‘My dream? A prosthesis that lets the blind see again’

Interview with Pieter Roelfsema

How is it possible that you see what you see? And that you see it at a single glance? Your eye jumps to a new location every 300 milliseconds, on the basis of information analyzed by the visual system. Pieter Roelfsema investigates how different brain regions collaborate to construct a correct image at super-high speed.

‘If something forms an entity it gets our attention, which makes it stand out’

Seeing consists of two phases, Roelfsema explains. ‘During the first phase the image is relayed, in 0.1 seconds, from the lower to the higher brain areas. The cells behave like simple detectors for colors, shapes, depths and movements. During the second phase, the thought phase, the higher brain areas give feedback information to the lower areas. We then establish what something is by determining which parts form an entity and which bits do not belong to this entity.’ In this way we are able to not only distinguish a knife from a fork on a table but also to plan our grasping movement.

For this so-called ‘perceptual grouping process’ the neurons involved must fire synchronously, was the theory put forth by Roelfsema in his dissertation. But when he tested his theory in practice, this synchronization did not take place. ‘That was a real blow. I thought it meant the end of my career,’ he says. But what animal experiments did show, was a strengthened activity of the nerve cells. ‘If something forms an entity it gets our attention, which makes it stand out. We push aside anything that is unimportant ‘noise’.’

Roelfsema is currently investigating in rhesus monkeys how the brain manages to do this. ‘We place electrodes in various brain areas, which enables us to measure the signals emitted by nerve cells. In this way we can see which cells in which areas collaborate with each other. We are beginning to understand how this works exactly, and expect that we will be able to see what goes wrong and to come up with a way to repair it. For instance in people with Alzheimer or schizophrenia. They suffer from disturbed interactions between various areas in their brain.’ But Roelfsema also sees beyond that. ‘I think that people who have gone blind may be able to regain a simple form of sight through artificial stimulation of brain cells in the primary visual cortex. They will see a very rough image, as if the pixel density is very low. But the more pixels we are able to create, the more detailed the image will become.’ He expects to be able to deliver the proof of principle for such a prosthesis for the cortex within five years.