Analysis on Recent Artificial Retina in Electrochemical and

Mechanical Approaches

Xinrui Zhang

Faulty of Art and Science, University of Toronto, ON L5L 1C6, Toronto, Canada

Keywords: Retinal Prosthesis, Retinal Degenerative Disease, Artificial Vision.

Abstract: Outer retinal degeneration diseases, such as retinitis pigmentosa and age-related macular degeneration, are

the main eye diseases that cause blindness. Although the molecular genetics of retinitis pigmentosa has made

great progress in recent years, no obvious breakthrough has been made in the treatment. The use of retinal

pigment epithelial cell transplantation, or the transplantation of retinal slices including photoreceptors, has

major problems in case selection, immune rejection, efficacy and safety. If a certain method is adopted to

generate the perception of light and the corresponding electric current or release neurotransmitters, the inner

retina, i.e., the inner nuclear layer and ganglion cells are activated, and nerve impulses are generated and

transmitted to the visual cortex. Vision, this device that can activate the inner retina is called an artificial retina

(retinal prosthesis). Activation of the inner retina by means of retinal prosthesis (artificial retina) will be a

promising approach for treatment of outer retina degenerative diseases. Great efforts have been paid to evolve

special new devices in recent years.

1 INTRODUCTION

The design principle of the retinal implanted

electrode system is to replace the damaged

photoreceptor function, that is, to effectively capture

the visual image of the surrounding environment,

convert the visual signal into a neuroelectric signal,

and activate the inner retina to form vision (Humayun

et al. 1994). The production of vision depends on

three major tissues and organs: the eyeball (mainly

the retina), the optic nerve, and the visual cortex. So,

it is necessary for developing prostheses that can

replace these three tissues, namely retinal prostheses,

optic nerve prostheses, and visual cortex prostheses

in order to restore vision. The current approaches are

mainly facing obstacles in the artificial retina implant

location, materials selection and how to efficiently

encode and decode chemical and electrical signals,

Therefore, this article will review recent studies and

try to promote the development of artificial retina in

achieving the capacities of wide field of view, high

resolution and low aberration sensitivity.

The three types of visual prostheses have different

stimulation sites. The visual cortex prosthesis uses

microelectrodes to directly stimulate the primary

visual cortex, which can produce light perception, but

cannot form an image. Optic nerve prosthesis

stimulates nerve bundles and does not require a

complete retinal structure. However, it has low

resolution, difficult surgery, and high risks. It is still

in the basic research stage. The retinal prosthesis is to

implant microelectrodes or photoelectric arrays at

specific locations on the retina, convert additional

video and image information into electrical impulses,

stimulate specific nerve cells, then transmit them to

the visual cortex and brain center through neural

pathways (Rizzo et al. 2014).

The main function of the retinal implanted

electrode system is to replace the damaged

photoreceptor layer, accept and convert the light

signal of the environment. The key to its successful

vision depends on the biological activity of the inner

retinal neurons. Electrode pulses can affect many

inner and outer layers of the retina. At present, it is

generally believed that the ganglion cell body, axon

and proximal segment are the first targets of

extracellular stimulation (Weiland et al. 1999). Based

on the principle that the inner retinal nerve

conduction is only limited to the electrical stimulation

area, only the axons of the retinal ganglion cells in the

electrode stimulation area are activated. This lays the

theoretical foundation for retinal positioning

electrical stimulation and makes it possible to form

Zhang, X.

Analysis on Recent Artiﬁcial Retina in Electrochemical and Mechanical Approaches.

DOI: 10.5220/0011370400003438

In Proceedings of the 1st International Conference on Health Big Data and Intelligent Healthcare (ICHIH 2022) , pages 373-378

ISBN: 978-989-758-596-8

373

geometric images with simultaneous multi-point

electrical stimulation.

Researchers used bullfrog, rabbit, and mouse

animal models to study the number, size, and current

threshold of electrodes needed to produce effective

vision. It is found that the current value of the 200μm

diameter distribution with an interval of 200μm is

within the safe current value range of long-term

retinal stimulation (Weiland et al. 1999). Electric

pulses below the lower limit of nerve stimulation

cannot produce light perception. Higher than the

upper limit of nerve stimulation may cause tissue

damage. The electrical stimulation threshold will be

manually adjusted after retinal chip implantation until

the patient has light perception. After the patient is

familiar with the device, the electrical stimulation

threshold will change. The adaptation program needs

to be re-adjusted (Humayun et al. 1999). The

adaptation program converts the grayscale value of

the image to a specific value. Then it can be projected

onto the retinal electrode chip through the visual

processing unit to generate the corresponding

electrical stimulation.

