The Bionic Eye: a review of multielectrode arrays › ... › 2015_12_bionic-review.pdfBionic Eye devices still in the preclinical research phase, this chapter will summarize the advances

C:/ITOOLS/WMS/CUP-NEW/5905702/WORKINGFOLDER/RRAR/9781107040830C24.3D 294 [294–312] 11.2.2015 2:19PM

24

The Bionic Eye: a reviewof multielectrode arraysKate Fox, Owen Burns, David J. Garrett, Mohit N. Shivdasaniand Hamish Meffin

24.1 Introduction

The notion of creating artificial vision using visual pros-theses has been well represented though science fictionliterature and films. When we think of retinal prostheses,we immediately think of fictional characters like TheTerminator scanning across a bar to assess patrons forappropriately fitting clothing, or Star Trek’s Geordi LaForge with his VISOR, a visual instrument and sensoryorgan replacement placed across his eyes and attachedinto his temples to provide him with vision. Such devicesare no longer farfetched. In the past 20 years, significantresearch has been undertaken across the globe in the racefor a “Bionic Eye”. Advances in Bionic Eye research havecome from improvements in the design and fabricationof multielectrode arrays (MEAs) for medical applications.MEAs are already commonplace in medicine with use inapplications such as the cochlear device, cardiac pace-makers, and deep brain stimulators where interfacing withneuronal cell populations is required.

The use of MEAs for vision prostheses is currently ofsignificant interest. For the most part, retinal prostheseshave dominated the research landscape owing to the easeof access and direct contact to the retinal ganglion nervecells. However, MEAs are also in use for direct stimulationinto the optic nerve [1]. Retinal prostheses bypass thedamaged photoreceptor cells within the retina and insteadreplace the degenerate retina with electrical stimulationto the nerve cells. Using electrical stimulation, stimulatedretinal ganglion cells have been shown to elicit a percept

in the form of a phosphene in blind patients [2–6].Accordingly, the two diseases commonly linked to the jus-tification for Bionic Eye research are age-related maculardegeneration (AMD) and retinitis pigmentosa (RP), dis-eases which lead to progressive loss of photoreceptor cellsand diseases where the patient has had previous vision andthus exhibits prior visual-brain pathways. At present, therehas been no reliable cure for any of the retinal diseases thattarget the photoreceptor cells, and thus the development ofprosthetic devices is a viable clinical treatment option [7–9].

The best physical location for a retinal implant remainsundecided. Although this chapter concentrates on retinalprostheses, other artificial vision devices have been inves-tigated at sites external to the eye such as the optic nerve[10] and the visual cortex [11]. Referring back to retinalprostheses, at present there have been three sites best iden-tified for potential implant positioning: (i) epiretinal; (ii)subretinal; and (iii) suprachoroidal.

Figure 24.1 shows a drawing of an eye, highlighting thethree device positions relative to the retinal ganglion nervecell layer. The implant sites have been selected for variousreasons such as distance from the nerve cells, surgical easeand mechanical stability. The majority of clinical groupshave opted for the epiretinal implantation of their BionicEye devices [3, 12–16]. The epiretinal position offers theadvantage that the surgical procedure is relatively wellunderstood and the device is closely aligned with the siteit is actively stimulating. In particular, the epiretinal posi-tion aligns the device against the intact retinal ganglion cells

Handbook of Bioelectronics, Sandro Carrara and Krzysztov Iniewski. Published by Cambridge University Press.© Cambridge University Press 2015.


and offers the added advantage of not having to overcomethe high resistance of the Mueller glial cell seal in theexternal limiting membrane which is seen in the laterphase of RP patients [17]. However, the epiretinal deviceposition also has its drawbacks. In order to position thedevice against the retina, external fixation components arerequired, such as tacks. Tacking through the retina so closeto the stimulated area is problematic as the trauma of theinsertion of the tack is detrimental to the retinal cells belowand around the tack.

The subretinal implant position also provides a closealignment of the device with the underlying neuronalpopulation it seeks to stimulate (taking advantage of thepre-existing neuronal network, albeit in a highly reorgan-ized manner) as well as providing a mechanically stableposition of the device without the need for external fixation.However, the use of the subretinal location has beenrestricted by the complexity of surgery, the physical disrup-tion of the pigment epithelium and the retina as part ofpositioning the device, and the limitation of space whichmeans that the devices need to be very thin. In comparison,suprachoroidal implantation of the devices offers amechanically stable position without the need for furtherfixation components or the need for wires to penetratethrough the eye wall via the pars planar; however, itsmajor limitation is that the device is positioned some dis-tance from the nerve population it seeks to stimulate. Thismeans that the currents required to stimulate the nerve cellsof the retina are significantly higher, since the current mustpass through much more tissue to reach its destination.

Across the globe, there are a number of retinal pros-theses or “Bionic Eye” devices in clinical trials. Probably themost successful to date have been designed in the UnitedStates of America by the Second Sight group. The secondgeneration device, the Argus II Retinal Prosthesis System,has achieved regulatory approval for market in Europe (CEMark) and has proven to the FDA that its benefit outweighsthe risks. This is the preliminary step towards achievingmarket approval in the United States of America (FDAapproval). Retinal prostheses devices operate like a stand-ard brain–machine interface (BMI). A stimulus is generated

by the prosthetic device and coded into an electrical signalby its electronics (machine) which is transferred into theneural interface at the device–tissue interface (Figure 24.2)[18]. This is detected as a visual percept by the brain, whichelicits an action by the user. Accordingly, for the most part,Bionic Eye devices share common components. The devi-ces each require a stimulating interface comprising an MEA(or multiphotodiode array) and circuitry. Further, eachrequires an input source, often in the form of a visualcamera and clever software and hardware to convert visualimages to electrical signals capable of being transmittedto the interfacing array. Finally, each device requires amethod of placement. In order to achieve such complexmechanisms, the retinal prostheses often require intrao-cular, extraocular, and external components. The intraocu-lar components are usually the stimulation MEA, but someretinal prostheses also implant themicrocircuit. Extraocularcomponents include the image acquisition system, dataprocessing components, and the power and data supplies.The extraocular components are connected to the intra-ocular components using either wireless or wired technol-ogy. Given the large number of investigatory MEAs forBionic Eye devices still in the preclinical research phase,this chapter will summarize the advances of the key groupswith a particular focus on those that have entered theclinical trial phase of development.

The key points of difference between the different visualprosthesis technologies highlighted in this chapter arematerial selection, interfacing, electrode number, and thefinal clinical acuity. The importance of acuity is shown inFigure 24.3. If we consider a visual scene as a combinationof pixels, it can easily be seen that themore pixels, the betterthe image. The goal of the Bionic Eye devices is to achievea visual phosphene, that is, a bright “spot” relative to astimulation (electrical, mechanical, magnetic, auditory).The phosphenes need to be distinguishable in order tobuild useful information, and thus the preferred outcomefor a Bionic Eye prosthesis is for a single electrode to be able

Figure 24.1 The three retinal implant positions. Modified image

courtesy of the Bionics Institute.

Brain Machine

Intent Action

Stimulus

Command

Neural interface Coding Physical interface

Percept

Figure 24.2 A standard brain–machine interface. Image repro-

duced from [18].

The Bionic Eye: a review of multielectrode arrays 295


to generate a single, non-overlapped spot that can be usedto excite a visual map similar to that achieved using indi-vidual pixels. Accordingly, if we consider one electrodeas one pixel, the number of electrodes to generate usefulinformation depends on the information. Referring toFigure 24.3a, the displayed image requires 25 pixels. Thisis likely not sufficient to enable a 25 pixel/electrode systemto read text, but may allow some level of unguided naviga-tion. However, as the pixels increase to 100, 400, and 1032(Figures 24.3b–d), it is clear that the applications increase.At 1032 pixels, it is likely that a retinal prosthesis recipientcan interpret large text.

Of course, it is not as simple as treating one electrode asone pixel, and thus the size of electrodes and pitch chosento elicit a phosphene differs greatly across the clinicaltrial devices. Further, considerable research has beenundertaken as to the best method to use the generatedphosphenes to gain maximum visual capacity (i.e. edgedetection, blending, illumination, and MEA stimulationstrategy). Referring to Figure 24.4, various image processingtechniques may be utilized to optimize phosphene-generated vision. Figure 24.4a shows a variety of ways asingle phosphene could be modeled relative to its peers.Figure 24.4b illustrates the variation in which a syntheticvision prosthesis may model an environment in order tomaximize visual output. In order to achieve the variousvisual “scenes”, different filtering has been applied to theoriginal input signal. Predictive modeling is beneficial tooptimize the mechanism of using the MEA of the retinal

prosthesis to drive the electrical signals through the retinalganglion cells to generate the best representation of theimage the prosthesis seeks to simulate. The predictiveresearch, however, does not necessarily translate to thefinal percept pattern seen by a patient. Reports from BionicEye groups in clinical trial have shown some concern aboutthe interactions between neighbouring phosphenes in termsof overlap, fading, and dominance [1, 2, 4–6, 19–21].

