons. The combination
oI inIrared diIIerential interIerence microscopy in brain slices and the
use oI multi-neuron patch clamping allowed the systematic
quantifcation oI the molecular, morphological and electrical
properties oI the diIIerent neurons and their synaptic pathways in
a manner that would allow an accurate reconstruction oI the column.
Over the past 10 years, the laboratory has prepared Ior this reconstruction
by developing the multi-neuron patch-clamp approach, recording Irom
thousands oI neocortical neurons and their synaptic connections, and
developing quantitative approaches to allow a complete numerical
breakdown oI the elementary building blocks oI the NCC. The recordings
have mainly been in the 14-16-day-old rat so mato
sensory cortex, which is a highly accessible region on which many
researchers have converged Iollowing a series oI pioneering
Blue Brain 13
studies driven by Bert Sakmann. Much oI the raw data is located in
our databases,but a major initiative is underway to make all these data
Ireely available in a publicly accessible database. The so-called `blue
print` oI the circuit, although not entirely complete, has reached a
suIfcient level oI refnement to begin the reconstruction at the cellular
level. Highly quantitative data are available Ior rats oI this age, mainly
because visualization oI the tissue is optimal Irom a technical point
oI view. This age also provides an ideal template because it can serve as
a starting point Irom which to study maturation and ageing oI the NCC.
As NCCs show a high degree oI stereotypy, the region Irom which the
template is built is not crucial, but a sensory region is preIerred because
these areas contain a prominent layer 4 with cells specialized to receive
input to the neo cortex Irom the thalamus; this will also be required Ior
later calibration within vivo experiments. The NCC should not be overly
specialized, because this could make generalization to other neocortical
regions diIfcult, but areas such as the barrel cortex do oIIer the advantage
oI highly controlled in vivo data Ior comparison. The mouse might have
been the best species to begin with, because it oIIers a spectrum
oI molecular approaches with which to explore the circuit, but mouse
neurons are small, which prevents the detailed dendritic recordings that
are important Ior modeling the nonlinear properties oI the complex
dendritic trees oI pyramidal cells (75-80 oI the neurons). The image
shows the Microcircuit in various stages oI reconstruction. Onlya small
Iraction oI reconstructed, three dimensional neurons is shown. Red
indicatesthe dendritic and blue the axonal arborizations. The columnar
structure illustrates the layer defnition oI the NCC.