Figure 1: The working principle of artificial retina

2 TYPES OF ARTIFICIAL

RETINA

There are three kinds of artificial retina devices which

are placed on the inner surface of the retina, placed on

the under the retina and chemical prosthesis. The

recent progress in three approaches to artificial retina

implementation: epiretinal prosthesis, subretinal

prosthesis and chemical prosthesis is reviewed.

2.1 Epiretinal Prosthesis

Epiretinal prosthesis is to arrange the electrodes for

electrical stimulation close to the inner limiting

membrane of the retina, without damaging the

structure of the eyeball other than the fixed point of

the electrode arrangement. The non-photosensitive

area on the device directly receives electrical signals

containing image information, and the electrodes

directly stimulate the axons of ganglion cells. If the

electronic device needs to be placed outside the

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

374

eyeball, a wire needs to be led out through the flat part

of the ciliary body. The current surface artificial

retina requires an external system for image

acquisition, image processing, data and power

conversion to the implant, so this information

conversion is more easily controlled by the outside

world. The composition of this device is a very small

field sensor like a camera, which is fixed outside the

eyeball or inside a plastic lens implanted in cataract

surgery. Moreover, a platinum-wrapped metal wire

connects it to the electrode arrangement on the top of

the inner retina. stand up. The implant on the retina is

first a readout chip, which receives the image

information from the sensor and the electrical signal

from the processing unit to generate electrical

impulses, which stimulate the axons of the ganglion

cells and pass into the brain through the axons.

A clinical trial of a retinal surface stimulator

approved by the FDA has recently begun. An

exclusive clinical trial is also underway at the Doheny

Retina Institute at the University of Southern

California, and 2 patients have received implants.

Studies have confirmed that microelectronic devices

have enough energy to directly stimulate retinal

neurons, so that patients who are completely blind can

feel the light consistent with the stimulation pulse.

Each electrode is controllable. Under the control of a

microelectronic device, when each electrode is

activated, it can cause light perception. In Germany,

the researchers temporarily placed a stimulating

device on the retina of a test subject (before eye

surgery, with normal photoreceptors, not blind). The

results showed that there was light under a very low

stimulation current. These results supports the results

of previous preclinical trials of multiple research

groups. What is more, in the case of outer retinal

degeneration, the retina requires higher current

stimulation, which is more difficult to stimulate than

the healthy retina without loss of photoreceptors

(Zhou et al. 2007).

2.2 Subretinal Prosthesis

Subretinal prosthesis is implanted between the retinal

pigment epithelium and the sensory layer of the

retina. The advantage is that the implanted electrodes

are close to the retinal bipolar cells, the stimulation

current required is small. Some devices are designed

to directly generate stimulus current from light

without an external power source; while others are

designed with an external power source to amplify the

electrical signal generated by the light. This device is

composed of thousands of microelectrodes,

containing photosensitive micro photodiodes,

integrated on a very thin board (thickness 50-100μm,

diameter 2-3mm). The photodiode is irradiated by

light and converted into a tiny current on each

microelectrode. The current is "injected" into the

remaining neurons of the retina, and the middle and

inner layers of the retina serve as the processing part

of visual information. In addition, coating

glycoproteins (such as laminin) on the surface of the

micro-photodiode can increase biocompatibility.

Besides, optobionics has been approved by the FDA

for clinical trials using artificial silicon retinas in

recent years (Hauer et al. 2007). This device is an

array of micro-photodiodes, implanted in the

subretinal space, and powered by light only. In June

2000 and July 2001, 3 people received implants on 2

occasions. It was reported at the ARVO meeting in

2002 that the implant still had electrical function and

remained in place, which could indirectly improve

visual function, but under normal light conditions, the

device itself could not directly activate the retina.

Researchers such as the University of Tubingen in

Germany have also developed a device. They

confirmed that in the long-term loss of photoreceptors

in the retina, passive light reception cannot generate

enough current to induce a direct response of retinal

neurons. In order to solve this problem, they used

external infrared energy to provide greater energy to

the implant to generate electrical stimulation pulses.

In addition to the energy supply device, the micro

photodiode located under the retina can receive light

and transmit stimulation pulses to the stimulation

electrodes under the retina. Recently, a research

group at the University of Houston used ceramic

optoelectronic materials to make implants. Japan's

Nidek Co., Ltd. produced a subretinal electrode array

with a wire connected to the electronic device in the

vitreous cavity through the retina (shown in Figure 2)

(Gu et al. 2020). Chemical prosthesis method can

release neurotransmitters in the target area of the

retina has been proposed, but its feasibility is still at

the stage of demonstration. The Kresge Institute of

Ophthalmology at Wayne State University has

proposed a microfluidic device that can be used to

stimulate the cortex and retina which can stimulate

neurons.