Considering retinal prosthesis devices in clinical trials,the EPIRET device has 25 electrodes [22], the Bionic VisionAustralia device has 35 electrodes, the Intelligent MedicalSystems device has 49 electrodes [23], Second Sight’s ArgusII has 60 electrodes [3], the Boston Retinal Group has 200electrodes [24], and the microphotodiodide array fromRetina Implant AG has 1500 elements [6]. Putting this intocontext, the best reported data from the epiretinal Argus II(Second Sight) and the subretinal Retina Implant AG pro-vide a visual acuity of 20/1260 [3] and 20/550 [5], respec-tively, which remains significantly lower than the 20/200acuity used to represent legal blindness [25].

24.2 Electrode materials for use in multielectrodearrays for retinal stimulation

Selection of electrode materials for neural stimulation isultimately governed by safety concerns. In other words,stimulating electrodes must effectively evoke action poten-tials, in the target tissue, without causing damage. The

(a) (b) (c)

(d) (e) (f)

Figure 24.3 Representative image illustrating the importance of electrode number relative to visual acuity. (a) 25 electrodes; (b) 100

electrodes; (c) 400 electrodes; (d) 1032 electrodes; (e) 2500 electrodes; and (f) original image.

296 Kate Fox, Owen Burns, David J. Garrett, Mohit N. Shivdasani and Hamish Meffin


functioning of a stimulating electrode in simple terms iseasily conceptualised. In most instances a current pulse isdriven into the electrode causing it to become charged.This localized charge affects the environment around theelectrode and can cause nearby neurons to depolarize. Inthis simple view, any electrically conducting material couldfunction as a neural stimulation electrode, the only questionbeing how much charge is needed to evoke a neuralresponse. Of course, the reality is more complicated.

Firstly, electrodes must be constructed from materialswhich are inherently biocompatible. Secondly, considera-tion needs to be given to the voltages that the electrodes willreach during stimulation pulses. On forcing charge into thesurface of an electrode, the energy of the surface is changed.In electrical terms, the difference in energy between theelectrode surface and its environment is described in joulesper coulomb (J/C) of electrons or volts. Inmedical terms it isthe volts that generate tissue damage.

Themost frequently used predictor of safe voltage limitsis the water window. The water window is the voltage rangebounded by the positive and negative voltage values at

which significant water splitting occurs. At high positivevoltages, the surface energy of the electrode is sufficientlylow that electrons are pulled from water molecules (oxida-tive water splitting) into the electrode surface, causing thereaction shown in Equation (24.1).

2H2O−4e−→4H+ + O2 (24.1)

The products of this reaction are protons, which alter pH,and gaseous oxygen, which is understandably undesirablein particular in a closed environment such as the eye. Whenelectrode voltages are sufficiently negative the surfaceenergy is so high that electrons are forced off the electrodeand into water molecules (reductive water splitting), caus-ing the reaction shown in Equation (24.2).

2H2O−2e−→H2 + 2OH

− (24.2)

Similarly, OH− ions alter pH, and gaseous hydrogen isequally undesirable inside the body. The water window isusually established using cyclic voltammetry, where anelectrode of the material in question is cycled from its

a

b

Bitmap Discretephosphenes

1

2

3

Boundedphosphene cells

Fusedphosphenes

Figure 24.4 Processing techniques to optimize the method of translating an image to a visual prosthesis. (A) Methods to control individual

phosphene spots. (B) Filtering the image and then phosphenizing the original image using: (1) mean filtering; (2) histogram equalization,

then mean filtering; and (3) edge detection using Laplacian of Gaussians filtering. Image sourced from [19].



resting (open circuit) potential to voltages where watersplitting is observed to occur in both the positive and neg-ative voltage directions. A typical cyclic voltammogram ofthis nature is shown in Figure 24.5

Water splitting is not the only undesirable reaction thatoccurs at high-magnitude electrode voltages. Many biomo-lecules can be electrochemically degraded, chloride ionsin saline solution can be oxidized according to Equation(24.3), and metal from metal electrodes can be oxidizedforming metal complexes in solution. The most studied ofthese is oxidation of platinum to form platinum hexachlor-ide ions according to Equation (24.4). This reaction not onlyintroduces a toxic foreign metal complex into the patient[27] but also results in dissolution of the electrode.

Cl−+H2O →ClO−+H+ (24.3)

Ptþ 6Cl�→PtCl2�6 þ 4e� ð24:4Þ

The voltage limits at which water splitting occurs arematerial-specific and are used to evaluate the charge injec-tion capacity of an electrode (Qc). Charge injection capacitycarries the units of coulombs per unit area, very oftenµC/cm2 or mC/cm2. This value represents the maximumamount of charge that can be delivered during a stimulationpulse to a cm2 of material without electrode voltage exceed-ing set limits. Usually those limits are derived directly fromthe water window. The charge injection capacity is directlyrelated to the electrochemical capacitance of an electrode,which is described as the rate of change of electrode chargewith changing electrode voltage or C = dQ/dV (C = capaci-tance (F), Q = charge (C), and V = volts (V)). Simply put,an electrode with high capacitance can accommodate a

large amount of charge whilst experiencing a small voltagechange. Such an electrode is more likely to effectively evokeaction potentials in nerve tissue whilst remaining withinsafe voltage limits.

In classical electrochemistry, electrode processes areoften defined as either Faradaic or non-Faradaic reactions.An example of a non-Faradaic reaction is double-layercapacitive charging. Initial charge injected into an elec-trode is accommodated by rearrangement of the electro-lyte solution in the immediate vicinity of the electrode. Theamount of charge that can be stored in this way is a com-plex function of the electrolyte, the material/ electrolyteinteraction and the real surface area of the electrode.Faradaic reactions are those that result in charge transferto or from a chemical species that exists either in thesolution adjacent to the electrode or within the electrodeitself. The definition “reversible Faradaic reaction” refersto a process where charge can be donated to (reduction) orwithdrawn from (oxidation) a chemical species reversibly.A non-reversible Faradaic process is where the products ofthe electrochemical reaction are electrochemically inert,and therefore the charge donated or extracted cannot berecovered.

Some of the most important historical work on thecharacterization of electrodes for neural stimulation, froman electrochemistry and materials point of view, was con-ducted at EIC laboratories by Brummer, Turner, and Roblee[28–30] in the late 1970s and 1980s. In more recent years,Stuart Cogan, also of EIC, has carried the torch and main-tained the laboratories’ leading role in this field [31–34], inparticular through development of iridium oxide as an elec-trode material. A recent review penned by Stuart Cogan [31]summarized the current leading stimulating electrode

0.5Hydrogen

desorptionOxide formation

Wateroxidation

Oxidereduction

Hydrogenadsorption

Oxygen reduction

Potential (V) vs. Ag/AgCl (pH 7.2)

Waterreduction

0.0

–0.5

–1.0Cur

rent

den

sity

(m

A]c

m2 )

–1.5–1.0 –0.5 0.0 0.5 1.0 1.5

Figure 24.5 Cyclic voltammograms recorded using a platinum electrode in pH 7.2 phosphate buffered saline, intended to emulate

biological conditions. CVs were recorded before (gray) and after (black) removal of dissolved oxygen by purging with N2 gas. The peaks

assigned to hydrogen adsorption/desorption and reversible platinum oxide formation are indicated. Also shown are the potential limits of

the water window bounded by the onset of water reduction (−0.6 V) and water oxidation (0.9 V). Image adapted from [26].



materials in terms of charge injection capacity and voltagelimits. The results of that summary are shown in Table 24.1.

24.2.1 Current clinical trial materials for retinalmultielectrode array electrodes

The list of materials in use as stimulating electrodes inretinal implants undergoing clinical trial currently featuresonly platinum (Pt) and titanium nitride (TiN) [6].

24.2.1.1 Platinum (Pt)Platinum iswell established as a biocompatiblemedicalmetal[26]. Accordingly, it is an obvious trial electrode material, andfortuitously it falls into the category of electrochemical pseu-docapacitors, giving it an unusually high Qc for a bare metal.The source of the additional charge injection capacity is thewell-understood electrochemically induced adsorption ofhydrogen that occurs on the platinum surface in aqueouselectrolyte solutions. The reaction is shown in Equation (24.5).

Pt + H2O + e− ↔ Pt − H + OH− (24.5)

The reaction is an electrochemically reversible Faradaicreaction and therefore under charge-balanced conditions(such as a charge-balanced biphasic current pulse) therewill be no net change in the chemical environment at theelectrode surface. Platinum also enjoys a long history ofsuccess in cochlear implants and is therefore the most com-prehensively tested of all the possible electrode materials.Platinum electrodes are also some of the easiest to fabricateeither by machining from readily available high-purity foilsor by conventional metal deposition such as electron beamevaporation. One projected problem with platinum is thatthe charge injection capacity (≈ 150 µC cm−2) is unlikelyto be high enough to be safe for very small electrodes(


high adhesion of diamond films may prove to be advanta-geous in biomedical implants.

24.2.2.3 PolymersConducting polymers are currently undergoing extensiveresearch as stimulating electrodematerials. The front runneramong these is poly(3,4-ethylene dioxythiophene) (PEDOT)[46]. PEDOT is electrochemically deposited directly on theelectrode and can exhibit very high charge injection values(1–4 mC cm−2) [46, 47]. The principal concern with conduct-ing polymers is stability over the very long duty cycleexpected of a vision prosthesis.

24.3 Interfacing systems for retinal implants

In order to generate a retinal implant system, it is criticalthat the interfacing between the input source (i.e. the stim-ulating electronics and the MEA) and the ocular tissue isappropriate. The electrodes are designed under materialrestrictions outlined above in Section 24.1 and often have

strict requirements as to the optimal size, shape, and place-ment. In addition to a series of electrodes, the retinal implantsystems require clever communication hardware and soft-ware to ensure that the appropriate signals are transferredfrom the system through the electrodes and into the appro-priate tissue or neuronal site. Accordingly, the next sectionwill discuss the various interfacing systems used by retinalimplant groups with consideration to the various surgicalalignment positions used by each group. The discussion hasbeen limited to the research groups who have moved frompreclinical research into human clinical trials.

24.3.1 Second Sight Medical Products (Epiretinal)

Second Sight Medical Products have clinically tested twodevice generations. The first generation device, Argus I [48,49], was implanted in six clinical trial patients. The Argus Iis designed and used to perform basic object differentiationand motion tasks [13, 50]. It follows well-establishedplatinum–silicone fabrication techniques. The Argus Iretinal prosthesis features a 4 × 4 platinum electrode array

Figure 24.6 IrOx films formed under a variety of conditions taken from [42].



suspended in a silicone carrier. The electrode array is madeup of a series of platinum disk electrodes 260–520 µm indiameter, with center to center electrode pitch of 800 µmand evenly spaced over a retinal area of 2.84 × 2.84 mm orapproximately 9.5° of visual angle. The electronics drivingthe array is located subdermally and connected to theelectrode array via a subcutaneous cable. There is noexternal unit.

The Argus II [51, 52] retinal prosthesis developed forSecond Sight’s second round of clinical trials is a scaled-upversion of the Argus I (Figure 24.8). The Argus II features a60-channel retinal MEA, an implanted coil for inductivepower, and data transfer between an external unit and elec-tronics unit fixed to the sclera of the eye. The external unitconsists of a video processor, battery, and small camera andtransmitter mounted on a pair of glasses. Connecting thearray to the implanted electronics is a metallized polymerribbon cable that penetrates the sclera in the pars plana [52].The device is held against the retina using a retinal tack.

24.3.2 Intelligent Medical Implants AG (Epiretinal)

Designed for use in acute human trials, the IntelligentMedical Implants (IMI) AG retinal implant is implanted in

the epiretinal location. Referred to as a microcontact film bythe developing researchers [54], the acute devices are 10 µmthick, 1 mm wide and 60 mm long thin-film electrode arraymanufactured using standard microfabrication methods.The implant consists of a visual processor mounted on apair of glasses which communicates wirelessly with theretinal stimulator positioned on the retina. Power is pro-vided by an inductive link between an external antenna anda episcleral receiver which communicates to the MEA by atranscleral cable. Data are communicated through an IRoptical link. The stimulation microchip is positioned epis-clerally providing a 232-channel stimulus (microchip foot-print of 22 m2). The electrodes are coated in IrOx, andelectrode diameters of 50 µm and 100 µm have been usedas well as electrode diameters of 200 µm, and 360 µm. Ahole in the end of themicrocontact films and adjacent to theelectrodes is used for retinal tack fixation in preclinicalstudies.

24.3.3 Retina implant AG (subretinal)

Combining both a microphotodiode array (MPDA) andthin-film techniques, the Alpha IMS subretinal implantdeveloped by the Tubingen group [55, 56] is currently in

Figure 24.7 N-UNCD films prepared for a retinal prosthesis taken from [45].



a second round of clinical trials [6, 57]. Designed to replacethe function of degenerated photoreceptor cells directlyby translating incident light falling onto the retina point bypoint, the Alpha IMS implant features a MPDA with 1500individual light sensitive pixel-generating elements on a20 µm thick polyimide foil also carrying 16 electrode testelectrodes as a 4 × 4 quadruple array (Figure 24.9). Byincorporating the two technologies, the Alpha IMS sub-retinal implant allows for functional testing of multiplestimulation paradigms in one device. The device connectsto an external power supply using a cable across thepigment epithelium, choroid, and sclera, and is thenconnected to a percutaneous plug via a subcutaneousconnector and lead. The microphotodiodes use theintensity of light on the photodiode to adjust the stimuluspulse [58].

24.3.4 EPI-RET GmbH (epiretinal)

Based in Germany, the EPIRET group has now developedthree generations of retinal prostheses. The EPIRET 3device, having now progressed to clinical trials, is fabri-cated using standard thin-film processes. The EPIRET 3consists of two parts: (i) an extraocular unit mounted on

a pair of glasses including a computer system and trans-mitter unit; and (ii) an intraocular unit comprising a poly-imide thin-film structure supporting a receiver module andstimulating electrode array (Figure 24.10). The electrodearray features 25 stimulating electrodes fabricated fromgold and coated with iridium oxide [22, 59]. The 100 µmelectrodes are arranged in a hexagonal array with a centerto center pitch of 500 µm. Power and data are transmittedusing an inductive link from the transmitter positioned atthe front of the eye to a received positioned in the intra-ocular lens.

24.3.5 Osaka University/NIDEK Co(suprachoroidal-transretinal)

Following a similar approach to Second Sight, the OsakaUniversity/NIDEK Co. device is of platinum–silicone con-struction with 49 “bullet” shaped electrodes in a planarMEA. Each individual electrode is connected to animplanted stimulator via a coiled platinum wire. The arrayof platinum electrodes is supported by a silicone substratedesigned to minimize surgical trauma. Each bullet shapedelectrode is 0.5 mm in diameter with a center to centerspacing of 0.7 mm. To date, two patients with retinitis

Electronics case – withstimulator and receiver

Electrode array

Receiving and

transmitting coil

1

Figure 24.8 Schematic of the Argus II implantable prosthesis, including electrode array, hermetic electronics case and RF coil for power and

data transfer. Also shown is the 60-channel device implanted inside an eye. Image sourced with permission from [53].



pigmentosa (RP) have been acutely implanted with theNIDEK device [62].

Earlier devices developed by the NIDEK group forpreclinical studies employed polyimide-based thin-filmprocesses [63, 64]. The device comprises intraocular elec-tronics whereby the microchips used to direct stimuluspulses to electrodes are placed on the reverse side of thestimulating electrode (Figure 24.11). The merit in decreas-ing the distance between microchip and stimulating elec-trode is to decrease (eliminate in this case) the numberand bulk of power and data lines, and the distance thatthey must travel through the human anatomy. This isachieved using a fabrication method known as flip-chipbonding.

Polyimide

Electrodes

MPDA

1 mm

photodiodeelectrode

contacthole

Figure 24.9 The Alpha IMS retinal implant. The device connects to a percutaneous connector behind the ear. The implant comprises a

polyimide foil featuring the 4 × 4 quadruple electrode array with direct electrical stimulation (DS test field) supported by a 3.0 × 3.0mm

CMOS-ship 1500 pixel Microphotodiode array (MPDA). Inset is a magnified image of four of the individual microphotodiodes. Image

modified from [6].

Figure 24.10 The EPIRET 3 system comprises two parts, an external transmitter coil mounted on a pair of glasses and an internal

stimulator chip which provides stimulations pulses to the electrode array. Image sourced from [60] modified from [61].

Microchips

Flexible polyimide

Figure 24.11 Image of the fabricated thin-film retinal array. The

array is positioned on a flexible polyimide cable. The image shows

the reverse side of device showing flip-chip bonded microchips. The

electrode array comprises nine Pt/Au pads per single microchip.

Image modified from [63].



24.3.6 Boston Retinal Implant Group (epi/subretinal)

Although the Boston Retinal Implant Group (BRIG) nowworks on a subretinal device, initial prototype devices andclinical testingwas performed epiretinally. In the early part ofthe 2000s, the BRIG conducted acute stimulation of five blindpatients and one normally signted patient. [65, 66]. Duringthese acute surgical trials, iterative changes to electrodegeometry were based on feedback from electrical stimula-tion. [65] The device comprises a receiver coil on the sclerasurface which communicates via a wire through the eye wallto the MEA and to an external transmitter coil positioned onthe rim of a pair of glasses. Recently the BRIG has beendeveloping a fully implantable epiretinal device for humantrials. [24, 67–69] The device illustrated in Figure 24.12 isintended to comprise 200 individually controlled stimulatingelectrodes powered wirelessly via an inductive link.

24.3.7 Bionic Vision Australia (suprachoroidal)

In 2012, three clinical trial patients were implanted with aprototype wide-view retinal prosthesis developed by theBionic Vision Australia (BVA) consortium[70]. The prototype

device, featuring thirty-five platinum electrodes, twentyindividually accessible 600 µm diameter electrodes, thirteen600 µm electrodes shorted together and two 2 mm returnelectrodes was developed to assess the safety efficacy of asuprachoroidal placement for a retinal prosthesis inhumans. Earlier work [71–74] conducted by the BionicsInstitute in preclinical research led to the development ofthe device seen in Figure 24.13 using platinum–siliconefabrication techniques. Individual insulated platinum wiresare welded to each electrode and are coiled to form a helicalcable. The device does not at present contain implantableelectronics and instead connects to an external stimulatorvia a percutaneous pedestal secured behind the ear suchthat the external stimulator can access the electrodesdirectly, allowing greater control of electrical stimulation.

24.3.8 Engineering development – past and future

24.3.8.1 Optobionics (subretinal)Developed by Alan and Vincent Chow, the Artificial SiliconRetina (ASR) microchip is a subretinal MEA comprising a2 mm diameter silicon microchip with a reported thicknessof 25 μm [75]. The microchip comprises individually isolated5000 microphotodiode pixels (20 × 20 μm) with a 9 × 9 μmiridium oxide electrode bonded to each pixel. The device ispowered solely by incident light without the need for anyexternal components such as wires. The reported illumina-tion is 800 foot-candles from a pixel current of 8–12 nA [75].Although the ASR showed promising results from its pilotclinical trial, Optobionics is currently not undertaking anyfurther clinical trials.

24.3.8.2 Seoul National University (suprachoroidal)Also developing a suprachoroidal retinal prosthesis, theKorean team from the Seoul National University are

Biocompatibleflex substrate

Power and datareceiver coils

Microfabricated thin-filmelectrode array

Controllermicrochip

Power supplycomponents

Figure 24.12 Artist’s rendering of the next-generation BRIG subretinal prosthesis. The prosthesis comprises a thin-film electrode array, a

hermetically sealed capsule containing stimulating electronics, and an inductive coil that receives power and data. Connecting the three

components is a polyimide-based flexible cable. Also pictured is the first-generation BRIG device. Image sourced from [67].

Figure 24.13 Prototype wide-view retinal prosthesis developed by the

Bionics Institute for chronic implantation in the suprachoroidal space.



delivering promising results in preclinical studies [76, 77],The group are continuing to developing new fabricationtechniques using liquid crystal polymers [78] to overcomethe reported biostability issues with polyimide, which tendsto absorb moisture and delaminate [79]. Chronic implanta-tion of crystal polymer substrates have shown to be welltolerated by the eye with no damage to the neural retinaafter 3 months of implantation [76, 77].

24.3.8.3 Bionic Vision AustraliaThe Bionic Vision Australia consortium is currently in thepreclinical testing phase of two additional retinal electrodedevices. The first prototype builds on the platinum–siliconefabrication methods developed in their first clinical device[70]. The device being explored is a 98-electrode retinalimplant for suprachoroidal implantation advanced usinglaser patterning techniques [80]. In this work, a platinum–silicone device is fabricated as with a polyimide thin-filmdevice, where a silicone elastomer is used as a base layer,platinum foil is then applied, and using a laser, stimulatingelectrodes and connection tacks are cut into the platinumfoil. Excess platinum is removed, and silicone elastomer isspun-coated over the platinum as an insulation layer.Finally the stimulating electrodes are exposed by laser pat-terning. To further increase charge injection limits, theexposed electrodes can either be roughed by laser ablation,or coated with conducting polymers [47].

The second prototype device comprises a MEAmade ofdiamond. The MEA is fabricated using microwave plasmaenhanced deposition of insulating polycrystalline and con-ductive ultrananocrystalline diamond in order to produceareas on the array of known conductivity [43, 45, 81]. TheMEA is produced using standard microfabrication technol-ogy, and accordingly multitudes of electrodes can be pro-duced. The second-generation device is expected to have1032 electrodes in a 16 × 16 array coupled directly toimplanted microelectronics and housed in a hermetic dia-mond capsule.

24.3.9 Microelectronics to drive the MEA

MEAs cannot be thoroughly discussed without considering,albeit briefly, the electronics required to induce a stimulusin the MEA. Ideally, the microelectronics package is bestused when intraocularly implanted as this is the best mech-anism to prevent wires from transversing the eye wall. Thekey design considerations for the microelectronics aresize, power consumption, and output; that is, the micro-electronics needs to be able to drive enough current intothe tissue to be able to excite a phosphene [82]. Size is oftenproblematic. In the push for more stimulation electrodes,the microelectronics remains limited by the minimumrequired feature size of its electrode driving circuitry.Surgically, the entire implanted device, included the micro-electronics package, is required to fit through an incision

wound. This places some limitation on the maximum sizewhich is probably no more than 5–6 mm2.

Microelectronics are generally comprised of highlytoxic materials such as copper, aluminum, and indium.Thus, in order to make the device fully implantable, her-metic housing is a crucial design consideration. There aregenerally two types of housing used at present: an encap-sulating capsule, or a thin-film coating. Hard capsules aremade of biocompatible or bioinert materials such asceramics, metals such as titanium, and glass. The largestproblems with having a secondary capsule are firstly thatthe capsule needs to be sealed within temperature limita-tions of the internal electronics, and secondly, that thecapsule requires feedthroughs to enable the microelec-tronics to connect to the MEA. Thin-film coatings providea less bulky alternative as the film will conform to theunderlying microelectronics. However, the majority ofthin-film coatings such as silicones do not provide a long-term hermetic seal [82]. Paralene, a conformal polymer, hasbeen used as a thin-film coating in retinal devices, with acoated device holding its function after 14 months ofimplantation [83].

24.4 Performance of retinal prostheses

24.4.1 Phosphenes

The term “phosphene” is defined as a visual percept whichcan be induced by mechanical, electrical, magnetic, orauditory stimulation to the eye without light having toenter the eye. In the case of retinal prostheses, phospheneselicited through electrical stimulation of a single electrodeare usually perceived as a whitish or yellowish flash of light,often without a discrete shape but sometimes with a well-defined shape. To this end, they have been reported toappear as points, spots, bars, or chaotic structures of bothcolored and colorless light [84–86]. Electrically inducedphosphenes have rarely been defined as simple percepts.In some cases, phosphenes have been described as perceptsthat can appear darker than the patients’ naturally per-ceived background light, sometimes with both bright anddark areas in the same phosphene and sometimes even indifferent colors. In terms of shape, phosphenes have typi-cally been described to be roundish but can also be elon-gated, arc-like, or just “blob-like” without any form.

It is unclear as to what ultimately determines the per-ceptual properties of a phosphene; however, the brightness,shape, size, duration, and location of perceived phosphenesare well known to vary with the location and the number ofelectrodes stimulated in the retina, as well as the parame-ters used for electrical stimulation. Despite the complexitiesin phosphenes induced through electrical stimulation, theyare considered to be the primitive building blocks of pros-thetic vision that can be used to display an image akin topixels on a computer display. It is thus generally accepted



that a large number of non-overlapping phosphenes willresult in improved resolution and better vision for a retinalprosthesis. Below we present some of the properties ofphosphenes and the performance levels achievable withphosphene vision in blind humans described in the litera-ture by themajor retinal prosthesis companies and researchgroups.

24.4.2 Second Sight medical products (epiretinal)

Initial results by Humayun et al. from acute stimulation ofthe retina, using platinum wire electrodes under local anes-thesia, provided the much-required proof of concept thatlocalized phosphenes can be induced from focal epiretinalstimulation [2]. Patients in this study described the phos-phenes as round, ring, line, or rectangle shapes that werethe size of a pin, pea, or match head, and mostly yellow incolor [2]. In a follow-up study using an array of platinumelectrodes, it was reported that simple patterns such ascontinuous lines, the letter H, and a box shape could beperceived upon sequential stimulation of multiple electro-des [87]. Although most of the thresholds reported in theirearlier study [2] were much above those considered safefor long-term stimulation, Humayun et al. posited thatimprovements in electrode design that reduced the dis-tance between the electrodes and the retina, and carefulselection waveform parameters would enable lower thresh-olds in the future. Subsequently, with the Argus I device [48,49], the group was able to demonstrate low threshold per-cepts and showed that long-term thresholds were depend-ent on impedance and distance of the electrode array fromthe retina, but not on electrode size and retinal thickness.Using just 16 electrodes present in this device, patientswere able to perform simple object counting and objectdifferentiation as well as some motion discrimination andorientation identification tasks [13, 50].

