Although a bionic eye could bring sight to the blind and greatly improve robotic vision, current visual sensors are far from nature’s impressive attributes. Scientists have now developed an artificial eye that mimics many of the functions of the eye.
An essential aspect of the eye’s design is its shape, but it is also difficult to copy. As light passes through the curved lens, the concave shape of the retina, the layer of photoreceptor-rich tissue at the back of the eye, allows the eye to pick up more light than if it were flat. However, replicating this curved sensor array has proven difficult.
A common problem with past approaches
For the most part, previous approaches first fold or transplant photosensors onto curved surfaces from flat surfaces. As a result, this approach limits bionic eye resolution, as space between sensors is required to transition from flat to curved.
According to the study published in Nature, an artificial retina with sensors built-in was created by researchers from Hong Kong University of Science and Technology. As a result, they created a device that mimicked the eye’s wide field of vision, responsiveness, and resolution.
“The structural mimicry of Gu and colleagues’ artificial eye is certainly impressive, but what makes it truly stand out from previously reported devices is that many of its sensory capabilities compare favorably with those of its natural counterpart,” wrote Hongrui Jiang, who’s an engineer at the University of Wisconsin in Madison, in a Nature perspective.
The key to its success
A dome-shaped artificial retina was implanted with photosensors in an ingenious way that led to the breakthrough. Aluminum oxide was mixed with nanoscale pores packed densely into a hemisphere. As the nanowires grew inside these pores, they used vapor deposition to create the photosensitive compound perovskite, ideal for solar cells.
A nanowire acts as the artificial equivalent of a photoreceptor. The light that passes over them transmits electrical signals that are picked up by the liquid metal wires attached to the retina at the back. Researchers created another aluminum hemisphere with a lens in the middle to represent the front of the eye and filled the space between it and the retina with an ionic liquid that mimics the fluid aqueous humor that makes up the bulk of the human eye.
Researchers then connected the bionic eye to a computer and demonstrated that it could recognize letters. Artificial eyes could not reach a field of view of 130 degrees, but they managed 100 degrees, which is higher than what a flat sensor can see.
This approach, however, might be able to improve upon biological eyes in other areas. Nanowires were found to have much higher-performing photodetectors than they were thought. In as little as 19.2 milliseconds, they were activated and recovered to the point where they could be activated again in 23.9 milliseconds. A human photoreceptor’s response and recovery time range from 40 to 150 milliseconds.
In addition, the density of nanowires in the artificial retina is more than ten times greater than the density of photoreceptors in the human eye, suggesting that the technology could achieve a resolution that far exceeds Nature.
A few setbacks
Currently, wiring these photosensors is the most significant limitation. Due to the liquid metal connections’ size could connect to many photosensors, but there is only enough room on the retina to attach 100 wires. Therefore, although the eye has dense photosensors, its retina only has 100 pixels of resolution.
Although the researchers succeeded in connecting nickel microneedles to three nanowires at a time with magnetic fields, it is a lengthy and challenging process that couldn’t handle the millions of nanowires in the artificial retina. Despite this, the discovery indeed suggests that one of Nature’s most exquisite designs may soon be duplicated and improved.
The research team concluded that it expects to see wide daily use of bionic and artificial eyes well within the upcoming decade.