As you read this article your eyes will move so the words fall on the central part of your vision. This region is called the fovea and it has excellent resolution when compared to your peripheral vision.
Vision in the periphery is much better at detecting moving objects and subtle differences in luminance. These factors can be beneficial if a rock is hurtling in your direction, say, or if you’re playing football.
So why does the fovea have a higher resolving power than our peripheral vision? Well, it’s all about the number of neurons – cells in the brain that process and transmit information through electrical and chemical signalling. There are many more neurons in the visual part of the brain dedicated to processing each visual degree of space at the centre of our vision, compared to periphery.
Our primary visual cortex – the part of the brain that receives visual information from the eyes – has a “map” of space, where adjoining neurons represent adjacent regions of space (this is known as a “retinotopic map” – see image below). This map has greater representation at the centre of our vision compared with the periphery.
When doing a difficult discrimination task in the periphery (“Are these two objects the same or different?”) we found information about these objects is also present at the foveal region of primary visual cortex.
That is, information about objects in the periphery is being sent to the part of the brain responsible for processing the centre of your vision.
This was a very surprising finding, as the retinotopic mapping of the visual cortex means this region should only receive input from the fovea or central vision.
Our finding also suggested that, in addition to direct input from the world, our foveal cortex also receives information from the periphery via other visual neurons (“feedback”).
Neuroimaging studies have the notorious disadvantage of being correlational – that is, a causal relationship cannot be established. So we needed to look to another technique to support our findings.
After lengthy discussions with a colleague at Cardiff University, Dr Chris Chambers, and thanks to the financial support of the Cardiff University International Collaboration Fund, I visited “sunny” Cardiff in 2008 to begin a three-year Transcranial Magnetic Stimulation (TMS) study.
TMS uses a very strong magnet to induce an electrical current in a small region of the cortex. When a pulse is given from the TMS, it transiently disrupts the brain under the magnet. This allows us to look at whether an area of the brain is critical in a particular task.
When we disrupted the foveal visual cortex early – 100 milliseconds (ms) – after the peripheral objects appeared on a black computer screen, people’s ability to compare them was unaffected. This is what we’d expect if the fovea responds only to incoming information from central vision.
But if we gave the TMS later, specifically 350-400ms after the stimuli appeared, then TMS affected performance. A timecourse of less than half a second might sound like a rapid pace, but for the brain, this is very slow – akin to a stroll in the park rather than a sprint.
The sprint happens for the “feedforward” visual processing (i.e. the first sweep of information that travels through the brain), which is completed by about 100ms. So this timecourse for the involvement of foveal cortex in discriminating peripheral objects pointed clearly to a feedback mechanism. That is, information about objects in the periphery being fed back to the portion of the visual cortex that processes central vision.
How might such a feedback mechanism work?
One take on it is that information from higher regions of the brain (areas involved in cognitive or semantic processes) feeds back a sort-of first guess or estimate of what might be out there in the periphery. But all previous theories of feedback suggest the information should go back to the same region of the cortex as the original input – not to the foveal cortex.
A second theory invokes the idea of a visual scratch-pad or high resolution buffer at the foveal region of visual cortex. The many neurons dedicated to processing visual information at the foveal region could provide extra power to process the information - a sort of supercomputer for vision.
The third major theory comes from the area of “saccadic updating”. Quite simply, a saccade is a movement of the eyes (or other body part) to a new location. When our eyes saccade, our visual system basically shuts down so the image we see doesn’t look blurred.
During this period (the length of which is still unknown) the brain updates the visual information to reflect the new scene that will fall on our eyes.
Our data may be evidence that even before you move your eyes, information in the periphery is transferred to the foveal region of the visual cortex in anticipation. When your eyes land on the new scene, a “guess” about what you’re going to see is already there.
It’s not yet clear which of these theories is correct. What is clear is that the foveal region of the primary visual cortex is involved in much more than simply processing what we are currently looking at. Indeed, it may support a whole other facet of our conscious vision.