chip in the retina restores sight
(appeared in April 2015)

(link to main website)

The material of glass spectacles is getting set to move into the retina, says S.Ananthanarayanan.

Silicon, whose oxide, silica is the material of the drinking glass or glass lenses, also gives off electric charges when exposed to light. This is the principle of the light meter, and also the solar panel. Implants made of silicon may now take the place of impaired light sensitive cells within human eyes.

Henri Lorach, Georges Goetz, Richard Smith, Xin Lei, Yossi Mandel, Theodore Kamins, Keith Mathieson, Philip Huie, James Harris, Alexander Sher and Daniel Palanker, a multidisciplinary team from Stanford University, University of California, Institut de la vision, Paris, Bar Ilan University in Israel and the University of Strathclyde, Glasgow, report in the journal, Nature Medicine that 1-2 mm wide, light-sensitive arrays of silicon can be placed inside the retina, to provide electrical signals and allow eyes t o see with functional clarity. The group has proved the device to work in the rat eye, and human clinical trials are planned next year in France.

The retina of the natural eye works as a panel of light sensitive nerve cells that create electrical signals that pass information to the brain. When light strikes the photoreceptor cell, there is a structural change in molecules in a pigment within the cell, and this affects the movement of charged sodium ions within the cells, creating electrical tension. This is the signal that gets transmitted down, to other nerve cells within the retina and then to the brain, as brightness in the part of the visual field that the photoreceptor cell represents. There are about 120 million photoreceptor cells in the average human eye, and these are connected to some 2 million nerve terminals leading to the brain.

Now, in older people or when there is disease, the photoreceptor cells get damaged and do not initiate an electrical signal when light falls on them. The rest of machinery is intact, but for want of the initial signal, sight is not possible and the person is blind in that eye. Age related macular degeneration (the macula is the centre of the retina) is a major cause of inability to read or recognize faces, in older people, affecting nearly 50 million. There is also a condition of retinitis pigmentosa, a hereditary disease where photoreceptor cells decline. The loss of vision progresses from the periphery to the centre and can strike at any age.

No medicines or curative procedures for these conditions are known so far, although the long term use of Saffron (Crocus sativus), a spice containing the antioxidant carotenoids crocin and crocetin has been found to give short term benefit in early ARMD. The sole recourse has hence been to retinal prosthetics. One system is cameras that generate electrical signals, which are passed by surgically implanted contacts to the nervous system. The result is not sight, but a substitute, and users are able to learn use of the neural stimuli received to approximate shapes. But bulky implanted electronics and passing of cables through the lining of the eye, with complex surgery, are involved. Another system is of placing the camera and related electronics in a chip that is implanted behind the retina. This device contains tiny photocells to capture light, amplifiers to boost their signal, and electrodes to stimulate retinal nerve cells. But the surgery required is even more complex. The cost in both cases is astronomical. And benefit is limited - users can learn just to make out the difference between a doorway and a person, or in some cases, between a fork and a spoon.

Implanting microchips

In contrast, the system reported in the paper in Nature Medicine consists of millimeter scale, silicon based chips, to be implanted with relatively less invasive surgery, into the retina, and, most significantly, does not involve any wiring from the eye to external devices, or from devices to the neural network. The chips, in fact, take the place of the degenerate photoreceptor cells and transmission of electrical signals is through the nerve cells in the retina itself, which communicates in turn with the optic nerve.

Each module of the microchip consists of an array of photocell pixels, each one 70 microns wide. Each pixel consists of two or three silicon based micro-devices that generate a voltage when illuminated, and these are arranged in series. Each pixel then, when activated, generates an electrical potential, which is passed down to the nerve cells immediately in contact with the pixel and the whole array stimulates different nerve cells as if by different natural photoreceptor cells.

One detail of the action of the photoreceptor cell is that it rapidly de-excites immediately after it is excited. If this were not the case, each bright point in the visual field would remain bright even after the bright object has moved away and movement could not be made out. The actual nature of the nerve cells in the eye is that their normal condition is to be polarized, or charged with the inflow of sodium ions. The effect light falling on the cells is that a gateway for the flow of ions get closed, and this changes the charge distribution, which gets passed on from cell to cell, as the visual signal. But the ion-gates stay closed only for an instant and the pixel becomes dark if it is not again stimulated by another particle of light. Flickering of light that is faster than the rate of relaxing of photoreceptors cannot be made out and this is the way the eyes see motion without breaks on the movie screen, for instance.

The same effect is created in the pixels of the microchip by providing a resistive shunt, or pathway which discharges the pixel a small instant after it has been charged by a particle of light. The pixel would then stop passing a signal to the nerve cells below, unless it is again stimulated by a particle of light, so that it behaves just like a natural photoreceptor cell.

The team of researchers tried out the scheme by implanting 1-2 mm chips into the retinas of experimental rats, both normally sighted as well as vision impaired. This was a simple surgical procedure, compared to implanting other devices which need to be wired and connected. With the insertion of a chip that is 1mm square, the field covered is of some 250 pixels, which can resolve a reasonable amount of detail.

That the system actually worked was verified first by sandwiching a healthy piece of retina between a photoreceptor chip and an array of electrodes wired into the terminal nerve cells in the patch of retina. Next, for testing in real life, implants were placed in the living retinas of sixteen experimental rats. The response to light stimulus was then estimated by connecting electrodes within the eye and the nerve pathway to the brain of the rats, and the response checked with stimulus by a pattern of ruled lines, and varying the thickness and spacing of the lines.

The results showed that the microchips did create sight and quality of sight is around 20/250. This a measure that means sight at 20 feet as good as a subject with normal vision would have at 250 feet. ( if our eye test gives a reading of 6/6, this means we can see at 6 metres what a normal person would see at the same 6 metres. But 6/12 would mean seeing half as well). In contrast, the elaborate prosthetics so far available are able to provide only 20/1200. The microchip method is thus about five times more effective.20/250,incidentally is like good enough to read the newspaper if the type is large

The team plans to improve the resolution by reducing pixel size and providing each one with a contact that touches fewer retinal nerve cells. Human trials are then planned in 2016 in collaboration with a French firm, using persons who have lost eyesight through retinitis pigmentosa as subjects.


Do respond to :