Analysis on Recent Artiﬁcial Retina in Electrochemical and Mechanical Approaches

375

Figure 2: Electrochemical eye with a hemispherical retina.

2.3 Chemical Prosthesis

Retinal implanted electrodes are the most successful

artificial vision implant system so far. The reasons are

as follows: (1) The complicated fatality rate of

intracranial optic cortex and optic nerve implanted

electrodes is high. With the continuous improvement

of vitreoretinal surgery technology, retinal implanted

electrode surgery. The risk of continuous reduction in

the incidence of surgical complications has gradually

been accepted by authoritative institutions. (2)

Retinal implanted electrodes avoid the complex

processing and transmission of visual signals in the

subretinal, midbrain and visual cortex. Stimulate at

the beginning of the visual pathway It is easier to

produce effective vision. (3) Due to the fusion of

multiple photoreceptor cells in the peripheral area of

the retina during the signal processing of the visual

pathway, and a bipolar cell, it is further fused and

connected to the retinal ganglion cells to transmit

visual signals. The ratio of photoreceptor cells,

bipolar cells and retinal ganglion cells is 1:1:1.

Therefore, placing a multi-electrode stimulation chip

in the macula area is more likely to produce Holmes

retinal topological vision.

3 LATEST RESEARCH AND

DISCUSSION

At present, the research of artificial retina has reached

an important stage, that is, implanting a device in the

blind to replace the lost function (Wang et al. 2020).

There are often robots with artificial eyes in science

fiction novels, and bionic eyes that are connected to

the human brain to restore the vision of the blind.

Scientists have spent a lot of energy to develop such

devices. However, making spherical human eyes-

especially hemispherical retinas-is a huge challenge,

severely limiting the functions of artificial and bionic

eyes. The team of Fan Zhiyong, the member of which

are from the Hong Kong University of Science and

Technology, the University of California at Berkeley

and Lawrence Berkeley National Laboratory reported

an innovative, concave hemispherical retina, which

consists of a series of nano-level light sensors

(photoreceptors) that mimic the human retina

Photoreceptor cells in Researchers apply this kind of

retina to electrochemical eyes, which have multiple

functions equivalent to human eyes and can complete

the basic functions of acquiring image patterns. The

retina of the human eye is hemispherical. In addition,

its optical layout is more sophisticated than the flat

image sensor in the camera: the dome shape of the

retina naturally reduces the light transmission through

the lens, thereby making the focus sharper. The core

component of the bionic electrochemical eye is an

array of high-density photosensitive elements as the

retina. Besides, the photosensitive element is formed

directly in the pores of the alumina hemispherical

film. Thin flexible wires made of liquid metal are

sealed in a soft rubber tube to transmit the signal from

the nanowire light sensor to an external circuit for

signal processing. These wires simulate the nerve

fibers that connect the human eye and brain. The most

impressive is the high-resolution imaging of this

artificial retina, which is due to the high density of the

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

376

nanowire array. In previous artificial retinas,

photoreceptors were first fabricated on a flat and rigid

substrate; later, they were either transferred to a

curved support surface or folded into a curved

substrate. This limits the density of imager units

because there must be space between them for

transport or folding. In contrast, the nanowires in this

new device are formed directly on the curved surface,

allowing them to be more closely bound together. In

fact, the density of nanowires is much higher than the

photoreceptors on the human retina. The signal from

each nanowire can be obtained separately, but the

pixels in current devices are composed of three or

four nanowires. The overall performance of artificial

eyeballs represents a leap forward for such devices.

Nevertherless, there is still a lot of work to be done.

Firstly, the photoelectric sensor array is only 10×10

pixels size large, and the gap between the pixels is

about 200-µm, which means the light detection area

is only 2 mm wide. In addition, the manufacturing

process involves some expensive and low-throughput

steps. For example, researchers used an expensive

process called focused ion beam etching to prepare

holes for the formation of each nanowire. In the

future, high-throughput manufacturing methods must

be developed to significantly reduce costs to produce

larger arrays of photosensitive elements.

Nonetheless, this work adds a strong touch to the

breakthroughs made in the past few decades. This

breakthrough was achieved by imitating camera-like

eyes and imitating insects-like eyes. Realized by

compound eyes. In view of these developments. It

seems possible for us to witness the widespread

application of artificial and bionic eyes in daily life in

the next decade (Stiles et al. 2010).

What is more, Chinese researcher Feng Miao’s

team proposed that a brain-like visual sensor based on

the vertical heterojunction of two-dimensional

materials can be built through the "atomic Lego"

method. These vertical structures can not only

naturally imitate the vertical layered structure of the

retina, but also contain the differences in the

heterojunction. Two-dimensional materials can be

used to simulate the functions of different cells in the

retina. And the number of electrodes in the artificial

retina system is an urgent hardware system difficulty

that needs to be overcome at present. Realize the

activation of independent electrodes corresponding to

independent retinal ganglion cells and improve the

visual resolution.

