Retina
Cornea
Iris Lens
Vitreous
humor
Optic
nerve
Vision basics
To understand how Campbell could
go from blindness to seeing some
light, let’s take a brief tour of the
human visual system. To see an
object, the light re;ected off of it must
pass through the cornea covering the
front of the eye, go through the lens
behind that and then pass through
the gel-like vitreous humor that ;lls
the eye. At the back of the eye, the
light hits the retina, which consists
of several layers of cells.
out that gathering visual information
from inside the eye creates several prob-
lems. “;e back of the eye is very dark,”
says Humayun, “and the light varies.”
No existing photovoltaic cells are e;-
cient enough to perform there.
Even if the photovoltaic cells could
do the job from inside the eye, it still
takes computation to turn that information into something that the optic
nerve can understand. “If you do all of
the signal processing on a chip inside
the eye, it gets pretty hot,” Humayun
says. Electronics-generated heat near the
surface of the retina could damage or
even kill some of the neurons. In thinking back about taking such an approach,
Humayun says, “It was getting pretty
complicated, so we backed out.”
Stimulating a sensation
Instead of trying to do all of the visual
processing inside of the eye, Humayun—along with University of California, San Francisco, ophthalmologist
Eugene DeJuan, M.D., and bioelectron-ics expert Wentai Liu, Ph.D., of the
University of California, Santa Cruz—
turned to an outside approach. “We
decided to use a camera outside of the
eye and have that drive most of the electronics and just put a delicate electrode
inside the eye.” But first, they needed to
see if stimulating the retina even generated any visual response in someone with
retinitis pigmentosa.
In the mid-1990s, Humayun and
his colleagues joined forces with Robert
Greenberg, M.D., Ph.D.—now president and CEO of Second Sight Medical
Products, an artificial-retina company
headquartered in Sylmar, Calif.—to test
a crucial question: If a person is blind
because of retinal damage, such as from
retinitis pigmentosa, can an electrical
stimulation to the front surface of the
retina elicit a sensation of light?
Pilot tests showed that using a small
probe to provide very low-level electrical
stimulation to the retina succeeded. ;e
patients saw flashes of light connected to
the stimulation.
But which retinal cells should be
stimulated? In retinitis pigmentosa,
the photoreceptors are damaged, so
they’re out of contention. To try to take
advantage of some of the visual process-
ing of the retina, it might make sense to
try to stimulate the bipolar cells, which
lie in the middle—front to back—of the
retina. As Humayun says, “Intuitively,
you might think it’s best to stimulate
the bipolar cells since they are the next
element in the retinal circuit after the
photoreceptors, but those cells make
some very unusual synapses when there
are no functional photoreceptors.”
In other words, in retinitis pigmen-
tosa, the lack of working photorecep-
tors—the rods and cones—causes the
next cells in the retinal circuit, the
bipolar cells, to make connections with
other cells that they would not other-
wise form. So, for some reason, the lack
of input from the photoreceptors leads
to some rewiring in the retina by the
bipolar cells. ;at would be like chang-
ing some connections in a computer
chip, thereby altering the output. In the
case of the eye, it means that the visual
processing in the bipolar cells might not
make any sense.
