Patient BAA, who is 35, lost her sight when she was 27. She can still detect light and dark, but for all intents and purposes, she is blind. Now, she—and other formerly sighted people—may one day regain a limited form of vision using electrodes implanted in the brain. In a new study, such electrodes caused parts of BAA’s and other people’s visual cortexes to light up in specific patterns, allowing them to see shapes of letters in their mind’s eyes.
The work is a step forward in a field that emerged more than 40 years ago but has made relatively little progress. The findings suggest technical ways to stimulate images in the brain “are now within reach,” says Pieter Roelfsema, a neuroscientist who directs the Netherlands Institute for Neuroscience in Amsterdam and wasn’t involved in the work.
Research to electrically spur blind people’s brains to see shapes began in the 1970s, when biomedical researcher William Dobelle, then at The University of Utah in Salt Lake City, first implanted electrodes in the brain to stimulate the visual cortex. Typically, the rods and cones in retinas translate light waves into neural impulses that travel to the brain. Specialized layers of cells there, known as the visual cortex, process that information for the rest of the brain to use.
Dobelle’s implants took advantage of a phenomenon known as retinal mapping. The visual field—the plane of space you see when you look out into the world—roughly maps onto a segment of the visual cortex. By electrically stimulating parts of this brain map, Dobelle could cause flashes of light called phosphenes to appear in the minds of people who were blind, but who had experienced at least a few years of vision. By stimulating different electrodes, he could get phosphenes to flash in different parts of a person’s visual field.
Dobelle dreamed of stimulating many phosphenes at once to create images, a bit like pixels on a computer screen. Yet this line of research stalled, says Michael Beauchamp, a neuroscientist at Baylor College of Medicine in Houston, Texas. When researchers tried to stimulate a complex image made of many phosphenes, such as a letter of the alphabet, people’s brains seemed to smudge the intended image into something unrecognizable.
He and Baylor colleagues William Bosking and Daniel Yoshor had an idea: What if, instead of stimulating all the phosphenes in a shape at once, they stimulated them in sequence, like the lights of a marquee that go off one after another? To test the idea, they found four sighted people with epilepsy who were having electrodes implanted in their brains to help control their symptoms. With the participants’ consent, the researchers added a panel of 24 tiny electrodes on the map segment of the visual cortex.
In a series of experiments, Beauchamp and colleagues stimulated these electrodes in sequences that mimicked the shapes of letters, such as C and Z. The flashes happened about 50 milliseconds apart and together, they traced the shape of a letter. The participants would then use a stylus to draw on a computer screen the images that flashed into their minds. They did phenomenally well, producing the patterns of letters with ease, Beauchamp and colleagues reported last week at the annual meeting of the Society for Neuroscience in San Diego, California, and on the preprint server bioRxiv.
Next, the researchers repeated the test with patient BAA. She had previously been outfitted with a brain implant similar to the ones the participants with epilepsy used. She used a touchscreen to draw the patterns that sprang into her mind during stimulation. Like the sighted participants, she managed to produce patterns that were close to those programmed by the researchers.
So far, the researchers have experimented only with letter patterns—between five and 10 letters per participant—but they hope to soon work with basic shapes such as squares and circles. For formerly sighted people, this work could one day lead to a device that uses a person’s cellphone camera to observe their environment and convert what it sees into basic shapes and outlines that could be beamed directly into their brains, Beauchamp says. “You could have a navigation mode, a face recognition mode, a reading mode—a variety of modes that could help blind people go about their lives.”
But the technique isn’t likely to work in people who were born blind. That’s because during early childhood development, their visual cortexes are often taken over by other functions, such as auditory processing. Roelfsema also points out that Beauchamp’s team can stimulate only about 20 phosphenes per second. That’s fine for relatively simple shapes and letters, but the kind of potential Beauchamp dreams of requires higher resolution. One possible solution is implanting the electrodes deeper within the cortex, Roelfsema says. Cortically implanted electrodes produce smaller phosphenes, he says, meaning researchers might be able to fire several electrodes at once without the phosphenes smudging together. That, combined with the new sequenced approach, could allow for more complex images.
