Hersenen versus computer

quote:

Simple mathematical computations underlie brain circuits

Discovery of how some neurons inhibit others could shed light on autism, other neurological disorders

Optical mapping of functional inhibition by neurons

MIT neuroscientists report that two major classes of brain cells repress neural activity in specific mathematical ways: One type subtracts from overall activation, while the other divides it.

The brain has billions of neurons, arranged in complex circuits that allow us to perceive the world, control our movements and make decisions. Deciphering those circuits is critical to understanding how the brain works and what goes wrong in neurological disorders.

“These are very simple but profound computations,” says Mriganka Sur, the Paul E. Newton Professor of Neuroscience and senior author of the Nature paper. “The major challenge for neuroscience is to conceptualize massive amounts of data into a framework that can be put into the language of computation. It had been a mystery how these different cell types achieve that.”

A fine balance

There are hundreds of different types of neuron in the brain; most are excitatory, while a smaller fraction are inhibitory. All sensory processing and cognitive function arises from the delicate balance between these two influences. Imbalances in excitation and inhibition have been associated with schizophrenia and autism.

“There is growing evidence that alterations in excitation and inhibition are at the core of many subsets of neuropsychiatric disorders,” says Sur, who is also the director of the Simons Center for the Social Brain at MIT. “It makes sense, because these are not disorders in the fundamental way in which the brain is built. They’re subtle disorders in brain circuitry and they affect very specific brain systems, such as the social brain.”

In the new Nature study, the researchers investigated the two major classes of inhibitory neurons. One, known as parvalbumin-expressing (PV) interneurons, targets neurons’ cell bodies. The other, known as somatostatin-expressing (SOM) interneurons, targets dendrites — small, branching projections of other neurons. Both PV and SOM cells inhibit a type of neuron known as pyramidal cells. (See “How the brain’s stem cells find out when to make new neurons” for the role of PV neurons in inhibiting neurogenesis.)

Observing reactions in activated neurons

[+]First, the researchers genetically programmed either PV or SOM cells in mice to produce a light-sensitive protein called channelrhodopsin.

To study how these neurons exert their influence, the researchers had to develop a way to specifically activate PV or SOM neurons, then observe the reactions of the target pyramidal cells, all in the living brain.

When embedded in neurons’ cell membranes, channelrhodopsin controls the flow of ions in and out of the neurons, altering their electrical activity. This allows the researchers to stimulate the neurons by shining light on them.

[+]

The team combined this with calcium imaging inside the target pyramidal cells. Calcium levels reflect a cell’s electrical activity, allowing the researchers to determine how much activity was repressed by the inhibitory cells.

“Up until maybe three years ago, you could only just blindly record from whatever cell you ran into in the brain, but now we can actually target our recording and our manipulation to well-defined cell classes,” Runyan says.

Decoding computations in a brain circuit: subtraction vs. division

In this study, the researchers wanted to see how activation of these inhibitory neurons would influence how the brain processes visual input — in this case, horizontal, vertical or tilted bars. When such a stimulus is presented, individual cells in the eye respond to points of light, then convey that information to the thalamus, which relays it to the visual cortex. The information stays spatially encoded as it travels through the brain, so a horizontal bar will activate corresponding rows of cells in the brain.

Those cells also receive inhibitory signals, which help to fine-tune their response and prevent overstimulation. The MIT team found that these inhibitory signals have two distinct effects: Inhibition by SOM neurons subtracts from the total amount of activity in the target cells, while inhibition by PV neurons divides the total amount of activity in the target cells.

“Now that we finally have the technology to take the circuit apart, we can see what each of the components do, and we found that there may be a profound logic to how these networks are naturally designed,” Wilson says.

These two types of inhibition also have different effects on the range of cell responses. Every sensory neuron responds only to a particular subset of stimuli, such as a range of brightness or a location. When activity is divided by PV inhibition, the target cell still responds to the same range of inputs. However, with subtraction by SOM inhibition, the range of inputs to which cells will respond becomes narrower, making the cell more selective.

“Conceptually, inhibition by subtraction and division is a very nice distinction,” says Tony Zador, a professor of neuroscience at Cold Spring Harbor Laboratory who was not involved in the research. “It’s a joy when something as theoretically appealing as division and subtraction actually maps onto the physiological substrate in such a fundamental way.”