To date, with the Argus I device implanted in sixpatients, the Argus II device in over 30 patients, and over10 years of experience with chronic stimulation in blindhumans, Second Sight have been able to establish them-selves as the leaders in retinal prostheses. They are the onlygroup to have developed a commercially available retinalprosthesis system that has been granted CE approval inEurope and HDE approval in the United States. Subjectsimplanted with the latest Argus II system containing 60electrodes have been reported [3, 51] to be able to performsignificantly better in object localization [88], motion dis-crimination [51, 89, 90], and discrimination of orientedgratings [91] with the system turned on than with it turnedoff, even two years after implantation [3]. In addition,patients have been able to identify different-sized shapes,both filled and outlined [92, 93], are able to consistentlyread short sentences [94], and are able to perceive complexpatterns with intersecting lines [95]. The best visual acuity todate is found to be 20/1260 [3]. Recently, it was reported that

patients can even perceive up to eight different combina-tions of colors depending upon the frequency of stimulation[96, 97] as well as simultaneous colors on two separateelectrodes [97]. Numerous challenges still remain withregards to the fading of percepts in many patients [20], theability to place electrodes close enough to the retinal sur-face, stimulation of fibers of passage leading to arc-likepercepts [98], and being able to systematically control thebrightness, shape, and size of phosphenes through appro-priate modulation of stimulation parameters [86].

24.4.3 Intelligent Medical Implants AG (Epiretinal)

A wireless powered 49-electrode implant dubbed the“learning encoder retinal prosthesis system” has beendeveloped by this group [23] and was implanted in a totalof four retinitis pigmentosa patients.[99] In at least onepatient, follow-ups have been conducted over a period ofnearly three years [14]. Thresholds were found to remainstable over several months of testing sessions and reportedto be well under the safe limits of charge [16, 99].Perceptions of simple patterns such as a vertical/horizontalbar or a cross and discriminability of individual phospheneshave been possible with stimulation of appropriate electro-des [99, 100]. A recent report by the group on acute stim-ulation in 20 patients under anesthesia using an MEAcontaining a range of electrode sizes indicated reliablethreshold measurements ranging between 20 and 768 nCin 15 patients [101]. The large variability in thresholdsin their study was attributed to a combination of varyingdegree of retinal degeneration, different electrode sizes, andvarying distances between the electrode array and retina[101]. Patients mostly perceived phosphenes that were lightin contrast to their background, with the exception of onepatient who reported a dark phosphene [101]. Phospheneswere found to have a wide variety of described shapesranging from simple percepts of a circle or ring to morecomplex shapes such as a hash mark or a fan-like arrange-ment of fluorescent tubes [101]. In functional assessment,the four patients were able to undertake basic localizationtasks and recognize simple light patterns [102].

24.4.4 Retina Implant AG (subretinal)

The first-generation device from Retina Implant AG wasimplanted in eight blind patients for a duration of 4 weeks,and in four more patients for a duration of 4 months [103].The implanted system contained two parts, one a 4 × 4electrode array that was directly accessed externallythrough a cable exiting the skin and the other an encapsu-lated microphotodiode chip with 1500 individual photodi-odes each with its own amplifier circuitry and stimulationelectrode, which was externally powered through the samecable [6]. Psychophysical results from direct stimulationof the 4 × 4 electrode array have been presented in Wilkeet al. [103], while results from task-based tests using the



photodiode chip have been presented by Zrenner et al. [6]Using the 4 × 4 array, six of the patients were able to perceivephosphenes from both single and simultaneous multipleelectrode stimulation; in a further two patients, activationof multiple electrodes was necessary to generate a phos-phene; and in three patients no phosphenes were per-ceived. One subject withdrew consent from the study[103]. Unlike the abovementioned epiretinal studies, phos-phenes implanted with the subretinal device were generallydescribed as simple and tended to be yellowish round oroval-shaped percepts [103]. Thresholds ranged between 2.1and 91.8 nC with most of the variance attributed to patientsand individual electrodes within a patient [104]. Thresholdswere also found to depend moderately on the number ofelectrodes stimulated simultaneously, time after implanta-tion (particularly in the early days post implantation), pulseduration, pulse shape, and frequency [104]. A number ofdifferent patterns such as single lines of different orienta-tions, parallel lines, geometric objects such as a triangle orsquare, and simple letters were shown to be perceived uponsequential stimulation of electrodes [103].

However, because of the reported problem of percep-tual fading upon repetitive stimulation of a single electrode[4, 6, 21], inter-stimulus intervals for sequential stimulationhave to be kept greater than 100 ms (


this time, they were able to perform acute stimulation in fivepatients with severe retinitis pigmentosa and one normallysighted human [114, 115]. Phosphenes were elicited inthree blind patients and in the normally sighted patient,with the lowest recorded threshold in the blind patients 4times higher than in the normally sighted patient [115].Sometimes, multiple phosphenes were perceived withsingle electrode stimulation [114]. Multiple electrode stim-ulation in a given pattern did not always induce the percep-tion of the pattern, but most phosphenes were reproducible[114]. Since abandoning the epiretinal approach, the grouphas developed two generations of subretinal devices capableof chronic stimulation through over 256 electrodes, one fortesting in animals and the other intended for humans [113].Animal studies with full wireless devices have been ongoingfor the past two years [113]. The group has recently set-uptwo companies, named Bionic Eye Technologies and VisusTechnologies, to commercialize their technologies [116].

24.4.8 Bionic Vision Australia (suprachoroidal)

Bionic Vision Australia has recently commenced pilot clin-ical trials in three RP patients using a suprachoroidallyimplanted platinum–silicone MEA device. The implantedpatients have reportedly been able to see phosphenes [70];however, any further details of the pilot studies are yet to bereported.

24.5 Future thoughts

Although retinal prostheses are now a reality, several chal-lenges remain to achieve useful vision that will assist intasks of daily living and provide navigation, face recogni-tion, and reading abilities for the blind. The biggest chal-lenge is designing an appropriate stimulation strategy thatwill make clever use of primitive phosphenes to display theintended image. Massive efforts are under way globally toimprove resolution through a combination of clever front-end vision processing, increasing the number of physicalelectrodes, placing electrodes as close as possible to thetarget neurons or using novel methods for electrical stim-ulation. Development of newer electrode materials, encap-sulation technologies, surgical techniques, and image andsignal processing as part of a multidisciplinary approach inthe next 10 years will all help towards achieving the goal ofrestoring vision to the blind.

References

[1] W.H. Dobelle, M.G. Mladejovsky, and J. P. Girvin, “Artificialvision for the blind: electrical stimulation of visual cortexoffers hope for a functional prosthesis,” Science, vol. 183,pp. 440–444, 1974.

[2] M. S. Humayun, E. de Juan Jr, G. Dagnelie et al., “Visual per-ception elicited by electrical stimulation of retina in blindhumans,” Arch. Ophthalmol., vol. 114, no. 1, pp. 40–6, 1996.

[3] M. S. Humayun, J. D. Dorn, L. da Cruz et al., “Interim resultsfrom the international trial of Second Sight’s visual prosthe-sis,” Ophthalmology, vol. 119, no. 4, pp. 779–88, 2012.

[4] E. Zrenner, H. Benav, A. Bruckmann et al., “Electronicimplants provide continuous stable percepts in blind volun-teers only if the image receiver is directly linked to eyemove-ment,” ARVO Meeting Abstr., vol. 51, no. 5, pp. 4319, 2010.

[5] E. Zrenner, K. U. Bartz-Schmidt, F. Gekeler et al., “Seeingwith subretinal electronic implants: study in ten patientswith wireless implant Alpha-IMS,” ARVO Meeting Abstr.,vol. 53, no. 6, pp. 6948, 2012.

[6] E. Zrenner, K. U. Bartz-Schmidt, H. Benav et al., “Subretinalelectronic chips allow blind patients to read letters and com-bine them to words,” Proc. Biol. Sci. Roy. Soc., vol. 278, no.1711, pp. 1489–97, 2011.

[7] A. Kusnyerik, K. Karacs, and A. Zarandy, “Vision restorationand vision chip technologies,” Proc. Computer Sci., vol. 7,pp. 121–124, 2011.

[8] A. J. Smith, J.W. Bainbridge, and R. R. Ali, “Prospects forretinal gene replacement therapy,” Trends Genet., vol. 25,no. 4, pp. 156–165, 2009.

[9] J. Menzel-Severing, “Emerging techniques to treat cornealneovascularisation,” Eye, vol. 26, no. 1, pp. 2–12, 2012.

[10] C. Veraart, M.-C. Wanet-Defalque, B. Gérard, A. Vanlierde,and J. Delbeke, J., “Pattern recognition with the optic nervevisual prosthesis,”Artif. Organs, vol. 27, no. 11, pp. 996–1004,2003.

[11] R. A. Normann, “Toward the development of a corticallybased visual neuroprosthesis,” J. Neural Eng., vol. 6, no. 3,pp. 035001, 2009.

[12] T. Stieglitz, “Development of a micromachined epiretinalvision prosthesis,” J. Neural Eng., vol. 6, no. 6, pp. 065005,2009.

[13] A. Caspi, J. D. Dorn, K. H. McClure et al., “Feasibility study ofa retinal prosthesis: spatial vision with a 16-electrodeimplant,” Arch. Ophthalmol., vol. 127, no. 4, pp. 398–401,2009.

