“Memory is built on concept neurones”
This interview was published in the PRBB monthly newspaper, Ellipse.
You can also read an earlier post about his talk here.
Figuring out how the brain works is the obsession of Rodrigo Quian, professor at the University of Leicester (UK). This challenge led him to apply his physics training and a PhD in maths to neuroscience. With the discovery of the “Jennifer Aniston neurone”, or concept cells, it seems we have taken a step towards the understanding of memory.
How can we “see” neurones?
We work on patients with epilepsy requiring hippocampus surgery. As part of this they have electrodes attached to the brain for several hours. This allows us to talk to them and detect how the neurones respond to stimuli we present them with.
Why the Jennifer Aniston neurone?
We did experiments where we showed patients people close to them like relatives and celebrities. The first neurone I found responded to pictures of Jennifer Aniston. It was a shock to discover that somewhere in the brain are neurones that respond in such a specific way to abstract concepts.
Did it only respond to photos?
It responded to various photos of Aniston, images as different in colour and format as we were able to find. The same with her name when written or spoken. Specifically, to the ‘concept’ of Jennifer Aniston. We found neurones that responded to different famous people depending on the person. The only neurones that did not respond were in an autistic patient.
One neurone per concept?
If I could find one neurone that responded to Jennifer Aniston, there must be more because if it was the only one, the probability of me finding it among the thousands of neurones in that area is practically zero. There has to be a network of neurones that encode a concept. These concept cells can quickly generate associations, so there are neurones that respond to two concepts, but they are always related to one another. This is a key mechanism for generating memories. I think they are the building blocks of memory and the link between perception and memory. This is a radically different idea to what was believed until now, that the basis of memory was distributed networks of millions of neurones.
Can you locate complex thoughts like phobias?
Often a complex thought is an association of simple thoughts. My old mentor at Caltech, Christof Koch, said it was necessary to break down the difficult problem of consciousness into related problems that are simpler and easier to attack. The consciousness of self is a very complex thing. One must first understand how the flow of consciousness works. That is, that one thing makes me think about another thing and that about another and so on. This can be studied in the neurones generating associations between two concepts and, from the moment we have made this association, we can see if the neurone also responds to the association and encodes it. In a few tests we have found that these concept cells begin to respond to the association we have created.
What other experiments are you working on?
We want to know if neurone response changes according to the presentation of the stimulus, for example the exposure time to the photos. The results demonstrate that neural response is closely related to the conscious perception of the patient. That is, if the patient believes that he has seen something, then the neurone is activated. In fact, it is even possible to predict beforehand when neurones will be activated and know what image a patient is looking at only from the neurone records.