Increased inhibition by PV neurons also changes a trait known as the response gain — a measurement of how much cells respond to changes in contrast. Inhibition by SOM neurons does not alter the response gain.

The researchers believe this type of circuit is likely repeated throughout the brain and is involved in other types of sensory perception, as well as higher cognitive functions.

Sur’s lab now plans to study the role of PV and SOM inhibitory neurons in a mouse model of autism. These mice lack a gene called MeCP2, giving rise to Rett Syndrome, a rare disease that produces autism-like symptoms as well as other neurological and physical impairments. Using their new technology, the researchers plan to test the hypothesis that a lack of neuronal inhibition underlies the disease.

quote:

Wetenschappers herstellen zicht door ontrafelen neurale code

Door Bauke Schievink, woensdag 15 augustus 2012 14:49, views: 22.766

Onderzoekers claimen een doorbraak in de ontwikkeling van netvliesprotheses. Met de ontcijfering van de signalen die de ogen naar het brein sturen, wordt het wellicht mogelijk om het zicht bij blinden tot op normaal niveau te herstellen.

De wetenschappers, werkzaam aan de Cornell University in New York, richten zich in hun onderzoek op de aansturing van de oogzenuw, die signalen van het oog naar de hersenen stuurt, waar het uiteindelijke beeld wordt samengesteld. Normaal gesproken wordt licht opgevangen op het netvlies, oftewel de retina, en vervolgens omgezet naar elektrische zenuwpulsen, maar bij veel blinden is dit systeem stuk. Er zijn al protheses die de functie van het netvlies kunnen overnemen, maar deze zijn beperkt omdat de gegenereerde elektrische signalen niet volledig compatibel zijn: de kunstmatige prikkels wijken te veel af van die van de lichaamseigen 'encoder'.

Een door de wetenschappers ontwikkeld algoritme, dat ervoor zorgt dat een prothese invallend licht omzet naar de 'native' neurale code, kan een aanzienlijke verbetering betekenen voor het herstellen van zicht. De oogzenuw wordt zo aangestuurd met conventionele biologische signaalpatronen. Het systeem werkt al bij muizen, waar zicht tot vrijwel het normale niveau werd hersteld, maar de neurale code van de ogen van apen is inmiddels ook 'gekraakt'. Omdat signalen afkomstig van het netvlies bij apen en mensen min of meer gelijk zijn, kan het systeem mogelijk al snel in menselijke protheses worden toegepast.

Protheses om zicht mee te herstellen bij mensen met beschadigde netvliezen zijn al jaren geleden ontwikkeld. Met de eerste modellen kregen patiënten slechts zeer beperkt zicht, maar nieuwere modellen kunnen steeds meer overgebleven zenuwcellen in het oog stimuleren, waardoor er een steeds hogere resolutie kan worden waargenomen. Volgens de wetenschappers kan hun algoritme in combinatie met een prothese die ongeveer 9800 zenuwcellen stimuleert, ervoor zorgen dat zicht tot normaal niveau wordt hersteld. Klinische onderzoeken, waarmee de wetenschappers snel hopen te beginnen, moeten uitwijzen of dit ook inderdaad het geval is.

Recentelijk zijn ook alternatieve implantaten ontwikkeld om het zicht bij blinden te herstellen. In plaats van met conventioneel binnenvallend licht te werken, hebben wetenschappers een systeem ontwikkeld dat extern licht met lasers reproduceert. Dit zorgt voor een grotere lichtsterkte en dus een sterkere stimulatie. Het is mogelijk dat de door Cornell University ontwikkelde encoder ook hier voor verbeteringen kan zorgen.

quote:

Onderzoekers simuleren brein met 2,5 miljoen neuronen

Door Willem de Moor, vrijdag 30 november 2012 11:29, views: 15.154

Onderzoekers van een Canadese universiteit hebben een computersimulatie ontwikkeld van een menselijk brein. Hun model bestaat uit tweeënhalfmiljoen neuronen, een fractie van een echt brein. Toch kan het model relatief complexe taken volbrengen.

De neurowetenschappers van de Waterloo-universiteit in Canada hebben, onder leiding van Chris Eliasmith, het kunstmatige brein ontwikkeld. Spaun, een afkorting voor Semantic Pointer Architecture Unified Network, bestaat uit gesimuleerde neuronen die met elkaar communiceren middels elektriciteit en chemische moleculen, net als de actiepotentialen en neurotransmitters in biologische hersenen. De virtuele neuronen zijn in subnetwerken onderverdeeld die de functies en structuren van echte hersenen nabootsen.

