Analog IC that mimics a single synapse (via Guy Rachmuth and MIT)
Modeling the human brain into a usable computer is the basis for much research to date. IBM is attempting to build a computer system that rivals the human mind through brute force, one for one, copying the 100 billion neurons and 100 trillion synapses. Statistically so, the computer will be similar, but could it learn like a brain?
MIT and Harvard have modeled a new IC to follow how the brain learns. In the team's chip, 400 transistors simulate a single synapse. This acts as a digital representation of the pathway that nerve impulses from an axon travels before reaching neurons. More specifically, the chip emulates the ion channels. Neurons change their cell characteristics by releasing neurotransmitters (glutamate or GABA) which bind the receptors of postsynaptic cell membranes. This opens or closes the ion channels, which changes the cell's electrical potential.
The team was most concerned with two features of the synaptic connections, long-term depression (LTD) and long-term potentiation (LTP). LTD reduces the synaptic efficiency for periods of time, while LTP enhances. These processes are controlled by the flow of charged atoms (ie: sodium, calcium, potassium) through ion channels. The MIT chip mimics these processes closely in an analog fashion.
MIT-Harvard Division of Health Sciences and Technology explained their achievement further, "We now have a way to capture each and every ionic process that’s going on in a neuron. [The IC is a] significant advance in the efforts to incorporate what we know about the biology of neurons and synaptic plasticity onto CMOS [complementary metal-oxide-semiconductor] chips. If you really want to mimic brain function realistically, you have to do more than just spiking. You have to capture the intracellular processes that are ion channel-based. We can tweak the parameters of the circuit to match specific ion channels."
Aside from using this technology study brain function from a different perspective, build communication networks between the brain and artificial prosthetics, or step closer to a true artificial intelligence. If this team could only partner with IBM's SyNAPSE program, and they could both catalyst their way to another level.
Cabe
Other emulated mind news:
