This chapter discusses retina implants currently under development that will be used to bypass degenerated photoreceptors of blind people. The chapter first provides some background information and reviews the general approaches of how to build these electronic implants. The approaches differ in the location of the stimulation electrodes. Possible locations are the epiretinal, the subretinal, the suprachoroidal space, the optic nerve and in or on the sclera. The chapter then illustrates the technical implementation of such devices in detail and shows how retina implants are clinically tested.
This chapter reports on retina implants for the blind. These electronic implants currently under development will be used to bypass degenerated photoreceptors and provide some level of functional vision for blind people with degenerative retinal diseases. Briefly, a retina implant consists of an image sensor, a processing and/or amplifying unit, and a set of electrodes.
In the following text, we provide an overview of retina implants, their components, and the different approaches currently under development. Section 15.2 focuses on background information, demonstrates the motivation behind the development of retina implants and describes the history of this development. In Section 15.3, basics about the eye and the retina are explained for a better understanding of the function of these devices. Section 15.4 illustrates the structure of retina implants. As there are several approaches to realising a retina implant, Section 15.4 also describes the most common ways of implementing such a device. In Section 15.5, technical implementations of the main system components are described. Retina implants are currently undergoing clinical testing, and Section 15.6 therefore provides an overview of the status of clinical research. After the conclusion of this chapter in Section 15.7, future trends in retina implant development are forecasted in Section 15.8, Section 15.9 provides sources for additional information.
In ophthalmology, there are a number of diseases that lead to severe visual impairment or complete blindness due to a decline of retinal photoreceptors, which are the light sensitive cells of the eye. However, despite the loss of the photoreceptors, which is the cause of blindness, in many diseases the remaining retina (neuronal tissue of the eye) remains functional. This is the underlying basis for a novel and effective treatment based on a retinal implant. The idea is to stimulate the remaining retinal nerve cells artificially and, therefore, bypass the degenerated photoreceptors. The progress made in bioengineering and micro-technology posed a challenge to researchers to develop devices to deal with the complex visual system and to partially restore visual perception. Since the first intraocular electrical retinal stimulation of blind patients with retinitis pigmentosa in 1996 (Humayun et al., 1996), numerous research groups have been developing visual prostheses. The differences between these research groups are mainly the choice of location for retinal stimulation and the progress of their development efforts. Today, this innovative technology has reached the point at which intraocular implants can be introduced to patients. Sophisticated microelectronic devices replace the missing or destroyed photoreceptor cells and electrically stimulate viable retinal tissue. This ‘artificial’ input to the retinal cells travels along the regular visual pathway to the brain where it leads to the perception of light patterns. The natural retina comprises approximately 130 million photoreceptors. As it is not possible to replace all of the photoreceptors with an artificial device, patients with a retina implant will never regain their original vision. Although the implant offers a reduced form of vision, it is a great leap forward for blind people to be able to perceive light and contours again, and to be able to react directly to their environment and movements.
This section provides an overview of the principle structures of the eye and the retina. There is a description of how the normal sighted see and an overview of the causes of blindness. Last but not least, we demonstrate how a blind patient can benefit from a retina implant.
The eye can be compared to a photographic camera that consists of an optical system and a recording medium. The transparent cornea and lens serve, together with the pupil, as the optical system. When light enters the eye, these parts focus the beam onto the retina. The retina represents the recording system. It consists of eight layers and is a cellophane-like, transparent membrane that covers the inside of the back of the eye and contains vessels (see Fig. 15.1). The supply of oxygen and nutrition to the retina is provided by two systems: the outside half of the retina is supplied by diffusion from the underlying tissue, the retinal pigment epithelium, and the choroid, tissue that consists mainly of vasculature. The inner half of the retina is supplied by vessels that can be seen on the retina. Between the lens and the retina is the vitreous, a transparent gel-like substance that constitutes about 80% of the volume of the eye.
One of the outermost layers of the retina is the photoreceptor layer that changes light into bioelectric signals. The human eye has about 130 million receptors. The signals are then relayed through the other layers of the retina and sent to the brain via the optic nerve at the back of the eye. During the transmission of signals through the other layers of the retina, the information, which is encoded in the signal, is transformed and converges on the 1.5 million ganglion cells. This step involves complex information processing, which is an active research topic. The ganglion cells have long extensions, the so-called nerve fibres, which cover the inside of the back of the eye, exit the eye at the optic disk, and form the optic nerve. The optic nerve represents the connection from the eye to the brain. The retina, the optic nerve and specific parts of the brain form the so-called visual pathway. The brain itself processes the information by interpreting it. These interpretations are then compared with the experience of the individual and the most verisimilar one is finally selected.
Diseases that lead to partial or complete blindness are malfunctions along the visual pathways. The most common cause of blindness in the western world was previously glaucoma, which, if not treated, is characterised by gradually increasing damage of the optic nerve. During the past decade, however, the most common causes of blindness are retinal diseases, such as macula and other retinal degenerations (WHO, 2011). These diseases are a compilation of various entities, which have in common that the photoreceptors decline. Macular degeneration, for example, affects the centre of the retina, the fovea, which transports the largest portion of visual information allowing us to see fine details or to read. Other diseases, most of which are inherited, can affect the whole retina. In addition, disorders that affect the brain along the visual pathway, such as a stroke, can also lead to blindness.
It becomes obvious that retinal implant devices can only be of use in retinal disorders. This is because they serve as a replacement for photoreceptors and there is a mandatory need for a functioning visual pathway to process the information. Retinitis pigmentosa (RP) is one of the diseases that especially lend itself to this technological application. RP is a family of inherited diseases that causes a decline in retinal photoreceptors and leads to incurable blindness. A progressive loss of visual fields is the hallmark of RP, which often begins as a donut like ring in the mid-periphery. As RP progresses both centrally and peripherally, the resultant tunnel vision begins to affect the patient’s activities, driving and mobility. Data about the number of people afflicted with RP vary greatly. Worldwide, around three million people suffer from RP, which equates to a prevalence of 1 in 3000–70 000 depending on the various study populations (Daiger et al., 2007). RP is most commonly found in isolation, but it can also be associated with systemic disease. The most common systemic form is called Usher syndrome and is associated with hearing loss (up to 30% of patients). RP is usually diagnosed in young adulthood, although it can present anywhere from infancy to the mid-30s to mid-50s. In a third of patients, the tunnel continues to narrow and causes complete or virtually complete blindness. Loss of vision has a major impact on the lives of individuals suffering from RP.
In patients with photoreceptor loss, a retinal implant stimulates the retina with electrical pulses. Each single electrical pulse is perceived as a light spot, a so-called phosphene. Current technology could allow an improvement from complete blindness to a level of visual impairment that allows rough orientation of the surrounding area (see Fig. 15.2). The picture given by retinal devices is at first unfamiliar to patients. Patients learn to adapt to the artificial stimulus provided by a training and rehabilitation programme. Figure 15.2 shows a simulation of what an implant user could see.
Development in this field has progressed, and the first experiences of patients are now available. Future developments will concentrate on bringing this technology to blind patients with other diseases.
This section provides an overview of the structure of retina implants, and several approaches regarding how to place the electrodes are introduced. Each approach is discussed in detail and advantages and disadvantages are examined.
The aim of retina implants is to bypass the degenerated photoreceptors by artificial stimulation of the remaining nerve cells. A retina implant system typically consists of an image sensor, a processing or amplifying unit, stimulation electronics (a set of current or voltage sources), and an array of electrodes (see Fig. 15.3). In addition, the implant needs to have a power supply unit. In those cases where the image sensor is outside the body, the devices also contain a wireless data transmission unit.
The image sensor captures images from the environment and converts them into electrical signals. These signals are then processed by a signal processor. Processing can be standard image processing, such as contrast enhancement, or special algorithms to adapt to the biological visual system. Some implants do not have a processing unit. In these devices, the light signal from the sensor is simply amplified and transferred to the corresponding stimulation electronics where current or voltage pulses are generated. These stimulation signals are applied by the electrodes to the retinal nerve cells. The electrodes are arranged in an array that covers a certain part of the retina.
Retinal nerve cells always have their receptive fields close to the cell bodies. This means that if these cells are stimulated electrically, the patient sees a phosphene whose location corresponds to the location of the electrode. The stimulation of neighbouring cells or cell groups results in neighbouring phosphenes. This very important detail is called a retinotopic arrangement of nerve cells. A retina implant without processing would simply activate an electrode if the image sensor detected light in its corresponding location of the visual field. Images are created by activating several electrodes. More sophisticated retina implants do not transfer the light information from the image sensor directly to the electrode array but process the data to optimise perception and ease pattern recognition.
The electrodes apply electric energy to the retina and are therefore the neural interface of a retina implant. There are several ways in which the electrodes can be positioned. The most straightforward method is to directly contact the retina from the epiretinal side. With this approach, the electrode array is positioned in the inner eye. The advantage of the epiretinal approach is that the major part of the necessary surgery is more or less standard. If the electrode array is placed on the other side of the retina, between the retina and the pigment epithelium, the device is called a subretinal implant. The subretinal space is surgically more difficult to access than the epiretinal space. To avoid any retinal surgery at all, and to avoid any risks associated with contacting the retina with a device, suprachoroidal implants have been developed. With this approach, the electrodes are placed more distantly, which reduces the chance of selectively stimulating neighbouring cell groups in the retina, but minimises surgical risks.
In the 1970s, eye surgeons developed techniques to remove the vitreous, which allowed them access to surgically treat the retina (Machemer et al, 1971; Machemer, 1978). The vitreous is removed using a very small cutting device through a small incision in the sclera close to the cornea. This procedure is called vitrectomy. The removal of the vitreous allows surgical manoeuvres on the retina. Today, vitrectomies are well established and are used worldwide in eye hospitals. A vitrectomy allows the placement of an electrode array on the epiretinal side of the retina where the electrodes are adjacent to the ganglion cell layer (see Fig. 15.4). This is the most obvious position for a retinal implant, as the photoreceptors and parts of the other network of the retina are degenerated and can be completely bypassed by stimulating the ganglion cells. The ganglion cell layer is widely intact, even after a long period of blindness (Stone et al, 1992). However, as the current does not stop behind the ganglion cell layer, other cells such as amacrine cells, bipolar cells and horizontal cells are also stimulated, albeit at a lower current density.
A thin axon layer is present between the electrodes and ganglion cells. The stimulation current of epiretinal implants needs to pass the axon layer without stimulating the axons. This is very important, as the stimulation of the ganglion cell axons can lead to large phosphenes whose location does not correspond directly to the location of the electrodes. In clinical trials, the perceptions of RP patients after electrical stimulation of the retina were found to be relatively small and localised in a certain area (Hornig et al., 2007; Richard et al, 2005). It was found that the perceptions corresponded well to the electrode position. This retinotopic arrangement indicates that the cell bodies, or at least the axons near the cell bodies, are stimulated rather than the axon layer. Fried et al. (2009, 2010) reported that the stimulation target is dependent on the stimulation frequency. With higher stimulation frequencies, ganglion cells and axons could be stimulated. Low frequencies result in stimulation of more distal retinal nerve cells. This would allow epiretinal implants to steer the target of the electrical stimulation.
Typically, the electrode arrays of epiretinal devices are held in place with a retinal tack (Taneri et al., 1999; Hornig et al., 2007). Such tacks were originally developed to treat retinal detachments (Puustjarvi and Terasvirta, 2001). They are pushed through the retina and anchored in the sclera. Most epiretinal electrode arrays are fixed in such a way that the tack is pushed through a hole in the device and then anchored in the sclera. The IMI Intelligent Medical Implants group developed a tack fixation where the tack is first set, the implant is positioned over the tack, and a retainer is fixed onto it (Tiedtke and Meyer, 2007; Ivastinovic et al., 2009).This technique has the advantage that the implant can be removed if necessary without removing the tack, as tack removal has the risk of retinal detachments. Another technique for attaching an epiretinal device, which is currently under development, is to glue the device onto the retina (Tunc et al., 2008). The biggest challenge with this technique is the need of a close connection between the electrode array and the retina combined with the possibility of later removing the device without damaging the retina. Glues have therefore been tested that change their adhesion properties dependent on temperature. These glues do not adhere below the transition point of 31 °C but do show adhesion at temperatures above the transition point (Tunc et al., 2008). To remove the device the glued area just needs to be cooled below 31 °C.
The circuitry to generate the stimulation current in epiretinal implants is typically not placed directly on the retina. Some groups decided to place this part of the implant outside of the eye, fixed at the sclera (Kelly et al., 2009; Richard et al., 2009) (see Fig. 15.5a). In this case, a small cable passes through the sclera. This is preferably done close to the anterior part of the eye where no retina exists. The attachment of electronics outside of the eye makes surgery for implantation and explanation much easier. The German EPI RET (epiretinal) group (Walter and Mokwa, 2005) chose to place this part in the lens capsule (Alteheld et al., 2007) (see Fig. 15.5b).
Epiretinal implants are a very straightforward way of placing an electrode array close to the retina because ganglion cell stimulation bypasses all the degenerated cells of the retina. The surgery, including the fixation of the array, is now established and the first clinical results confirm that the devices function correctly.
Another location to place an electrode array close to retinal nerve cells is the subretinal space (Zrenner et al., 1997; Chow et al., 1999; Schubert et al., 1999; Shire et al., 2004) (see Fig. 15.6). To do this, the retina must be detached from the pigment epithelium, the electrodes put in place and then the retina reattached (Sachs et al., 2005). Typically, this is done using an ab externo approach. This means that first the vitreous is removed, then the sclera is opened from the outside and the retina is detached by injecting fluid in the subretinal space. The electrode array is placed and the retina is reattached by removing the liquid from the subretinal space.