Spaun krijgt input via een klein 28 bij 28 pixels tellend scherm, waarop reeksen getallen worden getoond. De hersensimulatie kan deze getallen niet alleen succesvol herkennen, maar de getoonde reeksen ook opslaan en patronen in de getallen achterhalen. Met een gesimuleerde robotarm kan een lijstje gereproduceerd worden, maar kunnen ook ontbrekende getallen in een reeks worden 'opgeschreven'. In totaal kan Spaun acht eenvoudige taken uitvoeren en presteert daarbij vergelijkbaar als een mens, die over tientallen miljarden neuronen beschikt.

Net als een menselijk brein, vertoont Spaun ook vergelijkbare fouten: zo kan het kunstmatige brein cijfers aan het begin of eind van een reeks beter onthouden dan die in het midden. Het model kan ook gebruikt worden om veroudering van hersenen te simuleren: door virtuele neuronen uit te schakelen, verandert de output van Spaun op een vergelijkbare manier als bij een ouder wordend brein. Spaun heeft wel een flink nadeel ten opzichte van menselijke hersenen: het simuleren van een seconde breinactiviteit kost enkele uren met een op 2,5GHz geklokt quadcoresysteem met 24GB geheugen.

Spaun brain activity and decoding for sample tasks

quote:

The Human Brain Project Wins Top European Science Funding

28.01.13 - The European Commission has officially announced the selection of the Human Brain Project (HBP) as one of its two FET Flagship projects. The new project will federate European efforts to address one of the greatest challenges of modern science: understanding the human brain.

The goal of the Human Brain Project is to pull together all our existing knowledge about the human brain and to reconstruct the brain, piece by piece, in supercomputer-based models and simulations. The models offer the prospect of a new understanding of the human brain and its diseases and of completely new computing and robotic technologies. On January 28, the European Commission supported this vision, announcing that it has selected the HBP as one of two projects to be funded through the new FET Flagship Program.

Federating more than 80 European and international research institutions, the Human Brain Project is planned to last ten years (2013-2023). The cost is estimated at 1.19 billion euros. The project will also associate some important North American and Japanese partners. It will be coordinated at the Ecole Polytechnique Fédérale de Lausanne (EPFL) in Switzerland, by neuroscientist Henry Markram with co-directors Karlheinz Meier of Heidelberg University, Germany, and Richard Frackowiak of Centre Hospitalier Universitaire Vaudois (CHUV) and the University of Lausanne (UNIL).

The Swiss Contribution
Switzerland plays a vital role in the Human Brain Project. Henry Markram and his team at EPFL will coordinate the project and will also be responsible for the development and operation of the project’s Brain Simulation Platform. Richard Frackowiak and his team will be in charge of the project’s medical informatics platform; the Swiss Supercomputing Centre in Lugano will provide essential supercomputing facilities. Many other Swiss groups are also contributing to the project. Through the ETH Board, the Swiss Federal Government has allocated 75 million CHF (approximately 60 million Euros) for the period 2013-2017, to support the efforts of both Henry Markram’s laboratory at EPFL and the Swiss Supercomputing Center in Lugano. The Canton of Vaud will give 35 million CHF (28 million Euros) to build a new facility called Neuropolis for in silico life science, and centered around the Human Brain Project. This building will also be supported by the Swiss Confederation, the Rolex Group and third-party sponsors.

The selection of the Human Brain Project as a FET Flagship is the result of more than three years of preparation and a rigorous and severe evaluation by a large panel of independent, high profile scientists, chosen by the European Commission. In the coming months, the partners will negotiate a detailed agreement with the Community for the initial first two and a half year ramp-up phase (2013-mid 2016). The project will begin work in the closing months of 2013.

---

A scientific portrait of the Human Brain Project

The Human Brain Project will provide new tools to help understand the brain and its fundamental mechanisms and to apply this knowledge in future medicine and computing.

Central to the Human Brain Project is Information and Computing Technology (ICT). The project will develop ICT platforms for neuroinformatics, brain simulation and supercomputing that will make it possible to federate neuroscience data from all over the world, to integrate the data in unifying models and simulations of the brain, to check the models against data from biology and to make them available to the world scientific community. The ultimate goal is to allow neuroscientists to connect the dots leading from genes, molecules and cells to human cognition and behavior.

