May 10, 2025
Progress in Artificial Vision
Xing Chen’s work is designed to prevent merging images and has the potential to aid those with blindness.
Stimulating neurons in the visual cortex can produce perceived spots of light called phosphenes. These are drawings of phosphenes made by a blind patient. Each panel depicts a phosphene created by stimulation from one electrode. (Credit: Eduardo Fernandez, Miguel Hernández University)
Designs on Aging-Ready
By Strategic Communications
Stimulating neurons in the visual cortex can produce perceived spots of light called phosphenes. These are drawings of phosphenes made by a blind patient. Each panel depicts a phosphene created by stimulation from one electrode. (Credit: Eduardo Fernandez, Miguel Hernández University)
Maybe you’re familiar with the toy Lite-Brite. Picture a kid at the kitchen table plugging backlit yellow pegs into a board to make the outline of a sun with a smiley face. At age 4, she’s proud of her art, and her parents probably think she’s a genius. But then she haphazardly fills in most of the holes inside and around the sun with more yellow pegs. Now it looks like a bright yellow blob.
That blob could represent phosphene fusion, which happens when perceived dots of light, called phosphenes, merge in the visual field. Scientists for decades have been trying to use phosphenes to let people “see” without their eyes but have found that the discernible images of objects they can create often blend to form a single unit of light.
Xing Chen, assistant professor of ophthalmology at the University of Pittsburgh School of Medicine, wants to better understand how phosphene fusion occurs—and how to circumvent it to improve artificial vision using a brain-computer interface. A recipient of a National Institutes of Health (NIH) Director’s New Innovator Award, Chen is pursuing these questions with a $1.5 million grant from the NIH’s High-Risk, High-Reward Research Program. She hopes her research will eventually assist those with blindness resulting from damage to the eye, optical nerve or both.
Stimulating neurons in the visual cortex produces phosphenes. Chen says that, typically, placing a small number of electrodes not too close together on the surface of or in the brain can create phosphenes in recognizable shapes. Research like this in nonhuman primate studies and human clinical trials dates to the 1960s and has included both people who are sighted and those who are blind. (Her current study involves macaque monkeys that can see.)
Chen and her team are using 16 arrays of 64 electrodes each, for a total of 1,024 silicon electrodes, and implanting those into the brains of the macaques. That’s more than 10 times the number used in a recent human trial that helped a woman blind for 16 years recognize various shapes.
[Chen] hopes her research will eventually assist those with blindness resulting from damage to the eye, optical nerve or both.
Xing Chen, assistant professor of ophthalmology
The high count not only allows the monkeys to perceive phosphenes in the shapes of letters but also to determine horizontal versus vertical orientation and even sense motion. To confirm that the monkeys recognize the phosphenes, the scientists use methods such as tracking eye movements. A brain-computer interface also allows the researchers to send signals to the brain and receive signals back so they can track neuronal activity.
Eduardo Fernandez, professor of cellular biology at Miguel Hernández University in Spain, who is working with Chen as a clinical collaborator, led the clinical trial that helped the woman “see.” He says while Chen’s study is ambitious, the technology still has a way to go.
“Although 1,024 electrodes are a significant achievement with current technology, this is still a low number in comparison to the more than one million axons each human eye [uses to send information] to the brain,” Fernandez says. “This vast difference likely underlies the current limitations of cortical visual prostheses—low visual acuity, poor spatial and temporal resolution and limited control over phosphene perception.” Innovative approaches like Chen’s are needed to move the field forward, he says.
Throughout five years, Chen will implement her technique bilaterally in the brain and elicit phosphenes across a larger area of the visual field. It’s uncharted territory in this type of research.
“Creating bilateral cortical visual implants presents significant challenges,” Fernandez says. The surgical complexity is doubled, requiring precise implantation in two brain regions. And the bilateral approach requires the researchers to manage the interaction and coordination of both implants to create coherent visual percepts.
“This raises many scientific and technological questions, which is probably why we haven’t seen successful bilateral implants yet,” he says.
Chen is intent on advancing this field, and not through baby steps: “So, the main goals are to be able to investigate how phosphene fusion occurs—how many electrodes could potentially be inserted, while we stimulate in a way that prevents [the fusion] from occurring, which will give us an overview of how far we can push the technology.”