Subretinal implants aim to stimulate the surviving retinal cells from underneath the retina. As the photoreceptors are degenerated, the subretinal electrode array stimulates bipolar, amacrine and ganglion cells. Subretinal electrode arrays are typically not specially fixed to the retina. They are held in place by the adhesion between the retina and the pigment epithelia or by the force of the power supply cable (Sachs et al., 2005).
One potential disadvantage of the subretinal approach is that the supply to the retina of oxygen and nutrients through the pigment epithelium can be interrupted by the electrode array. One solution to this problem is to make holes in the substrate of the electrode array so that nutrients and oxygen can diffuse through the holes. Another solution is to make the electrode array relatively small to allow diffusion around the device. However, only long-term clinical trials with functional electrical stimulation will show if the supply from the subretinal side is indeed needed, or if the degenerated retina is sufficiently supplied from the epiretinal side. A potential advantage of subretinal electrode arrays might be that the electrodes are very close to the retinal nerve cells, even if the electrode array is not curved.
Subretinal implants are more difficult to implant but initial clinical experience has shown that surgery is feasible (Zrenner et al., 2008,2010). Longterm clinical trials will show if interruption of the supply of oxygen and nutrients from the subretinal side will have a negative effect.
To minimise surgical efforts and risks, retina implants where the electrodes are not directly placed on the retina are under development. Locations are between the choroid and sclera, inside the sclera or attached to the sclera from outside the eye (Inomata et al., 2008; Fujikado et al., 2011; Liang et al., 2011; Morimoto et al., 2011). According to the position of the electrode array, these devices are called suprachoroidal or transscleral retinal prostheses.
For suprachoroidal positioning of an electrode array, the choroid is exposed by opening the sclera from outside of the eye. The electrode array is then placed onto the choroid and the sclera is closed again (see Fig. 15.7).
Transscleral implants stimulate the retina through the sclera (Inomata et al., 2008). This is a very safe approach because the eye does not need to be opened, which avoids intraocular complications. The electrode array can be placed outside the eye or, alternatively, a sclera pocket is prepared and the electrode array is placed in the sclera.
One potential disadvantage of suprachoroidal and transscleral devices is the larger distance between the electrode array and the target cells in the retina. The choroid has a thickness of 0.1–0.3 mm. Current also needs to pass the pigment epithelium, which is only one cell layer thick but has a relatively high specific resistivity of 100 000 Ωcm (Zhou and Greenberg, 2009). In transscleral stimulation, the current also needs to pass the up to 1 mm thick sclera (completely or partly). These circumstances lead not only to high stimulation thresholds but also to a blurring effect of the stimulation focus. This needs to be compensated for by a larger pitch of the electrodes. However, with a larger pitch, the whole electrode array becomes larger. For a large number of electrodes, a flat electrode array cannot fit the curvature of the eye anymore. Ohta et al., (2007) solved this problem by splitting the electrode array into strips, which are in relatively close contact with the retina.
Suprachoroidal and transscleral implants aim to have fewer surgical risks because the eyeball is not opened. However, because of the larger distance from the electrodes to the retinal nerve cells, activation thresholds will be higher and resolution could be worse than in epiretinal or subretinal implants.
In the retina, the optical signal is converted into a bioelectrical signal that is then transported via the optic nerve to the brain. Another possible way of bypassing degenerated photoreceptors is to directly stimulate the optic nerve. This approach avoids retina surgery completely. It can particularly be used if, due to additional eye diseases, the view inside the eye is poor and retina surgery is not possible. The optic nerve can be stimulated with penetrating or non-penetrating electrodes (Delbeke et al., 2001; Chai et al, 2008; Wu et al., 2010). Non-penetrating electrodes are arranged as cuff electrodes that are fixed around the nerve (Delbeke et al., 2001) (see Fig. 15.8). These electrode arrays have contacts that are distributed around the optic nerve in a circular fashion. By single or synchronous activation of multiple contacts, different parts of the optic nerve can be stimulated (Brelen et al., 2005). However, the selectivity of the stimulation is relatively low as it is difficult to steer the electric field within the optic nerve. To increase the selectivity of optic nerve implants, penetrating electrodes are used (Chai et al., 2008; Wu et al., 2010). Clinical results are, however, not yet available for penetrating optic nerve implants.
Retina implants consist of several components. This section explains the main elements of retina implant systems and their features. All current retina implant systems consist of external and implanted components. External components are all those parts of the system that are not implanted. Typically, the image capturing and data processing is done by the external components. In this case, data needs to be transmitted wirelessly to the implanted device. Systems that have no or only limited data processing often have the photo sensor array implanted. However, as a minimum, the power to operate the system is transmitted wirelessly to the implant. The implanted device can be divided into two main sections: stimulation electronics and electrode arrays. The stimulation electronics are used to receive data and power as well as to transform the signal into electrical stimulation pulses. The electrode array applies the current to the human body. Below we provide an explanation of the different possible technical implementations for all these components, and an examination of the advantages or disadvantages for certain variants. The order of the description is chosen according to the method of signal processing from the optical image to the electrical stimulation of the retina.
For most designs, retina implant systems consist of one or more external (i.e. non-implanted) components, in addition to the implanted components. Typically, a pair of glasses and an external, small processor are used (Hornig et al., 2007; Zhou and Greenberg, 2009) (see Fig. 15.9). A camera mounted on the pair of glasses captures images from the environment and converts them into digital signals. These signals are transferred via cable to the external processor. The processor converts the digital signals into stimulation commands, encodes them and transfers them back to the glasses. The encoded stimulation commands are then sent wirelessly to the implanted device where the commands are decoded and converted into electrical stimulation pulses. Power to the implanted device is provided wirelessly via RF signals that are generated by a transmitter integrated into the glasses (Hornig et al., 2007).
Installing the camera in the glasses is one of the easier tasks in developing a retinal prosthesis. Modern handheld applications possess micro-miniaturised cameras with much better performance than can be utilised in any of the existing retina implant projects. Typically, CCD photo sensors or CMOS cameras are used (Schwarz et al., 1999, 2000).
The external processor consists of a high-performance signal processor board and a set of batteries that provides the whole system with energy. The signal processing is described in the next section. External components possess the important advantage of being easy to access, making adaptation and updating during clinical trials possible.
The processing of the images is done by a portable processor. On one hand, it is based on standard image processing algorithms such as contrast enhancement or edge detection and, on the other hand, special encoding algorithms are used to convert the image signal into stimulation commands.
Standard image processing is often used to reduce the huge amount of information from the camera to a level that can be transferred by the electrode array to the visual system. This can be very important because an overload of information might confuse the implant carrier. It can also be used to emphasise important objects (Al-Atabany et al., 2009; Fink and Tarbell, 2009; Parikh et al., 2009).