Blue Brain 14

3.4 Simulating the Microcircuit
Once the microcircuit is built, the exciting work oI making the
circuit Iunction can begin. All the 8192 processors oI the Blue Gene are
pressed into service, in a massively parallel computation solving the
complex mathematical equations that govern the electrical activity in
each neuron when a stimulus is applied. As the electrical impulse travels
Irom neuron to neuron, the results are communicated via inter processor
communication (MPI). Currently, the time required to simulate the circuit
is about two orders oI magnitude larger than the actual biological time
simulated. The Blue Brain team is working to streamline the computation
so that the circuit can Iunction in real time - meaning that 1 second oI
activity can be modeled in one second.
3.5 Interpreting the Results
Running the Blue Brain simulation generates huge amounts oI
data. Analyses oI individual neurons must be repeated thousands oI times.
And analyses dealing with the network activity must deal with data
that easily reaches hundreds oI gigabytes per second oI simulation. Using
massively parallel computers the data can be analyzed where it is created
(server-side analysis Ior experimental data, online analysis during
simulation) Given the geometric complexity oI the column, a visual
exploration oI the circuit is an important part oI the analysis. Mapping the
simulation data onto the morphology is invaluable Ior an immediate
verifcation oI single cell activity as well as network
phenomena. Architects at EPFL have worked with the Blue Brain
developers to design a visualization interIace that translates the Blue
Gene data into a 3Dvisual representation oI the column. A diIIerent
Blue Brain 15
supercomputer is used Ior this computationally intensive task. The
visualization oI the neurons` shapes is a challenging task given the Iact
that a column oI 10,000 neurons rendered in high quality mesh accounts
Ior essentially 1 billion triangles Ior which about 100GB oI management
data is required. Simulation data with a resolution
oI electrical compartments Ior each neuron accounts Ior another 150GB.
As the electrical impulse travels through the column, neurons light up
and change color as they become electrically active. A visual interIace
makes it possible to quickly identiIy areas oI interest that can then be
studied more extensively using Iurther simulations. A visual
representation can also be used to compare the simulation results with
experiments that show electrical activity in the brain.
3.6 Data Manipulation Cascade
Building the Blue Column requires a series oI data manipulations
.The frst step is to parse each three-dimensional morphology and correct
errors due to the in vitro preparation and reconstruction. The repaired
neurons are placed in a database Irom which statistics Ior the diIIerent
anatomical classes oI neurons are obtained. These statistics are used to
clone an indefnite number oI neurons in each class to capture the Iull
morphological diversity. The next step is to take each neuron and insert
ion channel models in order to produce the array oI electrical types. The
feld has reached a suIfcient stage oI convergence to generate eIIorts to
classiIy neurons, such as the Petilla Convention - a conIerence held
in October 2005 on anatomical and electrical types oI neocortical
interneuron, established by the community. Single cell gene expression
studies oI neocortical inter neurons now provide detailed predictions oI
the specifc combinations oI more than 20 ion channel genes that underlie
electrical diversity. A database oI biologically accurate Hodgkin-Huxley
Blue Brain 16
ion channel models is being produced. The simulator NEURON is used
with automated ftting algorithms running on Blue Gene to insert ion
channels and adjust their parameters to capture the specifc electrical
properties oI the diIIerent electrical types Iound in each anatomical
class. The statistical variations within each electrical class are also used
to generate subtle variations in discharge behaviour in each neuron. So,
each neuron is morpho-logically and electrically unique. Rather than
taking 10,000 days to ft each neuron`s electrical behaviour with a unique
profle, density and distribution oI ion channels, applications are being
prepared to use Blue Gene to carry out such a ft in a day. These
Iunctionalized neurons are stored in a database. The three-dimensional
neurons are then imported into Blue Builder, a circuit builder that loads
neurons into their layers according to a 'recipe oI neuron numbers and
proportions.
A collision detection algorithm is run to determine the structural
positioning oI all axo dendritic touches, and neurons are jittered and spun
until the structural touches match experimentally derived
statistics. Probabilities oI connectivity between diIIerent types oI neuron
are used to determine which neurons are connected, and all axo-
dendritic touches are converted into synaptic connections. The manner in
which the axons map onto the dendrites between specifc anatomical
classes and the distribution oI synapses received by a class oI neurons are
used to veriIy and fne tune the biological accuracy oI the
synaptic mapping between neurons. It is thereIore possible to place 10-
50 million synapses in accurate three-dimensional space, distributed on
the detailed three dimensional morphology oI each neuron. The synapses
are Iunctionalized according to the synaptic parameters Ior diIIerent
classes oI synaptic connection within statistical variations oI each class,
Blue Brain 17
dynamic synaptic models are used to simulate transmission, and synaptic
learning algorithms are introduced to allow plasticity. The distance Irom
the cell body to each synapse is used to compute the axonal delay, and the
circuit confguration is exported. The confguration fle is read by a
NEURON subroutine that calls up each neuron and eIIectively inserts
the location and Iunctional properties oI every synapse on the axon, soma
and dendrites. One neuron is then mapped onto each processor and the
axonal delays are used to manage communication between neurons
and processors. EIIectively, processors are converted into neurons,
and MPI(message-passing interIace)- based communication cables are
converted into axons interconnecting the neurons - so the entire Blue
Gene is essentially converted into a neocortical
microcircuit. We developed two soItware programs Ior simulating such
large-scale networks with morphologically complex neurons. A new
MPI version oI NEURON has been adapted by Michael Hines to run on
Blue Gene. The second simulator uses the MPI messaging component oI
the large-scale Neo Cortical Simulator (NCS), which was developed by
Philip Goodman, to manage the communication between NEURON-
