Deconstructing vision: Motion, critical windows and curing blindness in India


What if blind eyes could see? What does that mean?

That’s the question neuroscientist Pawan Sinha and his team at MIT has begun to answer in a uniquely humanitarian and scientific endeavor.

Project Prakash (named for the Sanskrit word for “light”) intended, at first, to cure blind children in India. It’s a noble effort, given that India has the world’s highest population of blind people, less than half of whom survive to their third birthday and less than one percent of whom are employable.

Sinha’s team screened 20,000 blind Indian children and treated 700 of them for correctable problems such as cataracts. As Sinha recounted at last month’s One Mind for Research forum, these 700 children now can see.

Sort of.

Their vision didn’t arrive, voilà, as with the Biblical character Bartimaeus. Instead, parts of vision have developed, gradually, and in surprising ways.

Many developmental scientists, including Nobel Laureates, have long posited that congenitally blind children older than 5 to 10 years of age will never see again, even if the actual cause of blindness is cured. That’s because sight doesn’t just involve the eyes. It also involves the brain.

If the brain loses the opportunity to take in what the eyes are seeing – and subsequently wire and fire up — the brain eventually stops trying and shuts down. The time frame in which this wiring and firing can occur is called a “critical period” of development. It’s a window of time.

Researchers had thought the critical window for sight was open only for a limited time. But investigators such as neurobiologist Takao Hensch have discovered that’s not necessarily the case. In mice, he probed the neuro-molecular changes that occur in the brain to open and close that critical window of sight. He found circuits that open the window, guided by an inhibitory neurotransmitter called GABA, a protein called Lynx 1 and special encapsulating cells called perineuronal nets that work to close it, putting the brakes on rewiring.

Mimicking or re-introducing these chemical changes might potentially serve to speed the window’s opening, delay its closing or simply shift the whole entity forward or backward in time. The results have all kinds of implications, not just for visual problems like amblyopia, but for diseases such as epilepsy and autism, where critical windows of development play out in other brain processes.

Meanwhile, the Indian children given “sight” are also offering eye-catching findings. Even though the recognized developmental window for vision has closed, some aspects of their vision still develop. Sinha shows a fascinating series of visual tests in a TED presentation, where his MIT team dissects out exactly what remains. For example, the children retain a latent ability to segment images. This means they can look at photograph and identify out of the whole picture, the shape that is the person, the one that is the tree and the one that is the car.

The skill relies on motion; the researchers move the parts of the picture around, and the children then understand what they are looking at. This discovery shows that the brain learns to see with the help of movement. Sinha calls this “dynamic information processing,” or motion processing. In essence, it scaffolds the mind for seeing. It tells us that the brain processes how things change in space over time. And that the visual system needs this dynamic information in order to begin parsing the world into distinct objects.

“This motion information serves as the bedrock for building the rest of the complexity of visual processing,” Sinha says.

Indeed, motion and visual processing also occupies the laboratory of neurobiologist Gabriel Kreiman. By observing the brains of patients undergoing surgery, Kreiman has discovered a “quick and dirty” way that the brain recognizes certain categories of objects — like vehicles, fruits or faces — while not necessarily detailing the actual object itself. It’s another facet of visual processing that is aided by motion. And Kreiman suggests that super-fast visual processing offers an evolutionary advantage. A Neanderthal’s mind might have signaled, “Predator to the left. RUN!” without having to identify exactly which kind, fur color or height.

Both Kreiman and Sinha are using their findings as recipes for constructing machine-based vision systems that can learn on their own, or see for people who cannot.  Sinha’s group is also applying its findings to autism, in which visual integration is impaired and might be a sign of a deficiency in dynamic information processing.

Beyond the lab or clinic, Sinha always has his eye on the many more blind children in India who need help. And there are many. Millions.

“It’s been just a phenomenal experience because we have gotten to do interesting research, while at the same time helping the many children that we have worked with,” he says.