Vision isn’t as easy as it looks, pardon the pun. Images are coming into to the eye and going to the brain where they are interpreted by the person. Seeing the image spurs the person to move away, if it is an obstacle in their path. Conversely, the image can spur the person tor move toward it, if it is a tasty treat, like a chocolate chip muffin.
What about those who are born blind? What is going on in their brains, specifically the visual cortex? Also how is it that the brain of a sighted person is able to identify objects in both color and black and white images without difficulty and those who had vision restored after being born blind find it difficult to identify black and white images.
Two research projects are working to answer these questions and what they found shows how intricately connected the eye and brain are to each other.
Work done by neuroscientists at Georgetown University shows that the visual cortex in people that were born blind develops a unique connectivity pattern, like a fingerprint. While scientists have known for years that the visual cortex of blind person responds to stimuli such as touch, smell, sound memory recall and response to language. What scientists didn’t know is what links these tasks that activate the visual cortex.
The researchers found that there are differences in how each blind individual’s brain organizes itself and that they are stable over time. Still, this type of variation is not seen in persons who can see. These findings show how the brain develops. In particular, it suggests that experiences after birth shapes the way the visual cortex develops, even if the person is growing up blind. The brain’s flexibility, in these instances, allows it to develop different uses for the visual cortex among individuals. This knowledge may lead to improved vision rehab techniques that take into account the person’s brain connectivity pattern.
What if you were blind and you had your vision restored. There should be no problems, right? Not exactly. Children who had congenital cataracts removed end up having trouble identifying images that are black and white. Yet, a person with full vision has no problem identifying images that are shown either in color or in black in white. A study at the Massachusetts Institute of Technology (MIT) found a possible explanation how the brain can identify both color and black and white images.
Researchers used both experimental data and computational modeling to find evidence indicating that this ability to see both in color and black and white lies in development. Newborns have limited color information and this forces the brain to learn how to tell objects apart based on the intensity of light they produce, instead of their color. When the baby becomes older, the retina and visual cortex are better able to process colors and the brain uses this color information. Yet, the brain keeps the ability to recognize images without using color cues.
Researchers hypothesized that since the newborn’s visual system received less color information, this forces the brain to recognize the image’s reduced color cues. Whereas children born with cataracts who later had them removed, relied too heavily on color cues. To test this, scientists used a convolutional neural network, known as AlexNet and trained it to recognize objects, while giving different kinds of input during training. In one part of the training, they showed only grayscale images, then they showed color images. This was done to mimic the progression of color vision as a baby’s eyesight develops during the first years of life. Another training regimen consisted of only color images and this mimics the experience of children who had cataracts removed, since they can process images in full color once their cataracts are removed.
What the scientists found was that the developmental model could recognize images that were in black and white and in color. As for the model that that mimicked the children who had cataract surgery, it wasn’t able to recognize the grayscale images. By examining the organization of the models, scientists learned that those that begin with grayscale images learn to rely on the light they produce. Once they receive color input, they don’t change their approach, since it works. The models that began with color did shift their approach with grayscale images, but the shift wasn’t enough to make the images as accurate as the models that first dealt with grayscale images.
This helps to explain why children who were born with cataracts and later had them removed had difficulty identifying black and white objects. These children most likely developed an overreliance on color that made them less able to adapt to changes or removal of color information.
Once again research is expanding our understanding of the coordinated work that the eye and brain does in order to interpret images. These projects may lead to improved vision rehab techniques and possibly aid in the development of bionic vision. While vision isn’t as easy as it looks, it continues to spur researchers to learn more and this knowledge benefits everyone.