Retina implants stimulate the remaining nerve cells of the retinal network. In a healthy person, these target cells would have already coded the visual information. It could therefore be useful if, in artificial stimulation of the retina, the image information is correctly encoded before it is fed into the human visual system. Today, signal processing of the retina is well understood (Rodieck, 1965). It has been discovered that signals are processed through ON and OFF pathways (Hubel and Wiesel, 1960; EnrothCugell and Robson, 1966), colours are encoded and time filtering is done (Lee et al., 1994). The cells also differ in the size of their receptive fields (Rodieck and Stone, 1965). Whenever these retinal nerve cells are activated artificially, it makes sense to activate them in the same manner that they were activated when the retina was still completely functional. An encoder for retinal prostheses was therefore developed which mimics the signal processing from the photoreceptor level to the ganglion cell level (Eckmiller et al., 1999, 2005; Hornig et al., 2007). Given the history of cochlear implants, pre-processing seems to be a very important feature in neural prostheses and could improve the performance of a device significantly (Clark, 2003).
The active implant needs to be provided with power to generate the stimulation current. As implantable batteries would be too heavy and too large for a retina implant, and a cable through the skin carries the risk of infections, wireless power transmission schemes were developed. Most of the current implant systems use electromagnetic energy transmission, where a primary coil is placed outside of the body (e.g. on the glasses), and a secondary coil is placed in the implantable device (see Fig. 15.10) (Hornig et al., 2007). This method is also known as inductive coupling. The primary coil generates an alternating magnetic field. This field induces a voltage in the secondary coil. In retina implants, transmission frequencies between 125 kHz and 49 MHz are used (Caspi et al., 2009; Kelly et al., 2009; Shire et al., 2009).
In compact implants, such as the IMI Intelligent Medical Implants device, the coil is positioned very close to the eye and is sutured on the sclera (Hornig et al., 2007). Other implants place the coil behind the ear, such as in a cochlear implant, and connect coil and implant with a cable (Clements et al., 1999; Rothermel et al., 2008). In this case, primary and secondary coils are in very close proximity; they are only separated by the skin of the patient. Energy transmission is therefore very effective because a very large part of the magnetic field of the primary coil reaches the secondary coil. A disadvantage is that a cable from the eye to the location beneath the skin is needed. These cables make surgery more difficult and introduce the risk of breakage during eye movement.
An alternative method of providing retinal implants with energy is by light. Here, the implant contains a set of solar cells that generate the necessary electrical power (Zrenner et al., 1997; Chow et al., 1999; Laube et al., 2004; Palanker et al., 2005). Some early implants tried to use ambient light passing into the pupil to operate the implant electronics (Zrenner et al., 1997; Chow et al., 1999). However, even when very efficient solar cells were used, the energy was not enough to supply the implants properly. To solve this problem, high-powered light was generated and transmitted through the pupil (Laube et al., 2004; Palanker et al., 2005).
The transmission of stimulation data from the external components to the implant can be carried out in a similar way to the transmission of power, but less energy is needed. If an electromagnetic field is used for data transmission, the same carrier frequency as for the power transmission can be used (Caspi et al., 2009), or a different frequency (Kelly et al., 2009; Shire et al., 2009).
Data transmission can also be done optically (Hornig et al., 2007), where an optical transmitter is placed in front of the eye. The light is transmitted through the pupil and is received by a photo sensor of the implant. Optical data transmission has the advantage that data transmission rates can be very high and the receiver circuitry can be designed to be relatively simple because no demodulation circuitry is needed.
Which of the various options of power and data transmission is selected depends on the requirements of the individual device. This is often also a compromise between transmission efficiency, which requires close proximity between transmitter and receiver, and ease of surgery, which requires a small implant.
The implantable device of a retina implant system needs to contain a powerful electronic circuit to receive the data and energy from the external components and to generate stimulation signals that are applied to the retina via the electrodes. As this circuit needs to be extremely small, application-specific integrated circuits (ASICs) are used. A possible architecture for such an ASIC is shown in Fig. 15.11.
Stimulation electronics with various numbers of stimulation channels, level of output current and level of output voltages exists. Chen (2010), for example, developed a 256-channel stimulator chip that is able to deliver a total maximum current of 3 mA and an individual maximum current of 500 μA per electrode. The chip has a voltage span of approximately ± 10 V to provide this stimulation current. The stimulation parameters for pulse current amplitude and pulse duration are received wirelessly.
To avoid charge accumulation and irreversible faradaic effects on the electrodes, it is necessary to deliver charge-balanced pulses. Typically, biphasic, rectangular current pulses are used. However, a charge can accumulate due to leakage currents and a mismatch of cathodic and anodic pulses. This is traditionally solved by placing blocking capacitors in a series of each of the stimulation electrodes. However, in retina implants with a large number of electrodes this is not possible due to the limited space available. This problem is solved by using active charge cancellation circuits. Chen et al., (2010) used a switch to shorten the stimulation electrodes with the return electrode. Sooksood et al., (2009) developed a circuitry where the voltage on an electrode is measured and, if necessary, an offset current is applied.
The implant electronics need to be encapsulated to protect them against body fluids and guarantee functionality for a sufficient lifetime of the device. For encapsulation, different materials such as polymers, glass, ceramics and metals can be used (Jiang and Zhou, 2010). However, it is a challenge to create the feedthroughs, which transfer the current out of the encapsulation to the electrodes, for a high number of electrodes.
Retina implants stimulate the remaining retinal nerve cells of blind patients electrically. Electric charges need to be applied at different locations on the retina to create an image perception. Electrode arrays are therefore used that consist of a number of electrodes, which are distributed over a certain area of the retina.
The types of electrode arrays can be divided into two groups: flexible and solid. Flexible electrode arrays consist of thin polymer films where a number of electrodes are embedded (Stieglitz et al., 1997; Stieglitz and Gross, 2002; Guven et al., 2005; Kim et al., 2009; Weiland et al., 2009). The electrodes are then connected via circuit paths in the polymer film to the stimulation electronics. In solid electrode arrays, the electrodes are directly connected to the stimulation electronics, for example by placing electrodes onto the silicon chip (Chow et al., 1999; Rothermel et al., 2008).
Electrode arrays that are larger than approximately 4 mm should be either flexible or pre-shaped to match the curvature of the eyeball. If this is not the case, the distance between the electrodes and the retina can be too large. To match the curvature of the eye, flat polymer electrode arrays can be formed thermally, for example (Weiland et al., 2009).
Flexible electrode arrays based on thin polymer films can be produced using photolithography and typically consist of a sandwich of several layers of polymer and metal (Kim et al., 2009; Weiland et al., 2009; Stieglitz and Gross, 2002). The metal layers are used as electrodes and tracks to lead the current from the current generator to the electrodes (Weiland et al., 2005). Possible polymer materials are polyimide, parylene and silicone (Weiland et al., 2009). As the processing of polyimide is well known from the industrial production of flexible circuit boards, and its biocompatibility has been shown in many studies (Richardson et al., 1993; Seo et al., 2004; Sun et al., 2009; Myllymaa et al., 2010), it is the most widely-used material for thin electrode arrays in retina implants (Walter and Mokwa, 2005; Hornig et al., 2007; Kim et al., 2009; Weiland et al., 2009). It is a very robust material and shows sufficient mechanical stability for surgery (Richard et al., 2007).