A novel medical informatics platform will federate clinical data from around the world, allowing medical researchers to unlock the clinically valuable information they contain and to incorporate it in computer models of disease. The goal is to develop techniques for the objective diagnosis of the brain’s diseases, to understand their underlying mechanisms and to speed up the search for new treatments.

Finally, the HBP will build new platforms for “neuromorphic computing” and “neurorobotics”, allowing researchers to develop new computing systems and robots based on the architecture and circuitry of the brain. The new systems will use detailed knowledge of the brain to address critical problems facing future computing technology: energy efficiency, reliability, the huge difficulties involved in programming very complex computing systems.

The HBP will fund independent scientists to use the new platforms for their own research, reserving a substantial part of its budget for this purpose. In brief, the HBP will create a CERN for the brain.

The Human Brain Project - Video Overview

quote:

The ultimate goal is to allow neuroscientists to connect the dots leading from genes, molecules and cells to human cognition and behavior.

quote:

Obama seeking to boost study of human brain

The Obama administration is planning a decade-long scientific effort to examine the workings of the human brain and build a comprehensive map of its activity, seeking to do for the brain what the Human Genome Project did for genetics, The New York Times reports.

The project, which the administration has been looking to unveil as early as March, will include federal agencies, private foundations, and teams of neuroscientists and nanoscientists in a concerted effort to advance the knowledge of the brain’s billions of neurons and gain greater insights into perception, actions and, ultimately, consciousness.

Scientists with the highest hopes for the project also see it as a way to develop the technology essential to understanding diseases like Alzheimer’s and Parkinson’s, as well as to find new therapies for a variety of mental illnesses.

Moreover, the project holds the potential of paving the way for advances in artificial intelligence.

The project, which could ultimately cost billions of dollars, is expected to be part of the president’s budget proposal next month. And, four scientists and representatives of research institutions said they had participated in planning for what is being called the Brain Activity Map project [apparently a planned but unannounced NIH project].

In his State of the Union address, President Obama cited brain research as an example of how the government should “invest in the best ideas.”

“Every dollar we invested to map the human genome returned $140 to our economy — every dollar,” he said. “Today our scientists are mapping the human brain to unlock the answers to Alzheimer’s. They’re developing drugs to regenerate damaged organs, devising new materials to make batteries 10 times more powerful. Now is not the time to gut these job-creating investments in science and innovation.”

Scientists involved in the planning said they hoped that federal financing for the project would be more than $300 million a year, which if approved by Congress would amount to at least $3 billion over the 10 years. The Human Genome Project cost $3.8 billion.

But a group of nanotechnologists and neuroscientists say they believe that technologies are at hand to make it possible to observe and gain a more complete understanding of the brain, and to do it less intrusively. In June in the journal Neuron,six leading scientists proposed pursuing a number of new approaches for mapping the brain.

One possibility is to build a complete model map of brain activity by creating fleets of molecule-size machines to noninvasively act as sensors to measure and store brain activity at the cellular level. The proposal envisions using synthetic DNA as a storage mechanism for brain activity.

“Not least, we might expect novel understanding and therapies for diseases such as schizophrenia and autism,” wrote the scientists, who include Dr. Church; Ralph J. Greenspan, the associate director of the Kavli Institute for Brain and Mind at the University of California, San Diego; A. Paul Alivisatos, the director of the Lawrence Berkeley National Laboratory; Miyoung Chun, a molecular geneticist who is the vice president for science programs at the Kavli Foundation; Michael L. Roukes, a physicist at the California Institute of Technology; and Rafael Yuste, a neuroscientist at Columbia University.

The initiative will be organized by the Office of Science and Technology Policy, according to scientists who have participated in planning meetings.

The National Institutes of Health, the Defense Advanced Research Projects Agency and the National Science Foundation will also participate in the project, the scientists said, as will private foundations like the Howard Hughes Medical Institute in Chevy Chase, Md., and the Allen Institute for Brain Science in Seattle.

A meeting held on Jan. 17 at the California Institute of Technology was attended by the three government agencies, as well as neuroscientists, nanoscientists and representatives from Google, Microsoft, and Qualcomm. According to a summary of the meeting, it was held to determine whether computing facilities existed to capture and analyze the vast amounts of data that would come from the project. The scientists and technologists concluded that they did.

quote:

How the brain quickly rebounds from injuries

Scientists at Carnegie Mellon University‘s Center for Cognitive Brain Imaging (CCBI) have used a new combination of neural imaging methods to discover exactly how the human brain adapts to injury.