[14] M. Keserue, N. Post, R. Hornig et al., “Long term tolerabilityof the first wireless implant for electrical epiretinal stimula-tion,” ARVO Meeting Abstr., vol. 50, no. 5, pp. 4226, 2009.

[15] S. Klauke, M. Goertz, S. Rein et al., “Stimulation with a wire-less intraocular epiretinal implant elicits visual percepts inblind humans: results from stimulation tests during theEPIRET3 prospective clinical trial,” Invest. Ophthalmol. Vis.Sci., vol. 52, no. 1, pp. 449–455, 2011.

[16] G. Richard, M. Keserue, R. Hornig et al., “Long-term stabilityof stimulation thresholds obtained from a human patientwith a prototype of an epiretinal retina prosthesis,” ARVOMeeting Abstr., vol. 50, no. 5, pp. 4580, 2009.

[17] A. Vugler, J. Lawrence, J. Walsh et al., “Embryonic stem cellsand retinal repair,” Mech. Dev., vol. 124, pp. 807–829, 2007.

[18] J. P. Donoghue, “Connecting cortex to machines: recentadvances in brain interfaces,” Nature Neurosci., vol. 5,pp. 1085–1088, 2002.

[19] S. C. Chen, G. J. Suaning, J.W. Morley et al., “Simulatingprosthetic vision: I. Visual models of phosphenes,” VisionRes. vol. 49, pp. 1493–1506, 2009.

[20] A. Perez Fornos, J. Sommerhalder, L. da Cruz et al.,“Temporal properties of visual perception on electrical stim-ulation of the retina,” Invest. Ophthalmol. Vis. Sci., vol. 53,no. 6, pp. 2720–31, May, 2012.



[21] R. G. Wilke, U. Greppmaier, K. Stingl et al., “Fading of per-ception in retinal implants is a function of time and spacebetween sites of stimulation,” ARVO Meeting Abstr., vol. 52,no. 6, pp. 458, April 22, 2011.

[22] F.H. Koch C, Nolten U, Gortz M, Mokwa W, “Fabricationand assemble techniques for a 3rd generation wireless epi-retinal prosthesis.,” in Proc of MME 2008, 19th Workshopon Micromachining, Micromechanics and Microsystems,Aachen, Germany, 2008, pp. 365–368.

[23] M. Feucht, T. Laube, N. Bornfeld et al., “[Development of anepiretinal prosthesis for stimulation of the human retina],”Ophthalmologe, vol. 102, no. 7, pp. 688–91, Jul, 2005.

[24] S. K. Kelly, D. B. Shire, J. Chen et al., “A hermetic wirelesssubretinal neurostimulator for vision prostheses,” IEEETrans. Biomed. Eng., vol. 58, no. 22, pp. 3197–3205, 2011.

[25] H. Lorach, O. Marre, J.-A. Sahel et al., “Neural stimulation forvisual rehabilitation: Advances and challenges,” J. Physiol.Paris, http://dx.doi.org/10.1016/j.jphysparis.2012.10.003, 2012.

[26] E.M. Hudak, J. T. Mortimer, and H. B. Martin, “Platinum forneural stimulation: voltammetry considerations,” J. NeuralEng., vol. 7, pp. 026005, 2010.

[27] W. F. Agnew, T. G.H. Yuen, R.H. Pudenz et al.,“Neuropathological effects of intracerebral platinum saltinjections,” J. Neuropathol. Exp. Neurol., vol. 36, no. 3,pp. 533–546, 1977.

[28] S. B. Brummer, L. S. Robblee, and F. T. Hambrecht, “Criteriafor selecting electrodes for electrical-stimulation – theoreticaland prectical considerations,” Ann. New York Acad. Sci., vol.405, pp. 159–171, 1983.

[29] S. B. Brummer, and M. J. Turner, “Electrochemical consider-ations for safe electrical-stimulation of nervous-system withplatinum-electrodes,” IEEE Trans. Biomed. Eng., vol. 24, no.1, pp. 59–63, 1977.

[30] J. McHardy, L. S. Robblee, J.M. Marston et al., “Electrical-stimulation with Pt electrodes: 4 factors influencing Pt dis-solution in inorganic saline,” Biomaterials, vol. 1, no. 3,pp. 129–134, 1980.

[31] S. F. Cogan, “Neural stimulation and recording electrodes,”Annu. Rev. Biomed. Eng., vol. 10, pp. 275–309, 2008.

[32] S. F. Cogan, A. A. Guzelian, W. F. Agnew et al., “Over-pulsingdegrades activated iridium oxide films used for intracorticalneural stimulation,” J. Neurosci. Meth., vol. 137, no. 2,pp. 141–150, 2004.

[33] S. F. Cogan, P. R. Troyk, J. Ehrlich et al., “In vitro comparisonof the charge-injection limits of activated iridium oxide(AIROF) and platinum-iridium microelectrodes,” IEEETrans. Biomed. Eng., vol. 52, no. 9, pp. 1612–1614, 2005.

[34] S. F. Cogan, J. Ehrlich, T. D. Plante et al., “Sputtered iridiumoxide films for neural stimulation electrodes,” J. Biomed.Mater. Res. B Appl. Biomater., vol. 89B, no. 2, pp. 353–361,2009.

[35] T. L. Rose, and L. S. Robblee, “Electrical-stimulation with Ptelectrodes: 8 electrolytically safe charge injection limits with0.2ms pulses,” IEEE Trans. Biomed. Eng., vol. 37, no. 11,pp. 1118–1120, 1990.

[36] X. Beebe, and T. L. Rose, “Charge injection limits of activatediridium oxide electrodes with 0.2ms pulses in bicarbonatebuffered saline,” IEEE Trans. Biomed. Eng., vol. 35, no. 6,pp. 494–495, 1988.

[37] S. F. Cogan, P. R. Troyk, J. Ehrlich et al., “Potential-biased,asymmetric waveforms for charge-injection with activated

iridium oxide (AIROF) neural stimulation electrodes,” IEEETrans. Biomed. Eng., vol. 53, no. 2, pp. 327–332, 2006.

[38] L. S. Robblee, M. J. Manguadis, E. D. Lasinky et al., “Chargeinjection properties of thermally-prepared iridium oxidefilms,”Mat. Res. Soc. Symp. Proc., vol. 55, pp. 303–310, 1986.

[39] S. F. Cogan, T. D. Plante, J. Ehrlich et al., “Sputtered iridiumoxide films (SIROFs) for low-impedance neural stimulationand recording electrodes,” Proc. 26th Annu. Int. Conf. IEEEEng. Med. Biol. Soc., pp. 4153–4156, 2004.

[40] E.M. Schmidt, F. T. Hambrecht, and J. S. McIntosh, “Intra-cortical capacitor electrodes – preliminary evaluation,”J. Neurosci. Meth., vol. 5, no. 1–2, pp. 33–39, 1982.

[41] J. D. Weiland, D. J. Anderson, and M. S. Humayun, “In vitroelectrical properties for iridium oxide versus titanium nitridestimulating electrodes,” IEEE Trans. Biomed. Eng., vol. 49,no. 12, pp. 1574–1579, 2002.

[42] G. Ganske, E. Slavcheva, A. van Ooyen et al., “Sputteredplatinum-iridium layers as electrode material for functionalelectrostimulation,” Thin Solid Films, vol. 519, no. 11,pp. 3965–3970, 2011.

[43] D. J. Garrett, K. Ganesan, A. Stacey et al., “Ultra-nanocrystalline diamond electrodes: optimization towardsneural stimulation applications,” J. Neural Eng., vol. 9, no.1, pp. 016002, 2011.

[44] A. E. Hadjinicolaou, R. T. Leung, D. J. Garrett et al., “Electricalstimulation of retinal ganglion cells with diamond andthe development of an all diamond retinal prosthesis,”Biomaterials, vol. 33, no. 24, pp. 5812–5820, 2012.

[45] K. Ganesan, D. J. Garrett, A. Ahnood et al., “An all-diamond,hermetic electrical feedthrough array for a retinal prosthe-sis,” Biomaterials, vol. 35, no. 3, pp. 908–915, 2014.

[46] R. A. Green, R. T. Hassarati, L. Bouchinet et al., “Substratedependent stability of conducting polymer coatings on med-ical electrodes,” Biomaterials, vol. 33, no. 25, pp. 5875–5886,2012.

[47] R. A. Green, F. Devillaine, C. Dodds et al., “Conducting poly-mer electrodes for visual prostheses,” in Proc. 32nd Annu.Int. Conf. IEEE Eng. Med. Biol. Soc., Argentina, 2010.

[48] C. de. Balthasar, S. Patel, A. Roy et al., “Factors affectingperceptual thresholds in epiretinal prostheses,” Investig.Ophthalmol. Visual Sci., vol. 49, no. 6, pp. 2303–2314, 2008.

