German boffins snap hi-res rat brain activity

Unique brain-chip interface

German scientists have taken a movie of a rat's mental clockworks using a unique brain-chip interface they say could pave the way for mind-controlled prosthetic limbs.

The 20ms movie, which can be slowed down by a factor of 1,000 if your powers of concentration are not sharp enough to catch it, captures brain activity at drastically higher resolutions than have ever been caught before, said Peter Fromherz of the Department of Membrane and Neurophysics at the Max Plank Institute for Biochemistry.

For this experiment, the details of which are yet to be published, a microscopic sliver of rat brain was peeled on to a CMOS chip that reads brainwaves across a 50 nanometre gap. The chip packs 16,000 transistors, spaced just 8 micrometres apart, onto a 1mm x 1mm insualted wafer developed in collaboration with Infinion Technologies (formerly Siemens Semiconductor Group).

Previous experiments using microelectrodes have achieved resolutions of 64 microelectrodes spaced 200 micrometers apart, he said.

"We could detect the dynamics of the brain and could see the electrical waves going through it," said Fromherz.

A 1mm x 1mm brain slice of a rat hippocampus, 5ms after stimulation. The blue and red signals are voltages on the chip that are created by synaptic activity (blue negative, red positive).

The chip records voltages. There are two phases: one area of the hippocampus sends a fast signal (action potential) along a fibre; then a second area of the hippocampus responds with slow synaptic signals. The picture shows the second phase.

Fromherz's previous experiments using the same chip technology have involved examining the ion channels working within an individual snail neuron, as well as the dynamics of small networks of up to 20 cells.

The rat brain movie captures the activity of a brain slice containing thousands of neurons. Fromherz said: "We see signals not from individual neurons, but from averages. The action potential spreads along and you can see the spatial distribution of the synaptic correlates."

"This tells us where the different activities of the brain are localised and the correlation in space and time, which will be important to study the memory of slices," he added.

Fromherz distinguished his experiment from maps of brain activity taken using fluorescent dyes both because of their toxicity, which means they cannot be repeated with the frequency of the transistor approach, and the fact that dyes are still not as sensitive.

Toxicity could also be an important distinction between the transistor approach and the conventional method of brain-chip interface that uses planar metal microelectrodes.

As the transistors are insulated, and take their measurements from the brain's electrical fields, researchers can study brain physics without any interference or potential toxicity that might cause problems for the kind of man-machine devices that are currently being developed by medical researchers.

Microelectrodes stimulate neurons directly with a voltage, which in combination with the brain material's water and the electrode's metal, cause chemical side effects that are still little understood.

If the chemistry can be understood better, said Fromherz, the safety of medical devices such as brain-controlled prosthetic limbs can be guaranteed.

He is also preparing experiments with a chip that provide an alternative method of stimulation, via capacitors and has a chip in production that has 400 of them.

The transistor-capacitor technology is still at an early stage of development, and it may be 20 years before it can be used in prosthetic devices. However, some help might come Fromherz's way through contacts he has begun forming through CIMIT, a consortium of hospitals and colleges that includes MIT and Harvard Medical School, with the aim of seeing this technology find real-world applications. ®