When one brain area loses functionality, a “back-up” team of secondary brain areas immediately activates, replacing not only the unavailable area but also its confederates (connected areas), the research shows.

For the study, Robert Mason, senior research psychologist at CMU, and Chantel Prat, assistant professor of psychology at the University of Washington, used functional magnetic resonance imaging (fMRI) to study precisely how the brains of 16 healthy adults adapted to the temporary incapacitation of the Wernicke area, the brain’s key region involved in language comprehension.

Applying TMS to momentarily disable language comprehension

They applied repetitive transcranial magnetic stimulation (rTMS) to temporarily disable the Wernicke area in the participants’ brains, during an fMRI scan.

The participants, while in the fMRI scanner, were performing a sentence comprehension task before, during and after the rTMS was applied. Normally, the Wernicke area is a major player in sentence comprehension.

The research found that as the brain function in the Wernicke area decreased following the application of rTMS, a “back-up” team of secondary brain areas immediately became activated and coordinated, allowing the individual’s thought process to continue with no decrease in comprehension performance.

“The human brain has a remarkable ability to adapt to various types of trauma, such as traumatic brain injury and stroke, making it possible for people to continue functioning after key brain areas have been damaged,” said Marcel Just, the D. O. Hebb Professor of Psychology at CMU and CCBI director.

“It is now clear how the brain can naturally rebound from injuries. It gives us indications of how individuals can train their brains to be prepared for easier recovery. The secret is to develop alternative thinking styles, the way a switch-hitter develops alternative batting styles. Then, if a muscle in one arm is injured, they can use the batting style that relies more on the uninjured arm.”

quote:

Computer Chip Based On Human Brain Developed

Today's computing chips are incredibly complex and contain billions of nano-scale transistors, allowing for fast, high-performance computers, pocket-sized smartphones that far outpace early desktop computers, and an explosion in handheld tablets.

Despite their ability to perform thousands of tasks in the blink of an eye, none of these devices even come close to rivaling the computing capabilities of the human brain. At least not yet. But a Boise State University research team could soon change that.

Electrical and computer engineering faculty Elisa Barney Smith, Kris Campbell and Vishal Saxena are joining forces on a project titled "CIF: Small: Realizing Chip-scale Bio-inspired Spiking Neural Networks with Monolithically Integrated Nano-scale Memristors."

Team members are experts in machine learning (artificial intelligence), integrated circuit design and memristor devices. Funded by a three-year, $500,000 National Science Foundation grant, they have taken on the challenge of developing a new kind of computing architecture that works more like a brain than a traditional digital computer.

"By mimicking the brain's billions of interconnections and pattern recognition capabilities, we may ultimately introduce a new paradigm in speed and power, and potentially enable systems that include the ability to learn, adapt and respond to their environment," said Barney Smith, who is the principal investigator on the grant.

The project's success rests on a memristor -- a resistor that can be programmed to a new resistance by application of electrical pulses and remembers its new resistance value once the power is removed. Memristors were first hypothesized to exist in 1972 (in conjunction with resistors, capacitors and inductors) but were fully realized as nano-scale devices only in the last decade.

One of the first memristors was built in Campbell's Boise State lab, which has the distinction of being one of only five or six labs worldwide that are up to the task.

The team's research builds on recent work from scientists who have derived mathematical algorithms to explain the electrical interaction between brain synapses and neurons.

"By employing these models in combination with a new device technology that exhibits similar electrical response to the neural synapses, we will design entirely new computing chips that mimic how the brain processes information," said Barney Smith.

Even better, these new chips will consume power at an order of magnitude lower than current computing processors, despite the fact that they match existing chips in physical dimensions. This will open the door for ultra low-power electronics intended for applications with scarce energy resources, such as in space, environmental sensors or biomedical implants.

Once the team has successfully built an artificial neural network, they will look to engage neurobiologists in parallel to what they are doing now. A proposal for that could be written in the coming year.

Barney Smith said they hope to send the first of the new neuron chips out for fabrication within weeks.

This material is based upon work supported by the National Science Foundation under Grant No. CCF-1320987 to Boise State University. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.