The results of this study are ground-breaking in two ways: the technique itself, that enabled them to investigate neural competition, and the fact that they found that competitive behaviour between the neurons was similar to that in human perception. The findings are published in the Journal of Neuroscience.
“In our daily life our eyes receive a lot of information, usually very noisy and ambiguous”, Kogo explains. “This gives conflicting information. The brain has to process every shape, colour, sound, smell and much more and choose the correct signals to establish a coherent interpretation of the world around them.” A pair of two competing neurons represents the simplest possible neural competition. Kogo found that the phenomenon called bi-stable activity emerges in the pair when activated. This has been assumed to underlie one of the optical illusions called bi-stable perception – a phenomenon that perception researchers have been studying for decades.
Switching choices
This internal and usually unconscious skirmish in our brain comes to the surface with optical illusions, such as the spinning ballerina or Necker Cube, where two interpretations of the image are equally possible. The ballerina randomly seems to change direction (see video) and the brain struggles to decide on the orientation of the cube (see illustration). Kogo: “In bi-stable perception your brain starts to switch back and forth between two possible interpretations of images. The brain has to make a choice, but this can instantly change.”
Research into this phenomenon is usually done at the level of human perception, where memories, experience and environmental input also influence a choice. However, it has been assumed that neural competition underlies the phenomenon. “But this never has actually been studied thoroughly at a neuron level.”
Connecting neurons
To properly measure this, a connected and competing pair of neurons need to be recorded. Selecting and measuring a connected pair with competition at this microscopic level is complicated, which pushed Kogo to think out of the box. “When I was struggling with that, I got the crazy idea to pick two neurons randomly from the visual region of a mouse brain tissue and to make them connected. I remembered a technique developed in the 1980’s that simulates neural mechanisms and render them to real-life neurons. So I used their approach and constructed a competing neuron pair.”
Once electrodes were attached to the brain cells and the computer-modelled circuit successfully launched their connection, Kogo could start recording. The pair immediately started to show bi-stability: “We could see one neuron spiking and that the other one was suppressed. And few seconds later it turned around. That is exactly what the neural competition mechanism has been assumed to show during bi-stable perception.” Kogo compared the dynamics of this bi-stability at neuron level with the known dynamics of bi-stable perception and found surprising similarity between the two: “It is surprising as they come from two very different systems, the simplest competition mechanism and the complex human visual system.”
Next steps
The technique can find a way into brain-computer interfaces or integration with AI. Or into medical applications, to investigate treatment and medication using cultured neural networks derived from patient induced pluripotent stem cells.
The other direction of future research is to develop a system to study groups of neurons. To make that happen, Kogo is expanding the approach with optical measurement and stimulation techniques. “Attaching electrodes to neurons is complicated, especially if you’re aiming to measure groups simultaneously. So I’m trying to develop a system that can record from neurons and stimulate them completely optically, without the use of electrodes. In this way, we can study the dynamics of groups of neurons, and manipulate interactions among them.”
Image: Natasha Connell via Unsplash.