Parylene is very flexible and has excellent long-term biocompatibility (Yu et al., 2009). In terms of stiffness, it is a good alternative to polyimide (Weiland et al., 2009). Another alternative substrate material is silicone. Often used in medical devices, a very flexible silicone is poly-(dimethyl siloxane) (PDMS) (Weiland et al., 2004).
Using polymer films as electrode arrays has the disadvantage that all the tracks need to be routed from the distant stimulation electronics to the electrodes. This can be a significant problem when several hundreds of electrodes are used. A method that avoids this routing problem is to combine the electrode array and the current sources, as is typically done in solid electrode arrays. In this case, the stimulation circuitry is placed directly onto the corresponding electrode. With this approach, it is possible to implement more than 1000 electrodes on one electrode array without any track routing issues (Rothermel et al., 2008). Some devices have also implemented photo sensors close to the electrodes that are used as a camera (Chow et al., 1999; Rothermel et al., 2008).
For retina implants, it is important to place the electrodes as close as possible to the retinal nerve cells. This leads, on one hand, to a lower activation threshold of the cells, which is beneficial because less energy is needed and the electrodes can be designed to be smaller. On the other hand, with close proximity to the nerve cells there are better chances of stimulating only a small group of cells, which can lead to a better image resolution. Implants that combine stimulation electronics and electrodes are disadvantaged by the fact that the devices are no longer mechanically flexible. Because of the curvature of the eye, these devices need to be of a limited size in order to avoid too large a gap between the electrodes and the retina. To achieve better proximity to the target nerve cells and, therefore, reach a lower activation threshold and more selective stimulation, Palanker et al., (2004a, 2004b, 2005) developed a three-dimensional electrode array with the intention of retina cells migrating into the array. Such a configuration leads to the lowest activation thresholds of nerve cells. It does, however, make explantation without damaging the retina nearly impossible.
The stimulation current is delivered to the retina by electrodes. Basically, the electrodes convert the electron current in the metal to an ion current in the body. As perception thresholds in clinical trials were reported often to be in the range of 100 nC (Mahadevappa et al., 2005; Richard et al., 2005) for an electrode diameter of 200 μm, a charge density of 0.32 mC/cm2 needs to be applied. Compared with other medical applications this is quite a high charge density (Clark, 2003). To avoid toxic electrochemical reactions at the electrode, this high charge density requires the use of advanced electrode materials. Possible materials are platinum, iridium oxide, titanium nitride and PEDOT (Polyethylenedioxythiophene). Platinum is very often used in neurostimulation devices, and long-term experience reports regarding stability are available (Stieglitz, 2004; Cogan, 2008). Iridium oxide, titanium nitride and PEDOT are new materials that show promising charge capacities (Cogan, 2008), but long-term use with high charge densities must still be demonstrated.
The design of an electrode array requires one of the greatest efforts in the development of a retina implant. In particular, it is essential to obtain a very close contact without damaging the sensitive retina. This has a significant influence on the performance of the devices. In addition, it affects the design of many other components because it changes the required electrical power and the number of reasonable stimulation channels.
Some devices use an implanted camera chip that is combined with the electrode array instead of an externally mounted camera (Zrenner et al., 1997). The advantage of this approach is that eye movements can be used similarly to natural vision. Hence, very quick scans and even micro saccades (very small, involuntary eye movements) can be accomplished. The disadvantage is that the visual field of the device is always limited according to the size of the implanted device. For mobility purposes, it would be beneficial to have a large visual field. However, for surgical and biocompatibility reasons, the implants should always be small and cannot therefore allow large visual fields. A typical size for an implanted camera device with an electrode array is 3 × 3 mm2, resulting in a visual field of only 11° × 11° (Zrenner et al., 2010). In many countries, people with a visual field of less than 20° are classified as legally blind.
Since the first electrical retinal stimulation of blind patients with retinitis pigmentosa was conducted in 1996 (Humayun et al., 1996), numerous research groups have been seeking to establish a visual prosthesis for blind people.
Dr Alan Chow was the first person to implant a retina device, which he called an Artificial Silicon Retina (ASR) and which was sponsored by Optobionics (Glen Ellyn, Illinois, USA) (Chow et al., 1999, 2004). He used the subretinal approach for a 2-year study on the safety of the device. The implant had a diameter of 2 mm, the thickness of a human hair, and consisted of approximately 5000 microscopic solar cells called ‘microphotodiodes,’ each with its own stimulating electrode. A visual acuity enhancement was not detected, but a rescue effect on the declining function of the photoreceptors in areas far from the implant site was suggested, meaning that the other parts of the retina might regain some more function. The missing effect of the implant was mainly due to the microphotodiodes not generating enough energy to stimulate the retina nerve cells.
Dr Mark Humayun and his group announced in 2002 the successful implantation of an epiretinal implant sponsored by Second Sight Medical Products (Valencia, CA, USA) (Humayun et al., 2003). This implant had 16 electrodes and served as a proof of concept. The six patients that were implanted were able to detect light and distinguish the direction of movements. Stimulation thresholds were found to vary between patients and between electrodes within one patient (Mahadevapper et al., 2005). This is an indicator that individual tuning is essential for proper functioning of the device.
To establish the appropriate size of the electrode and the necessary energy to perceive phosphenes in a larger group of RP patients, a European multicentre trial was conducted from 2002 to 2003 by IMI Intelligent Medical Implants (Bonn, Germany) (Hornig et al., 2005). A vitrectomy was performed on twenty RP volunteers and the retina was electrically stimulated intra-operatively. Threshold charges from 20 to 768 nC with single or multiple electrodes were identified. The optimal size of electrodes was determined to be 200–300 μm. In 2005, a wireless epiretinal implant of 49 active electrodes was implanted in four patients by IMI Intelligent Medical Implants (Richard et al., 2007). In this study, the stimulation signals were generated by a computer that was connected to the external components. The implants proved to have good biocompatibility. As all of the patients had also participated in the multicentre trial before, a direct comparison of stimulation thresholds between an acute and a chronic device was possible. It was found that stimulation thresholds with the chronically implanted device were lower than in the previous acute trial. Identification and location of single and several light spots as well as simple pattern recognition were verified and stimulation strategies evaluated (Richard et al., 2008).
In the same year, Dr Eberhart Zrenner and his group started a study (Retina Implant AG, 2007) on the safety and efficacy of subretinal implants conducted by Retina Implant AG (Reutlingen, Germany). He used microphotodiodes as Chow did before, but supplied the device with electric power by implanting retro-auricular transdermal cable endings (a cable that exits the skin behind the ear). Eleven patients were enrolled, and the study was completed in 2009 (Zrenner et al., 2010). Three patients could locate bright objects on a dark table. One patient was even able to recognise letters 11 cm in size. With this trial, the proof of concept was established for subretinal implants.