[49] M. Mahadevappa, J. D. Weiland, D. Yanai et al., “Perceptualthresholds and electrode impedance in three retinal pros-thesis subjects,” IEEE Trans. Neural Systems Rehab. Eng.,vol. 13, no. 2, pp. 201–206, 2005.

[50] D. Yanai, J. D. Weiland, M. Mahadevappa et al., “Visualperformance using a retinal prosthesis in three subjectswith retinitis pigmentosa,” Am. J. Ophthalmo.l, vol. 143,no. 5, pp. 820–827, 2007.

[51] J. D. Dorn, A. K. Ahuja, A. Caspi et al., “The detection ofmotion by blind subjects with the epiretinal 60-electrode(Argus II) retinal prosthesis,” Arch. Ophthalmol., pp. 1–7,2012.

[52] M. S. Humayun, J. D. Dorn, L. d. Cruz et al., “Interim resultsfrom the international trial of second sight’s visual prosthe-sis,” Ophthalmology, no. 119, pp. 779–788, 2012.

[53] Fernandes R, Diniz B, Ribeiro R et al., “Artificial visionthrough neuronal stimulation,” Neurosci. Lett., vol. 519,no. 2, pp. 22–128, 2012.

[54] R. Hornig, T. Laube, P. Walter et al., “Amethod and technicalequipment for an acute human trial to evaluate retinal



implant technology,” J. Neural Eng., vol. 2, no. 1, pp. S129–134, 2005.

[55] F. Gekeler, P. Szurman, S. Grisanti et al., “Compound sub-retinal prostheses with extra-ocular parts designed forhuman trials: successful long-term implantation in pigs,”Graefe’s Archive Clin. Exp. Ophthalmol., vol. 245, no. 2,pp. 230–241, 2006.

[56] L. L. Albrecht Rothermel, N. P. Aryan, M. Fisher et al., “ACMOS chip with active pixel array and specific test featuresfor subretinal implantation,” IEEE J. Solid-State Circuits, vol.44, no. 1, pp. 290–300, 2009.

[57] F. Gekeler, A. Kopp, H. Sachs et al., “Visualisation of activesubretinal implants with external connections by high-resolution CT,” Br. J. Ophthalmol., vol. 94, no. 7, pp. 843–847, 2010.

[58] G. J. Chader, J. D. Weiland, and M. S. Humayun, “Artificialvision: needs, functioning, and testing of a retinal electronicprosthesis,” Progress in Brain Research, Verhaagen J., ed.Elsevier, 2009.

[59] M. Schwarz, L. Ewe, R Hausschild et al., “Single chip CMOSimagers and flexible microelectronic stimulators for a reti-nal implant system,” Sensors Actuators, no. 83, pp. 40–46,2000.

[60] Weiland JD, A. K. Cho, and M. Humayun, “Retinal pros-theses: current clinical results and future needs,”Opthalmology, vol. 118, no. 11, pp. 2227–2237, 2011.

[61] S. Klauke, M. Goertz, S. Rein et al., “Stimulation with a wire-less intraocular epiretinal implant elicits visual percepts inblind humans,” Invest. Ophthalmol. Vis. Sci., vol. 52, no. 1,pp. 449–455, 2010.

[62] T. Fujikado, M. Kamei, H. Sakaguchi et al., “Testingof semichronically implanted retinal prosthesis bysuprachoroidal-transretinal stimulation in patients withretinitis pigmentosa,” Invest. Ophthalmol. Vis. Sci., vol. 52,no. 7, pp. 4726–4733, 2011.

[63] J. Ohta, T. Tokuda, K. Kagawa et al., “Laboratory investiga-tion of microelectronics-based stimulators for large-scalesuprachoroidal transretinal stimulation (STS),” J. NeuralEng., vol. 4, no. 1, pp. S85–91, 2007.

[64] T. Tokuda, R. Asano, S. Sugitani et al., “In vivo stimulationon rabbit retina using CMOS LSI-based multi-chip flexiblestimulator for retinal prosthesis,” Proc. 29th Annu. Int. Conf.IEEE Eng. Med. Biol. Soc., pp. 5791–5794, 2007.

[65] J. F. Rizzo, J. wyatt, L. J. et al., “Methods and perceptualthresholds for short-term electrical stimulation of humanretina with microelectrode arrays,” Invest. Ophthalmol. Vis.Sci., vol. 44, no. 12, pp. 5355–5361, 2003.

[66] J. F. Rizzo, J. Wyatt, L. J. et al., “Perceptual efficacy of elec-trical stimulation of human retina with a microelectrodearray during short-term surgical trials,” Invest. Ophthalmol.Vis. Sci., vol. 44, no. 12, pp. 5362–5369, 2003.

[67] S. K. Kelly, D. B. Shire, J. Chen et al., “Communication andcontrol system for a 15-channel hermetic retinal prosthesis,”Biomed. Signal Processing Control, vol. 6, no. 4, pp. 356–363,2011.

[68] S. K. Kelly, D. B. Shire, J. Chen et al., “Realization of a 15-channel hermetically-encased wireless subretinal prosthesisfor the blind,” in Proc. 31st Annu. Int. Conf. IEEE Eng. Med.Biol. Soc., pp. 200–203, 2009.

[69] D. B. Shire, S. K. Kelly, J. Chen et al., “Development andimplantation of a minimally invasive wireless subretinal

neurostimulator,” IEEE Trans. Biomed. Eng., vol. 56, no. 10,pp. 2502–2511, 2009.

[70] Bionic Vision Australia, “All of a sudden I could see a littleflash of light. It was amazing.” http://bionicvision.org.au [30August, 2012].

[71] R. Cicione, M.N. Shivdasani, J. B. Fallon et al., “Visual cortexresponses to suprachoroidal electrical stimulation of theretina: effects of electrode return configuration,” J. NeuralEng., vol. 9, no. 3, pp. 036009, 2012.

[72] M.N. Shivdasani, C. D. Luu, R. Cicione et al., “Evaluation ofstimulus parameters and electrode geometry for an effectivesuprachoroidal retinal prosthesis,” J. Neural Eng., vol. 7,no. 3, pp. 036008, 2010.

[73] J. Villalobos, P. J. Allen, M. F. McCombe et al., “Developmentof a surgical approach for a wide-view suprachoroidal retinalprosthesis: evaluation of implantation trauma,” Graefe’sArchive Clin. Exp. Ophthalmol., vol. 250, no. 3, pp. 399–407,2011.

[74] J. Villalobos, D. A. X. Nayagam, P. J. Allen et al., “A wide-fieldsuprachoroidal retinal prosthesis is stable and well toleratedfollowing chronic implantation,” Invest. Ophthalmol. Vis. Sci.vol. 54, no. 5, pp. 3751–3762, 2013.

[75] A. Y. Chow, V. Y. Chow, K.H. Packo et al., “The artificialsilicon retina microchip for the treatment of vision lossfrom retinitis pigmentosa,” Arch. Ophthalmol., vol. 122,no. 4, pp. 460–469, 2004.

[76] J.-M. Seo, S. J. Kim, H. Chung et al., “Biocompatibility ofpolyimide microelectrode array for retinal stimulation,”Mater. Sci. Eng. C, vol. 24, no. 1–2, pp. 185–189, 2004.

[77] J. A. Zhou, S. J. Woo, S. I. Park et al., “A suprachoroidal elec-trical retinal stimulator design for long-term animal experi-ments and in vivo assessment of its feasibility andbiocompatibility in rabbits,” J. Biomed. Biotechnol., 547428,2008.

[78] S.W. Lee, J.-M. Seo, S. Ha et al., “Development of micro-electrode arrays for artificial retinal implants using liquidcrystal polymers,” Invest. Ophthalmol. Vis. Sci., vol. 50,no. 12, pp. 5859–5866, 2009.

[79] E. T. Kim, C. Kim, S.W. Lee et al., “Feasibility of microelec-trode array (MEA) based on silicone-polyimide hybrid forretina prosthesis,” Invest. Ophthalmol. Vis. Sci., vol. 50, no. 9,pp. 4337–41, 2009.

[80] M. Schuettler, S. Stiess, B. V. King et al., “Fabrication ofimplantable microelectrode arrays by laser cutting of siliconerubber and platinum foil,” J. Neural Eng., vol. 2, no. 1,pp. S121–8, 2005.

[81] K. Ganesan, A. Stacey, H. Meffin et al., “Diamond penetrat-ing electrode array for epi-retinal prosthesis,” in Proc.Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., pp. 6757–6760,2010.

[82] J. D. Weiland, Liu W, and M. S. Humayun, “Retinal prosthe-sis,” Annu. Rev. Biomed. Eng., vol. 7, pp. 361–401, 2005.

[83] T. Stieglitz, W. Haberer, C. Lau et al., “Development of aninductively coupled epiretinal visual prosthesis,” in Proc.Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., pp. 4178–4181,2004.

[84] I. Bokkon, “Phosphene phenomenon: A new concept,”BioSystems, vol. 92, pp. 168–174, 2008.