simulated neurons distributed on diIIerent processors.
The latter simulator will allow embedding oI a detailed NCC
model into a simplifed large scale model oI the whole brain. Both oI
these soItwares have already been tested, produce identical results and
can simulate tens oI thousands oI morphologically and electrically
complex neurons (as many as 10,000 compartments per neuron with more
than a dozen Hodgkin-Huxley ion channels per compartment). Up to 10
neurons can be mapped onto each processor to allow simulations oI the
NCC with as many as 100,000neurons. Optimization oI these algorithms
could allow simulations to run at close to real time. The circuit
Blue Brain 18
confguration is also read by a graphic application, which renders the
entire circuit in various levels oI textured graphic Iormats. Real-time
stereo visu-alization applications are programmed to run on the terabyte
SMP (shared memory processor) Extreme series Irom SGI (Silicon
Graphics, Inc.). The output Irom BlueGene (any parameter oI the model)
can be Ied directly into the SGI system to perIorm in silico imaging oI the
activity oI the inner workings oI the NCC. Eventually, the simulation oI
the NCC will also include the vasculature, as well as the glial network, to
allow capture oI neuronglia interactions. Simulations oI extracellular
currents and feld potentials, and the emergent electroencephalogram
(EEG) activity will also be modeled.
3.7 Whole Brain Simulations
The main limitations Ior digital computers in the simulation oI biol
ogicalprocesses are the extreme temporal and spatial resolution demanded
by somebiological processes, and the limitations oI the algorithms that ar
e used to model biological processes.
II each atomic collision is simulated, the most powerIul super-
computers still take days to simulate a microsecond oI protein Iolding, so
itis,oIcourse, not possible to simulate complex biological systems at the at
omic scale.However, models at higher levels, such as themolecular or cell
ular levels, can capture lower-level processes and allow complex large-
scale simulations oI biological processes. The Blue Brain Project`s Blue
Gene can simulate a NCC oI up to 100,000 highly complex neurons at the
cellular or as many as 100 million simple neurons(about the same number
oI neurons Iound in a mouse brain). However, simulating neurons
embedded in microcircuits, microcircuits embedded in brain regions, and
brain regions embedded in the whole brain as part oI the process oI
Blue Brain 19
understanding the emergence oI complex behaviors oI animals is an
inevitable progression in understanding brain Iunction and dysIunction,
and the question is whether whole-brain simulations are at all
possible. Computational power needs to increase about 1-million-Iold
beIore we will be able to simulate the human brain, with 100 billion
neurons, at the same level oI detail as the Blue Column. Algorithmic and
simulation eIfciency (which ensure that all possible FLOPS are
exploited) could reduce this requirement by two to three orders oI
magnitude. Simulating the NCC could also act as a test-bed to refne
algorithms required to simulate brain Iunction, which can be used to
produce feld programmable gate array (FPGA)-based chips. FPGAs
could increase computational speeds by as much as two orders
oI magnitude. The FPGAs could, in turn, provide the testing
ground Ior the production oI specialized NEURON solver application-
specifc integrated circuits (ASICs) that could Iurther increase
computational speed by another one to two orders oI magnitude. It could
thereIore be possible, in principle, to simulate the human brain even with
current technology. The computer industry is Iacing what is known as a
discontinuity, with increasing processor speed leading to unacceptably
high power consumption and heat production. This is pushing a
qualitatively new transition in the types oI processor to be used in Iuture
computers. These advances in computing should begin to make genetic-
and molecular-level simulations possible. SoItware applications and data
manipulation required to model the brain with three orders oI
magnitude. Simulating the NCC could also act as a test-bed to
refnealgorithms required to simulate brain Iunction, which can
be used to produce feldprogrammable gate array (FPGA)-based
chips. FPGAs could increase computationalspeeds by as much
as t wo or der s oI magni t ude. The FPGAs coul d, i n t ur n,
Blue Brain 20
provi dethe testing ground Ior the production oI specialized
NEURON solver application-specifc integrated circuits (ASICs)
that could Iurther increase computational speed by another one to
two orders oI magnitude.
It could thereIore be possible, in principle, to simulate the
human brain even with current technology. The computer
industry is Iacing what is known as a discontinuity, with
increasing processor speed leading to unacceptably high power
consumption and heat production. This is pushing a qualitatively new
transition in the types oI processor to be used in Iuture
computers. These advances in computing should begin to make genetic
and Molecular level simulations possible. SoItware applications and data
manipulation required to model the brain with biological accuracy.
Experimental results that provide the elementary building blocks oI the
microcircuit are stored in a database. BeIore three-dimensional
neurons are modeled electrically, the morphology is parsed Ior errors, and
Ior repair oI arborizations damaged during slice preparation. The
morphological statistics Ior a class oI neurons are used to clone multiple
copies oI neurons to generate the Iull morpho-logical diversity and the
thousands oI neurons required in the simulation. A spectrum oI ion
channels is inserted, and conductances and distributions are
altered to ft the neurons electrical properties according to known
statistical distributions, to capture the range oI electrical classes
and the uniqueness oI each neurons behaviour (model
ftting/electrical capture). A circuit builder is used to place neurons
within a three-dimensional column, to perIorm axo-dendritic collisions
and, usingstructural andIunctionalstatistics oI synaptic connectivity, to co
nvert a Iraction oI axo dendritic touches
Blue Brain 21
into synapses. The circuit confguration is read by NEURON, which calls
up each modeled neuron and inserts the several thousand synapses onto
appropriate cellular locations. The circuit can be inserted into a brain
region usingthe brainbuilder. Anenvironmentbuilderisusedtosetupthestim
ulusandrecordingconditions.Neurons are mapped onto processors, with
integer numbers oI neurons per processor. The output is visualized,
analysed and/or Ied into real-time algorithms Ior Ieedback stimulation.

