NEWRON vol I issue II (2-16-07)

Edited by: Natan Davidovics

Chipping In
Scientific American (02/07) Vol. 296, No. 2, P. 18; Griffith, Anna

http://web.ebscohost.com/ehost/detail?vid=6&hid=5&sid=5b2cd464-20bd-405f-8ff1-6808e53373f7%40sessionmgr9
(Full article is included below if the link gives you trouble)

Scientists are working on a "brain chip" designed as a memory aid, especially in cases where the patient has suffered neural damage. A team from the University of Southern California is getting ready for live tests of a neural prosthesis in brain-damaged rats, which may be carried out in the spring. In January 2006 USC researcher Theodore W. Berger and his team engineered a silicon chip that imitates biological neurons in tissue slices of rat hippocampus as a replacement for a section of brain that was surgically removed, and that returns function by processing neural input into appropriate output with a 90 percent rate of accuracy.



The Mind Chip
New Scientist (02/03/07) Vol. 193, No. 2589, P. 28; Fox, Douglas

Click Here to View Full Article (may not work so I included an extended summary)

A notable achievement in computer vision has been made by researcher Kwabena Boahen and colleagues at the University of Pennsylvania in Philadelphia, who constructed a device that can see via chips that physically imitate the electrical activity of neurons in the primary visual cortex. "I want to figure out how the brain works in a very nuts-and-bolts way," explains Boahen. "I want to figure it out such that I can build it." Boahen aims to top his accomplishment of building an artificial retina with the creation of an artificial cerebral cortex through the generalization of the chip's function; such a breakthrough may be an important step in helping restore neural function to people impaired by disease or injury. The concept of the artificial neuron as a technology for enabling brain-like computing in real time was first suggested in the late 1980s by California Institute of Technology scientist Carver Mead, who discovered he could build such circuits by having digital processors use transistors in their analog amplifier phase instead of their on/off switching phase. Mounted on the surface of Boahen's artificial retina are photosensitive transistors that translate incoming light into analog voltages with a value determined by the light's intensity and which last for as long as the light is beamed onto the transistors; these transmissions are routed to the artificial retina neurons where motion and regions of contrast are recognized, signaling the edges of objects in the image. Processing information about edges and movement in the visual scene is carried out by the low-power visual cortex chips, which build object outlines out of the signals. A successful cortical implant will have to be able to mimic the plasticity of the brain's neural network, in which connections between neurons are created and adapted on the fly.

Loving with all your ... brain

Cnn.com, Elizabeth Cohen

Loving with all your ... brain - CNN.com*

hese areas of the brain, while little known to most people, are helping scientists explain the physiological reasons behind why we feel what we feel when we fall in love. By studying MRI brain scans of people newly in love, scientists are learning a lot about the science of love: Why love is so powerful, and why being rejected is so horribly painful.

BRAIN CHIP FOR MEMORY REPAIR CLOSES IN ON LIVE TESTS

Supplementing the human brain with computer power has been a staple of science fiction. But in fact, researchers have taken several steps in melding minds with machines, and this spring a team from the University of Southern California may replace damaged brain tissue in rats with a neural prosthesis.

For the past few years, researchers have demonstrated the ability to translate another creature's thoughts into action. In 2000 neurologist Miguel Nicolelis of Duke University wired a monkey with electrodes so that its thoughts could control a robotic arm. Brain-machine interfaces developed by Niels Birbaumer, a neuroscientist at the University of Tübingen in Germany, already help some paralyzed patients move a computer cursor with their brain waves to select letters for writing a message.

Theodore W. Berger and his U.S.C. colleagues have developed the first brain-machine interface to communicate back to the brain. Last January they used a silicon chip to mimic biological neurons in tissue slices of rat hippocampus, the hub for memory sorting and storage. The chip replaced a surgically removed section of the hippocampus and restored function by processing incoming neural signals into appropriate output with 90 percent accuracy.

The biomedical engineers had been on the verge of testing a chip in hippocampal slices for several years, but roadblocks slowed work. Existing electrode array technology would not function well in tissue slices, forcing the researchers to construct their own. Cutting the hippocampus slices just right to keep the neural pathway intact was also difficult.

Because building the one-millimeter-square chip costs tens of thousands of dollars and takes several months, the planned spring test will actually rely on a model of that chip --specifically a larger, reprogrammable device linked to a computer called a field programmable gate array (FPGA). The FPGA will allow investigators to easily test and modify their new mathematical model of neural communication for living rats before committing it to a chip. Sam Deadwyler, a professor of physiology and pharmacology at Wake Forest University and a collaborator in the study, has demonstrated that stimulating the hippocampus of living rats with a certain pattern of activity can increase performance on a memory task, such as recalling which lever will dispense water. In a few months he will use the FPGA mathematical model to predict hippocampal activity. If the model is correct, the artificial implant should restore memory for such tasks in rats with drug-induced amnesia.

For more complicated animal models, U.S.C. physicist Armand Tanguay suggests a multichip module to facilitate the transition. Light beams would transmit signals between neuron units on multiple chip layers. Unlike wires, light beams pass directly through one another without interference, allowing for many more interconnections. The result: a web of light between silicon chips mimicking a dense neural network.

"Many challenges will be encountered as the researchers move from in vitro to in vivo studies in the rat," says Grace Peng, a program director at the National Institutes of Health's Division of Discovery Science and Technology. In fact, the team is not quite sure what to expect once it goes to live animals. Avoiding rejection by the immune system might mean anchoring cell adhesion molecules to the chip so that the surface of the implant looks like tissue, says U.S.C. chemist Mark Thompson. Neural plasticity, or the brain's ability to reorganize its connections, could also pose a problem by preventing the formation of stable connections between the neurons and chip. "In other application areas such as motor control or perception, plasticity and adaptability of the brain usually facilitate the effects of artificial interfaces," Peng notes optimistically.

One other possible concern, if such implants make it to human testing: Might bypassing damaged neurons in the hippocampus also bypass connections with other areas of the brain that filter what we remember? In other words, would the brain become unable to purge memories? If so, that would make the implant a truly unforgettable device.

DIAGRAM: HEADSTRONG: Implants communicating with the hippocampus might someday restore or improve memory, if such devices succeed in rat tests.

~~~~~~~~

By Anna Griffith

Anna Griffith is based in Chico, Calif.