At the beginning of 2007, IMI Intelligent Medical Implants started a clinical trial on epiretinal implants with 49 electrodes. The external components consist of a tiny camera and a transmitter mounted in glasses. This multicentre study concentrated on pattern recognition, the safety verification of stimulation parameters and the performance of the external components (IMI Intelligent Medical Implants, 2007). The receiving and stimulating components were entirely implanted into the eye. In 2006, Second Sight Medical Products began a multi-centre study with the Argus II Retinal Stimulation System (Second Sight Medical Products, 2006). This study concentrated on visual acuity and safety. The receiver was implanted around the eye and connected to the epiretinal electrode array in the eye. The first results of this study have been made public. Most of the patients had improved visual information, and 40% were able to read large letters (Humayun et al., 2010).
The latest study was announced in November 2009 by Retina Implant AG, Germany, concerning the Retina Implant model Alpha, which uses the subretinal approach (Retina Implant AG, 2009). It is a multicentre trial on safety and efficacy, using daily life activities as the primary measuring tool. These studies have already proven that retina implants can partially restore visual perception in blind RP patients. It is not clear which technology, concept or combinations of concepts, will provide the optimum patient performance and reliability. It is clear, however, that tuning of the implant already appears to be one of the key factors in improving results.
Blindness is one of the most serious disabilities in our community. For many blind people, such as patients with retinitis pigmentosa, no treatment is currently available. Retina implants can aid this group of patients through electrical stimulation of their remaining retinal nerve cells. The degenerated photoreceptors are then completely bypassed. To stimulate the remaining retinal nerve cells, electrodes need to be placed in close proximity to these cells and, for this, four different locations exist: epiretinal, subretinal, suprachoroidal/transscleral and the optic nerve. All of these electrode positions have advantages and disadvantages and, up to now, it is not clear which one should be preferred.
The technological requirements of retina implants are high. This is especially true regarding the size of the devices and the charge transfer of the electrodes. The development of electrode arrays with a high number of electrodes and with additional flexible behaviour is new. In clinical trials, it has been shown that blind patients have some level of visual function when their retina is stimulated with a retina implant. It is expected that vision can be significantly improved by using image processing algorithms and by individualising the processing to the individual needs of the patient.
To improve the performance of retina implants, there are a number of aspects to consider. It seems to be evident that an increase in the number of electrodes will result in a better visual function. However, it is only an advantage to have a large number of electrodes if each one stimulates a separate set of cells. This can be achieved by either a very close proximity of the electrode to the target cells or a large distance between the electrodes (large electrode pitch). A large pitch has the disadvantage that the resulting large array size is more difficult to implant. Future retina implants will therefore aim to bring the electrodes closer to the retina rather than increase the electrode pitch.
To increase the number of electrodes by decreasing the electrode pitch requires small electrodes. These small electrodes must be able to deliver enough charge to stimulate the cells. High charge levels are needed, especially if, after surgery, a gap between electrode array and retina exists. For a higher charge capacity, new electrode materials need to be established or existing materials need to be improved, for example by enhancing their effective surface areas. One way to reach a better proximity to the target cells is to use penetrating electrodes instead of planar surface electrodes. To use penetrating electrodes clinically, however, surgical methods need to be developed to safely place and explant them without damaging the retina.
If the future of retina implants follows the same path as cochlear implants, the pre-processing and coding of visual information will play a major role. Sophisticated image processing algorithms are necessary to automatically discriminate between important objects (e.g. people or obstacles) and unimportant objects (e.g. textures). Last but not least, clinical trials are necessary to find optimal stimulation coding.
Although a lot of progress has been made in the development of retina implants and in their respective clinical trials to show that these devices do work in principal, there remains a big potential for improvement. Improved image pre-processing and encoding of signals will probably result in a most effective enhancement of retina implant function.
As research in the field of retina implants has been progressing since the 1990s, a lot of fundamental and detailed information is available. This is the case for technical development, as well as pre-clinical and clinical testing. However, research remains ongoing and, therefore, no single document that gives a detailed and complete overview of the material is available.
A useful summary about retina research can be found in the book Artificial Sight: Basic Research, Biomedical Engineering and Clinical Advances (Humayun et al., 2007). In this book, comprehensive background information is given and many approaches are discussed in detail. In Visual Prosthesis and Ophthalmic Devices (Tombran-Tink et al., 2007) some additional approaches are explained. The fundamentals of electrical stimulation of the retina and some results of clinical trials can be found in the book Visual Prosthetics: Physiology, Bioengineering and Rehabilitation (Dagnelie, 2011).
The newest information about retina implants can be found from two conferences. One is the annual meeting of ‘The Association for Research in Vision and Ophthalmology (ARVO)’. Information about the meeting and abstracts can be found at www.arvo.org. The other important congress for retina implants is ‘The Eye and the Chip World Congress of Artificial Vision’ which take places biannually in Detroit. More information can be found at www.eyeson.org/index.php/research/eye-and-chip.
To find detailed information about retina implants, three special issues of the Journal of Neural Engineering can be recommended. These content-selected topics are from ‘The Eye and the Chip World Congress of Artificial Vision’ for the years 2004, 2006 and 2008 (Journal of Neural Engineering 2005, Journal of Neural Engineering 2007, Journal of Neural Engineering 2009).
More information about the authors’ project can be found on the website www.imidevices.com.
Al-Atabany, W.I., Memon, M.A., Downes, S., Mushtaq, B., Degenaar, P. Image enhancements to improve the visual recognition ability of patients with retinal photoreceptor degeneration. Invest Ophthalmol Vis Sci. 50, 2009. [ARVO E-Abstract 4222.].
Caspi, A., Dorn, J.D., McClure, K.H., Humayun, M.S., Greenberg, R.J., McMahon, M.J. Feasibility study of a retinal prosthesis: spatial vision with a 16-electrode implant. Arch Ophthalmol. 2009; 127(4):398–401.
Chow, A.Y., Chow, V.Y., Pardue, M.T., Peyman, G.A., Liang, D.H., Pearlman, J.I., Peachey, N.S. ‘The semiconductor-based microphotodiode array artificial silicon retina. IEEE Int Conf Syst Man Cybernetics. 1999; 4:404–408.
Chow, V.Y., Packo, K.H., Pollack, J.S., Peyman, G.A., Schuchard, R. The artificial silicon retina microchip for the treatment of vision loss from retinitis pigmentosa. Arch Ophthalmol. 2004; 122(4):460–469.
Clements, M., Vichienchom, K., Wentai, Liu, Hughes, C., McGucken, E., DeMarco, C., Mueller, J., Humayun, M., De Juan, E., Weiland, J., Greenberg, R. An implantable power and data receiver and neuro-stimulus chip for a retinal prosthesis system. IEEE Int Symp Circ Syst. 1999; 1:194–197.
Delbeke, J., Pins, D., Michaux, G., Wanet-Defalque, M.C., Parrini, S., Veraart, C. Electrical stimulation of anterior visual pathways in retinitis pigmentosa. Invest Ophthalmol Vis Sci. 2001; 42(1):291–297.