Blue Brain 22
CHAPTER 4
APPLICATIONS OF BLUE BRAIN PRO1ECT
4.1 What can we learn from Blue Brain?
Det ai l ed, bi ol ogi cal l y accur at e brai n si mul at i ons oI I e
r t he oppor t uni t y t o answer some Iundamental questions about
the brain that cannot be addressed with any current experimental or
theoretical approaches. These include,
4.1.1 Defning Iunctions oI the basic elements
Despite a century oI experimental and theoretical research,
we are unable to provide a comprehensive defnition oI the
computational Iunction oI diIIerent ion channels, receptors,
neurons or synaptic pathways in the brain. A detailed model will
allow fne control oI any oI these elements and allow a systematic
investigation oI their contribution to the emergent behaviour.
4.1.2 Understanding complexity
At present , det ai l ed, accur at e br ai n si mul at i ons ar e
t he onl y approach t hat coul d allow us to explain why the brain
needs to use many diIIerent ion channels, neurons and synapses, a
spect rum oI re cept or s, and compl ex dendr i t i c and axona
l ar bori zat i ons, rather than the simplifed, uniIorm types Iound in
many models.
5.1.3 Exploring the role oI dendrites.
This is the only current approach to explore the dendritic object
theory, which proposes that three-
dimensional voltage objects are generated continuously across dendritic
segments regardless oI the origin oI the neurons, and that spikes are used
to maintain such dendritic objects.
Blue Brain 23
5.1.4 Revealing Iunctional diversity
Mos t model s engi neer a speci fc I unct i on, wher eas a
spect rum oI I unct i ons might be possible with a biologically
based design. Understanding memory storage and retrieval. This
approach oIIers the possibility oI determining the manner in which
representations oI inIormation are imprinted in the circuit Ior storage and
retrieval, and could reveal the part that diIIerent types oI neuron play in
these crucial Iunctions.
5.1.5 Tracking the emergence oI intelligence
This approach oIIers the possibility to re-trace the steps taken by
a network oI neurons in the emergence oI electrical states used to
embody representations oI the organism and its world.
5.1.6 IdentiIying points oI vulnerability
Although the neocortex conIers immense computational
power to mammals, deIects are common, with catastrophic
cognitive eIIects. At present, a detailed model is the only approach
that could produce a list oI the most vulnerable circuit parameters,
revealing likely candidates Ior dysIunction and targets Ior treatment.
5.1.7 Simulating disease and developing treatments
Such si mul at i ons coul d be used t o t est hypot heses I or
t he pat hogenesi s oI neurological and psychiatric diseases, and to
develop and test new treatment strategies.
5.1.8 Providing a circuit design platIorm
Detailed models could reveal powerIul circuit designs that
could be implemented into silicone chips Ior use as intelligence devices
in industry.
Blue Brain 24
5.2 Applications of Blue Brain
5.2.1 Gathering and Testing 100 Years oI Data
The most immediate beneft is to provide a working model into
which the past100 years knowledge about the microstructure and
workings oI the neocortical column can be gat hered and
t est ed. The Bl ue Col umn wi l l t hereI or e al so
produce a vi rt ual l i br ary t o expl ore i n 3D t he mi cr o
ar chi t ect ur e oI t he neo cort ex and acces s al l key research
relating to its structure and Iunction.
5.2.2 Cracking the Neural Code
The Neur al Code re Ier s t o how t he brai n bui l ds
obj ect s usi ng el ect r i cal pat t er ns. In the same way that the neuron
is the elementary cell Ior computing in the brain, the NCC is
the elementary network Ior computing in the neo cortex. Creating
an accurate replica oI the NCC which IaithIully reproduces the emergent
electrical dynamics oI the real microcircuit, is an absolute requirement to
revealing how the neocortex processes, stores and retrieves inIormation.
5.2.3 Understanding Neocortical InIormation Processing
The power oI an accurate simulation lies in the predictions that can
begener at ed about t he neocort ex. Indeed, i t er at i ons bet we
en si mul at i ons and exper iments are essential to build an accurate c
opy oI the NCC. These iterations are therIore expected to reveal the
Iunction oI individual elements (neurons, synapses, ion channels,
receptors), pathways (mono-synaptic, disynaptic,
multisynapticloops)and physiological processes (Iunctionalproperties, lea
rning, reward, goaloreintedbehavior).
Blue Brain 25
5.2.4 A Novel Tool Ior Drug Discovery Ior Brain Disorders
Understanding the Iunctions oI diIIerent elements and pathways oI
the NCC will provide a concrete Ioundation to explore the cellular
and synaptic bases oI a wide spect r um oI neur ol ogi cal and
psychi at r i c di seases. The i mpact oI recept or , i on channel ,
cellular and synaptic defcits could be tested in simulations and the
optimal experimental tests can be determined.
5. 2. 5 A Gl obal Faci l i t y
A soItware replica oI a NCC will allow researchers to
explore hypotheses oI brain Iunction and dysIunction accelerating
research. Simulation runs could determine which parameters should
be used and measured in the experiments. An advanced 2D,3D
and 3D immersive visualization system will allow 'imaging oI
many aspects oI neural dynamics during processing, storage and
retrieval oI inIormation. Such imaging e x p e r i me nt s ma y b e
i mp o s s i b l e i n r e a l i t y o r ma y b e p r o h i b i t i v e l y
e x p e n s i v e t o p e r I o r m.
5.2.6 A Foundation Ior Whole Brain Simulations
With current and envisageable Iuture computer technology
it seems unlikely that a mammalian brain can be simulated with
Iull cellular and synaptic complexity(above the molecular
level). An accurate replica oI an NCC is thereIore required in
order to generate reduced models that retain critical Iunctions and
computational capabilities, which can be duplicated and interconnected to
Iorm neocortical brain regions. Knowledge oI the NCC architecture
can be transIerred to Iacilitate reconstruction oI sub cortical brain
regions.
5.2.7 A Foundation Ior Molecular Modeling oI Brain Function
Blue Brain 26
An accurate cellular replica oI the neocortical column will provide
the frst and essential step to a gradual increase in model
complexity moving towards a molecular l eve l
descr i pt i on oI t he neocort ex
wi t h bi ochemi cal pat hways bei ng si mul at ed. A mol ecul ar
l evel model oI t he NCC wi l l pr ovi de t he subs t r at e I or
i nt er Iaci ng gene expr essi on wi t h t he
net wor k st ruct ure and I unct i on. The NCC l i es at t he
i nt er Iace between the genes and complex cognitive
Iunctions. Establishing this link will allow predictions oI the
cognitive consequences oI genetic disorders and allow reverse
engineering oI cognitive defcits to determine the genetic and molecular
causes. This level oI simulation will become a reality with the most
advanced phase oI Blue Gene development.