[85] J. Walker, “The amateur scientist: about phosphenes: pat-terns that appear when the eyes are closed,” Scient. Am.,vol. 244, pp. 142–152, 1981.



[86] D. Nanduri, I. Fine, A. Horsager et al., “Frequency andamplitude modulation have different effects on the per-cepts elicited by retinal stimulation,” Invest. Ophthalmol.Vis. Sci., vol. 53, no. 1, pp. 205–14, Jan, 2012.

[87] M. S. Humayun, E. de Juan, Jr., J. D. Weiland et al., “Patternelectrical stimulation of the human retina,” Vision Res., vol.39, no. 15, pp. 2569–76, Jul, 1999.

[88] M. J. McMahon, J. D. Dorn, A. K. Ahuja et al., “The Argus IIretinal prosthesis enables blind subjects to localizeobjects,” ARVO Meeting Abstr., vol. 50, no. 5, pp. 4589,2009.

[89] A. K. Ahuja, J. D. Dorn, A. Caspi et al., “The Argus II retinalprosthesis enables blind subjects to identify the directionof motion,” ARVO Meeting Abstr., vol. 50, no. 5, pp. 4590,2009.

[90] A. K. Ahuja, J. D. Dorn, A. Caspi et al., “Subjects implantedwith the ArgusTM II retinal prosthesis are able to improveperformance in a spatial-motor task,” ARVOMeeting Abstr.,vol. 51, no. 5, pp. 4322, 2010.

[91] S. Mohand-Said, A. Caspi, F. Merlini et al., “Comparisonof ETDRS, Landolt C, and grating visual acuity testsbetween sighted volunteers using a pixelized image sim-ulator and blind subjects implanted with the ArgusTm IIretinal prosthesis,” ARVO Meeting Abstr., vol. 52, no. 6,pp. 4931, 2011.

[92] M. Arsiero, L. da Cruz, F. Merlini et al., “Subjects blinded byouter retinal dystrophies are able to recognize shapes usingthe Argus II retinal prosthesis system,” ARVO MeetingAbstr., vol. 52, no. 6, pp. 4951, 2011.

[93] L. daCruz, F. Merlini, M. Arsiero et al., “Subjects blinded byouter retinal dystrophies are able to recognize outlinedshapes using the Argus(R) II retinal prosthesis system: acomparison with the full shapes recognition task,” ARVOMeeting Abstr., vol. 53, no. 6, pp. 5507, 2012.

[94] J. A. Sahel, L. da Cruz, F. Hafezi et al., “Subjects blind fromouter retinal dystrophies are able to consistently read shortsentences using the ArgusTM II retinal prosthesis system,”ARVO Meeting Abstr., vol. 52, no. 6, pp. 3420, 2011.

[95] J. D. Dorn, A. K. Ahuja, M. Arsiero et al., “The ArgusTM IIretinal prosthesis provides complex form vision for a sub-ject blinded by retinitis pigmentosa,” ARVO Meeting Abstr.,vol. 51, no. 5, pp. 3020, 2010.

[96] P. E. Stanga, F. Hafezi, J. A. Sahel et al., “Patients blinded byouter retinal dystrophies are able to perceive color using theArgus™ II retinal prosthesis system,” ARVO Meeting Abstr.,vol. 52, no. 6, pp. 4949, 2011.

[97] P. E. Stanga, J. A. Sahel, L. daCruz et al., “Patients blinded byouter retinal dystrophies are able to perceive simultaneouscolors using the Argus(R) II retinal prosthesis system,” ArvoMeeting Abstr., vol. 53, no. 6, pp. 6952, 2012.

[98] D. Nanduri, J. D. Dorn, M. S. Humayun et al., “Perceptproperties of single electrode stimulation in retinal pros-thesis subjects,” ARVO Meeting Abstr., vol. 52, no. 6,pp. 442, 2011.

[99] G. Richard, M. Keserue, M. Feucht et al., “Visual perceptionafter long-term implantation of a retinal implant,” ARVOMeeting Abstr., vol. 49, no. 5, pp. 1786, 2008.

[100] G. Richard, R. Hornig, M. Keseru et al., “Chronic epiretinalchip implant in blind patients with retinitis pigmentosa:long-term clinical results,” ARVO Meeting Abstr., vol. 48,no. 5, pp. 666, 2007.

[101] M. Keseru, M. Feucht, N. Bornfeld et al., “Acute electricalstimulation of the human retina with an epiretinal elec-trode array,” Acta Ophthalmol, vol. 90, no. 1, pp. e1–8,2012.

[102] R. Hornig, T. Zehnder, M. Velikay-Parel et al., “The IMIretinal implant system,”in Artificial Sight: Basic Research,Biomedical Engineering, and Clinical Advances.Humayun, M., Weiland, J., Chader, G., Greenbaum, E, ed.,pp. 111–128, New York: Springer, 2008.

[103] R. Wilke, V. P. Gabel, H. Sachs et al., “Spatial resolution andperception of patterns mediated by a subretinal16-electrode array in patients blinded by hereditary retinaldystrophies,” Invest. Ophthalmol. Vis. Sci., vol. 52, no. 8,pp. 5995–6003, 2011.

[104] R. Wilke, U. Greppmaier, A. Harscher et al., “Factors affect-ing perceptual thersholds of subretinal electric stimulationin blind volunteers,” ARVO Meeting Abstr., vol. 51, no. 5,pp. 2026, 2010.

[105] W. Mokwa, M. Goertz, C. Koch et al., “Intraocular epireti-nal prosthesis to restore vision in blind humans,” Conf.Proc. IEEE. Eng. Med. Biol. Soc., vol. 2008, pp. 5790–3,2008.

[106] G. Roessler, T. Laube, C. Brockmann et al., “Implantationand explantation of a wireless epiretinal retina implantdevice: observations during the EPIRET3 prospective clin-ical trial,” Invest. Ophthalmol. Vis. Sci., vol. 50, no. 6,pp. 3003–3008, 2009.

[107] H. Kanda, T. Morimoto, T. Fujikado et al.,“Electrophysiological studies of the feasibility ofsuprachoroidal-transretinal stimulation for artificial visionin normal and RCS rats,” Invest. Ophthalmol. Vis. Sci., vol.45, no. 2, pp. 560–566, 2004.

[108] T. Fujikado, T. Morimoto, H. Kanda et al., “Evaluation ofphosphenes elicited by extraocular stimulation in nor-mals and by suprachoroidal-transretinal stimulation inpatients with retinitis pigmentosa,” Graefe’s ArchiveClin. Exp. Ophthalmol., vol. 245, no. 10, pp. 1411–1419,2007.

[109] M.N. Shivdasani, J. B. Fallon, C. D. Luu et al., “Visual cortexresponses to single- and simultaneous multiple-electrodestimulation of the retina: implications for retinal pros-theses,” Invest. Ophthalmol. Vis. Sci., vol. 53, no. 10,pp. 6291–6300, 2012.

[110] J. A. Zhou, S. J. Woo, S. I. Park et al., “A suprachoroidalelectrical retinal stimulator design for long-term animalexperiments and in vivo assessment of its feasibility andbiocompatibility in rabbits,” J. Biomed. Biotechnol.,vol. 2008, pp. 547428, 2008.

[111] Y. Terasawa, K. Osawa, M. Ozawa et al., “Large-surface-areaelectrodes based on bulk micromachining,” ARVO MeetingAbstr., vol. 49, no. 5, pp. 3020, 2008.

[112] Y. Terasawa, H. Tashiro, K. Osawa et al., “Characterizationof electrochemically-treated platinum bulk electrodes,”ARVO Meeting Abstr., vol. 51, no. 5, pp. 3033, 2010.

[113] J. F. Rizzo, 3rd, “Update on retinal prosthetic research: theBoston Retinal Implant Project,” J. Neuro-ophthalmol.Official J. North Am. Neuro-Ophthalmol. Soc., vol. 31,no. 2, pp. 160–168, 2011.

[114] J. F. Rizzo, 3rd, J. Wyatt, J. Loewenstein et al., “Perceptualefficacy of electrical stimulation of human retina with amicroelectrode array during short-term surgical trials,”



Invest. Ophthalmol. Vis. Sci., vol. 44, no. 12, pp. 5362–5369,2003.

[115] J. F. Rizzo 3rd, J. Wyatt, J. Loewenstein et al., “Methodsand perceptual thresholds for short-term electrical stim-ulation of human retina with microelectrode arrays,”

Invest. Ophthalmol. Vis. Sci., vol. 44, no. 12, pp. 5355–5361, 2003.

[116] J. F. Rizzo, J. Chen, D. B. Shire et al., “Overview of progresson the 256+ channel Boston retinal prosthesis,” ARVOMeeting Abstr., vol. 53, no. 6, pp. 1313, 2012.


/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages false /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 400 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages false /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice

Documents

The Bionic Eye: a review of multielectrode arrays › ... › 2015_12_bionic-review.pdfBionic Eye devices still in the preclinical research phase, this chapter will summarize the advances