Fink, W., Tarbell, M.A. μAVS2: microcomputer-based artificial vision support system for real-time image processing for camera-driven visual prostheses. Invest Ophthalmol Vis Sci. 50, 2009. [ARVO E-Abstract 4748.].
Fried, S.I., Desai, N.J., Eddington, D.K., Rizzo, J.F. The distribution of voltage across the proximal axon underlies spike initiation in response to electric stimulation of retinal ganglion cells. Invest Ophthalmol Vis Sci. 50, 2009. [ARVO E-Abstract 4568.].
Fujikado, T., Kamei, M., Sakaguchi, H., Kanda, H., Morimoto, T., Ikuno, Y., Nishida, K., Kishima, H., Maruo, T., Konoma, K., Ozawa, M., Nishida, K. Testing of semichronically implanted retinal prosthesis by suprachoroidal-transretinal stimulation in patients with retinitis pigmentosa. Invest Ophthalmol Vis Sci. 2011; 52(7):4726–4733.
Guven, D., Weiland, J.D., Maghribi, M., Davidson, J.C., Mahadevappa, M., Roizenblatt, R., Qiu, G., Krulevitz, P., Wang, X., LaBree, L., Humayun, M.S. Implantation of an inactive epiretinal poly (dimethyl siloxane) electrode array in dogs. Exp Eye Res. 2005; 82(1):81–90.
Hornig, R., Laube, T., Walter, P., Velikay-Parel, M., Bornfeld, N., Feucht, M., Akguel, H., Rössler, G., Alteheld, N., Lütke Notarp, D., Wyatt, J., Richard, G. A method and technical equipment for an acute human trial to evaluate retinal implant technology. J Neural Eng. 2005; 2:129–134.
Hornig, R., Zehnder, T., Velikay-Parel, M., Laube, T., Feucht, M., Richard, G. The IMI Retina Implant System. In: Humayun M., Weiland J.D., Chader G., Greenbaum E., eds. Artificial Sight: Basic Research, Biomedical Engineering, and Clinical Advances. New York: Springer, 2007.
Humayun, M.S., de Juan, E., Jr, Dagnelie G., Greenberg, R.J., Propst, R.H., Phillips, D.H. Visual perception elicited by electrical stimulation of retina in blind humans. Arch Ophthalmol. 1996; 114(1):40–46.
Humayun, M.S., Weiland, J.D., Fujii, G.Y., Greenberg, R.J., Williamson, R., Little, J., Mech, B., Cimmarusti, V., Van Boemel, G., Dagnelie, G., de Juan, E. Visual perception in a blind subject with chronic microelectronic retinal prosthesis. Vision Res. 2003; 43(24):2573–2581.
Humayun, M., da Cruz, L., Dagnelie, G., Mohand-Said, S., Stanga, P., Agrawal, R., Greenberg, R., Argus II Study Group. Interim performance results from the Second Sight® ArgusTM II Retinal Prosthesis Study. Invest Ophthalmol Vis Sci. 51, 2010. [E-Abstract 2022.].
Inomata, K., Tsunoda, K., Hanazono, G., Kazato, Y., Shinoda, K., Yuzawa, M., Tanifuji, M., Miyake, Y. Distribution of retinal responses evoked by transscleral electrical stimulation detected by intrinsic signal imaging in macaque monkeys. Invest Ophthalmol Vis Sci. 2008; 49(5):2193–2200.
Jiang, G., Zhou, D. Technology advances and challenges in hermetic packaging for implantable medical devices. In: Zhou D., Greenbaum E., eds. Implantable Neural Prostheses 2. New York: Springer; 2010:27–61.
Kelly, S.K., Shire, D.B., Chen, J., Doyle, P., Gingerich, M.D., Drohan, W.A., Theogarajan, L.S., Cogan, S.F., Wyatt, J.L., Rizzo, J.F., Realization of a 15-channel, hermetically-encased wireless subretinal prosthesis for the blind. IEEE Eng Med Biol Soc. 3rd. 2009:200–203.
Kim, E.T., Kim, C., Lee, S.W., Seo, J.M., Chung, H., Kim, S.J. Feasibility of microelectrode array (MEA) based on silicone-polyimide hybrid for retina prosthesis. Invest Ophthalmol Vis Sci. 2009; 50(9):4337–4341.
Laube, T., Brockmann, C., Buss, R., Lau, C., Hock, K., Stawski, N., Stieglitz, T., Richter, H.A., Schilling, H. Optical energy transfer for intraocular microsystems studied in rabbits. Graefe’s Arch Clin Exp Ophthalmol. 2004; 242(8):661–667.
Liang, T., Zhao, L., Sui, X., Zhou, C., Ren, Q., Qi, Y. Threshold suprachoroidal-transretinal stimulation current required by different-size electrodes in rabbit eyes. Ophthalmic Res. 2011; 45(3):113–121.
Mahadevappa, M., Weiland, J.D., Yanai, D., Fine, I., Greenberg, R.J., Humayun, M.S. Perceptual thresholds and electrode impedance in three retinal prosthesis subjects. IEEE Trans Neural Syst Rehabil Eng. 2005; 13(2):201–206.
Morimoto, T., Kamei, M., Nishida, K., Sakaguchi, H., Kanda, H., Ikuno, Y., Kishima, H., Maruo, T., Konoma, K., Ozawa, M., Nishida, K., Fujikado, T. Chronic implantation of newly developed suprachoroidal-transretinal stimulation (STS) prosthesis in dogs. Invest Ophthalmol Vis Sci. 2011; 52(9):6785–6792.
Myllymaa, S., Myllyma, K., Korhonen, H., Lammi, M.J., Tiitu, V., Lappalainen, R. Surface characterization and in vitro biocompatibility assessment of photosensitive polyimide films. Colloids Surf B Biointerfaces. 2010; 76(2):505–511.
Ohta, J., Tokuda, T., Kagawa, K., Sugitani, S., Taniyama, M., Uehara, A., Terasawa, Y., Nakauchi, K., Fujikado, T., Tano, Y. ‘Laboratory investigation of microelectronics-based stimulators for large-scale suprachoroidal transretinal stimulation (STS). J Neural Eng. 2007; 4:85–91.
Palanker, D.V., Huie, P., Vankov, A., Aramant, R., Seiler, M., Fishman, H., Marmor, M., Blumenkranz, M. Migration of retinal cells through a perforated membrane: implications for a high-resolution prosthesis. Invest Ophthalmol Vis Sci. 2004; 45(9):3266–3270.
Palanker, D.V., Huie, P., Vankov, A.B., Freyver, Y., Fishman, H., Marmor, M.F., Blumenkranz, M.S. Attracting retinal cells to electrodes for high-resolution stimulation. Proc SPIE, Ophthal Tech XIV. 2004; 5314:306–314.
Parikh, N.J., McIntosh, B.P., Tanguay, A.R., Humayun, M.S., Weiland, J.D., Biomimetic image processing for retinal prostheses: peripheral saliency cues. Conf Proc IEEE Eng Med Biol Soc. 2009:4569–4572. [3–6 September].