Blue Brain 27
CHAPTER 6
ADVANTAGES AND LIMITATIONS
6.1 Advantages
O We can remember things without any eIIort.
O Decision can be made without the presence oI a
person.
O Even aIter the death oI a man his intelligence can be
used.
The activity oI diIIerent animals can
be understood. That means by interpretation oI the electric
impulses Irom the brain oI the animals, their thinking can be
understood easily.
O It would allow the deaI to hear via direct nerve stimulation,
and also be helpIul Ior many psychological diseases. By down
loading the contents oI the brain that was uploaded into the
computer, the man can get rid Irom the madness.
. Limitations
Further, there are many new dangers these technologies
will open. We will be susceptible to new Iorms oI harm.
O We become dependent upon the computer systems.
O Others may use technical knowledge against us.
O Computer viruses will pose an increasingly critical
threat.
The real threat, however, is the Iear that people will have
oI new technologies. That Iear may culminate in a large resistance.
Blue Brain 28
Clear evidence oI this type oI Iear is Iound today with respect to
human cloning.

Blue Brain 29
CHAPTER 7
FUTURE PERSPECTIVE
The synthesis era in neuroscience started with the launch oI
the Human Brain Project and is an inevitable phase triggered by
a critical amount oI Iundamental data. The dat a set does not
need t o be compl et e beI ore such a phase
can begi n. Indeed, i t is essential to guide reductionist research
into the deeper Iacets oI brain structure and Iunction. As
a complement to experimental research, it oIIers
rapid assessment oI the probable eIIect oI a new fnding on
preexisting knowledge, which can no longer be managed
completely by any one researcher. Detailed models will probably
become the fnal Iorm oI databases that are used to organize all
knowledge oI the brain and allow hypothesis testing, rapid diagnoses
oI brain malIunction, as well as development oI treatments Ior
neurological disorders. In short, we can hope to learn a great
deal about brain Iunction and disIunction Irom accurate models oI the
brain .The time taken to build detailed models oI the brain depends
on the level oI detail that is captured. Indeed, the frst version oI
the Blue Column, which has 10,000 neurons, has already been
built and simulated; it is the refnement oI the detailed properties
and calibration oI t he ci rcui t t hat t akes t i me. A model
oI t he ent i re brai n at t he cel l ul ar l evel wi l l
probabl y t ake t he next decade. Ther e i s
no I undament al obst acl e t o model i ng t he brain and it is
thereIore likely that we will have detailed models oI mammalian brains,
including that oI man, in the near Iuture. Even iI overestimated by a
decade or two, this is still just a `blink oI an eye` in relation to the
Blue Brain 30
evolution oI human civilization. As with Deep Blue, Blue Brain will
allow us to challenge the Ioundations oI our understanding oI intelligence
and generate new theories oI consciousness.