At the same time, researchers the United States

designed and constructed a spherical anterior retinal

chip that conforms to the macular curvature to expand

the retinal stimulation area and reduce the electrode-

retinal distance. Research on other hardware systems,

such as the development of intraocular cameras, is

used to replace external Glasses and cameras can be

set to improve the patient’s perception of spatial

positioning, and it also has huge development

potential. Studies have shown that it is a feasible new

method to elicit visual cortex action potentials

through retinal implants, but it is not certain whether

this device can achieve a certain degree of

independence for the blind. There are still many

problems to be solved. For example, it is necessary to

know whether the sense of orientation, motor

perception, and feature positioning are in the visual

cortex, how to achieve the long-term stability of the

implant, whether the retinal neurons can withstand

long-term stimulation without changes in their own

shape and function, and blind people receive

implants. The type of image that can be felt after

entering.

4 CONCLUSIONS

As the first way for humans to obtain and process

information, the human visual system has a

physiological mechanism that is significantly better

than that of optical systems. For the research of the

human visual system, people can achieve a wide

range of applications in biomedicine, machine

intelligence, and visual simulation. It is a research

topic with far-reaching prospects. The final goal of

artificial retina research is to use microelectrode

arrays to directly stimulate internal nerve cells to

replace the diseased retina, thereby restoring the

patient's vision. People need to study the biological

structure of the retina and the information processing

mechanism. Furthermore, people also need to make

microelectrodes that can be integrated with biological

nerves in terms of electronics. Although artificial

retina chips have been successfully developed, they

have been transplanted into human eyes for

experiments and have achieved certain results. There

is still a certain distance from practicality and

economy. And there is no unified model in the basic

theory.

ACKNOWLEDGMENTS

Many thanks to my professors and teachers, without

their help, this paper could not be completed.

Analysis on Recent Artiﬁcial Retina in Electrochemical and Mechanical Approaches

377

REFERENCES

Gu L L, Poddar S, Lin Y J, et al. (2020). A biomimetic eye

with a hemispherical perovskitenanowire array retina,

Nature, 581: 278.

Hauer, M. C., Tanguay, A. R., Nasiatka, P. J., & Stiles, N.

R. B. (2007) Intraocular camera for retinal prostheses.

Humayun, M., Propst, R., Juan, E. D., Mccormick, K., & D

Hickingbotham. (1994). Bipolar surface electrical

stimulation of the vertebrate retina. Archives of

Ophthalmology, 112(1), pp.110-116.

Humayun MS, de Juan E Jr, Weiland JD, Dagnelie G,

Katona S, Greenberg R, Suzuki S. (1999). Pattern

electrical stimulation of the human retina. Vision Res.

Jul;39 (15):2569-76.

Luo, H. L., & Cruz, L. D. (2016). The argus ii retinal

prosthesis system. Progress in Retinal & Eye Research,

50.

Rizzo, S., Belting, C., Cinelli, L., Allegrini, L., Genovesi-

Ebert, F., & Barca, F., et al. (2014). The argus ii retinal

prosthesis: 12-month outcomes from a single-study

center. American Journal of Ophthalmology, 157(6),

pp.1282-1290.

Sekirnjak, & C. (2006). Electrical stimulation of

mammalian retinal ganglion cells with multielectrode

arrays. Journal of Neurophysiology, 95(6), pp.3311-

3327.

Stiles, N., Mcintosh, B. P., Nasiatka, P. J., Hauer, M. C., &

Tanguay, A. (2010). An Intraocular Camera for Retinal

Prostheses: Restoring Sight to the Blind.

U.S. Department of Energy Office of Science. (2018). How

the Artificial Retina Works. Retrieved from

https://artificialretina.energy.gov/howartificialretinawo

rks.shtml

Wang, C. Y., Liang, S. J., Wang, S., Wang, P., Li, Z., &

Wang, Z., et al. (2020). Gate-tunable van der waals

heterostructure for reconfigurable neural network

vision sensor. arXiv.

Weiland, J. D., Humayun, M. S., Dagnelie, G. , Juan, E. D.,

Greenberg, R. J., & Iliff, N. T. (1999). Understanding

the origin of visual percepts elicited by electrical

stimulation of the human retina. Graefes Archive for

Clinical & Experimental Ophthalmology, 237(12),

pp.1007-1013.

Zhou, D. & Greenberg, R. (2005). Microsensors and

microbiosensors for retinal implants. Frontiers in

bioscience: a journal and virtual library. 10. 166-79.

10.2741/1518.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

378