Richard, G., Feucht, M., Bornfeld, N., Laube, T., Rossler, G., Velikay-Parel, M., Hornig, R. ‘Multicenter study on acute electrical stimulation of the human retina with an epiretinal implant: clinical results in 20 patients. Invest Ophthalmol Vis Sci. 2005; 46(5):1143.
Richard, G., Hornig, R., Keserü, M., Feucht, M., Chronic epiretinal chip implant in blind patients with retinitis pigmentosa: long-term clinical results. Invest Ophthalmol Vis Sci; 48. 2007. [E-Abstract 666.].
Richard, G., Keserü, M., Zeitz, O., Hornig, R., Post, N. Surgical aspects of a long-term implantation of a wireless chip in blind patients with retinitis pigmentosa. Conf Proc Euretina Congress. 2009. [Nice, 14–17 May.].
Rothermel, A., Wieczorek, V., Liu, L., Stett, A., Gerhardt, M., Harscher, A., Kibbel, S. A 1600-pixel subretinal chip with DC-free terminals and ± 2 V supply optimized for long lifetime and high stimulation efficiency. Conf Proc IEEE Int. Solid-State Circuits. 2008; 144–146. [San Francisco, 3–7 February].
Sachs, H.G., Schanze, T., Brunner, U., Sailer, H., Wieseneck, C. Transscleral implantation and neurophysiological testing of subretinal polyimide film electrodes in the domestic pig in visual prosthesis development. Artif Organs. 2005; 2:57–64.
Schubert, M.B., Stelzle, M., Graf, M., Stert, A., Nisch, W., Graf, H.G., Hammerle, H., Gabel, V.P., Hofflinger, B., Zrenner, E. Subretinal implants for the recovery of vision. IEEE Int Conf Syst Man Cybernet. 1999; 4:376–381.
Schwarz, M., Hauschild, L.E.R., Hosticka, B.J., Huppertz, J., Kneip, T., Kolnsberg, S., Ewe, L., Trieu, H.K. S. Single chip CMOS image sensors for a retina implant system, IEEE Trans Circuit Syst II Analog Digit Signal Proc. 1999; 46(7):870–877.
Schwarz, M., Hauschild, L.E.R., Hosticka, B.J., Huppertz, J., Kolnsberg, S., Mokwa, W., Trieu, H.K. Single chip CMOS imagers and flexible microelectronic stimulators for a retina implant system. Sensors Actuators. 2000; 83:40–46.
Shire, D., Gingrich, M., Retterer, S., Theogarajan, L., Kelly, S., Markova, M., Raj, M., Cogan, S.F., Wyatt, J., Rizzo, J.F. Design and fabrication of an abexterno retinal prosthesis. Invest Ophthalmol Vis Sci. 45, 2004. [ARVO E-Abstract 4177.].
Shire, D.B., Kelly, S.K., Chen, J., Doyle, P., Gingerich, M.D., Cogan, S.F., Drohan, W.A., Mendoza, O., Theogarajan, L., Wyatt, J.L., Rizzo, J.F. Development and implantation of a minimally invasive wireless subretinal neurostimulator. Conf Proc IEEE Trans Biomed Eng. 2009; 56(10):2502–2511.
Stone, J.L., Barlow, W.E., Humayun, M.S., de Juan, E., Milam, A.H. Morphometric analysis of macular photoreceptors and ganglion cells in retinas with retinitis pigmentosa. Arch Ophthalmol. 1992; 110(11):1634–1639.
Sun, Y., Lacour, S.P., Brooks, R.A., Rushton, N., Fawcett, J., Cameron, R.E. Assessment of the biocompatibility of photosensitive polyimide for implantable medical device use. J Biomed Mater Res. 2009; 90(3):648–655.
Taneri, S., Bollmann, F.P., Uhlig, C.E., Thelen, U., Gerding, H. The retina implant project (epi-ret): in vitro and in vivo testing of different tack types for intraocular fixation of retina implants. Invest Ophthalmol Vis Sci. 40(733), 1999. [Abstract number 3877.].
Tunc, M., Humayun, M., Chen, X., Ratner, B.D. Reversible thermosensitive adhesive for retina implants: in vivo experience with plasma deposited poly(N-isoprophyl acrylamide). Retina. 2008; 28(9):1338–1343.
Weiland, J.D., Guven, D., Maghribi, M., Davidson, C., Pannu, S., Mahadevappa, M., Krulevitch, P., Sanchez, R.A., Humayun, M.S. Chronic implantation of an inactive epiretinal poly (dimethyl siloxane) electrode array in dogs. Invest Ophthalmol Vis Sci. 45, 2004. [ARVO E-Abstract 4210.].
Weiland, J.D., Humayun, M.S., Eckhardt, H., Ufer, S., Laude, L., Basinger, B., Tai, Y.C. A comparison of retinal prosthesis electrode array substrate materials. Conf Proc IEEE Eng Med Biol Soc. 4140–4143, 2009.
Weiland, J.D., Guven, D., Magrhibi, M., Davidson, C., Pannu, S., Mahadevappa, M., Krulevitch, P., Sanchez, R.A., Humayun, M.S. Chronic implantation of an inactive poly (dimethyl siloxane) electrode array in dogs. Invest Ophthalmol Vis Sci. 45, 2004. [ARVO E-Abstract 4210.].
International, W.H.O. [Accessed 27 July 2011]. Priority of eye diseases. 2011. Available from:. http://www.who.int/blindness/causes/priority/en/index8.html
Yu, W., Wang, X., Zhao, C., Yang, Z., Dai, R., Dong, F.M. Biocompatibility of subretinal parylene-based Ti/Pt microelectrode array in rabbit for further artificial vision studies. J Ocul Biol Dis Infor. 2009; 2(1):33–36.
Zrenner, E., Miliczek, K.D., Gabel, V.P., Graf, H.G., Guenther, E., Haemmerle, H., Hoefflinger, B., Kohler, K., Nisch, W., Schubert, M., Stett, A., Weiss, S. The development of subretinal microphotodiodes for replacement of degenerated photoreceptors. Ophthalmic Res. 1997; 29(5):269–280.
Zrenner, E., Wilke, R., Sachs, H., Bartz-Schmidt, K.U., Gekeler, F., Besch, D., Greppmaier, U., Harscher, A., Peters, T., Wrobel, W.G., Wilhelm, B., Stett, A. Visual sensations mediated by subretinal microelectrode arrays implanted into blind retinitis pigmentosa patients. Proceedings of the 13th Annual Conference of the IFESS. 2008; 53(Suppl 1):218–220.
Zrenner, E., Bartz-Schmidt, K.U., Benav, H., Besch, D., Bruckmann, A., Gabel, V.P., Gekeler, F., Greppmaier, U., Harscher, A., Kibbel, S., Koch, J., Kusnyerik, A., Peters, T., Stingl, K., Sachs, H., Stett, A., Szurman, P., Wilhelm, B., Wilke, R. Subretinal electronic chips allow blind patients to read letters and combine them to words. Proc Biol Sci. 2010; 278:1489–1497.