Blue Brain 31
CHAPTER 8
CONCLUSION
In concl usi on, we wi l l be abl e t o t r ans Ier our sel ve s
i nt o comput er s at some point. Most arguments against this outcome
are seemingly easy to circumvent. They are either simple minded, or
simply require Iurther time Ior technology to increase. The onl y
ser i ous t hreat s r ai sed a re al so overcome as we not e t he
combi nat i on oI biological and digital technologies

Blue Brain 32
REFERENCES
| 1 | ' En g i n e e r i n g i n Me d i c i ne a n d Bi o l o g y Soc i e t y
, 2 0 0 8 . EMBS2 0 0 8. 3 0
t h
An n u a l International ConIerence oI
the IEEE
|2| Henry Markram, 'The Blue Brain Project, Nature Reviews
Neuroscience 2006February.
|3| Simulated brain closer to thought BBC News 22 April 2009.
|4| 'ProjectMilestones. BlueBrain.http://bluebrain.epf.ch/Jahia/
site/bluebrain/op/edit/pid/19085
|5| Graham-Rowe, Duncan. 'Mission to build a simulated brain
begins, New Scientist, June 2005. pp. 1879-85.
|6| Blue Gene: http://www.research.ibm.com/bluegene
|7| The Blue Brain Project: http://bluebrainproject.epf.ch

Documentation

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Documentation

Enviado por

Direitos autorais:

Formatos disponíveis

Blue Brain 1

Dept. OI InIormation Technology YCET`11

Você também pode gostar