by John Williamson
Contributing Editor, MEMS Investor Journal
Retinal disease, or degeneration of the retinal cells, is exemplified by genetically caused retinitis pigmentosa (RP) and age-related macular degeneration (MD). These are leading causes of blindness affecting an estimated 1-3 million and 25-30 million people worldwide, respectively. As MD progresses, the tissue of the macula is destroyed leading to a complete loss of vision in the center. In the case of RP, early symptoms include night blindness, loss of contrast and steadily diminishing peripheral vision. Both can be traced to degenerating light-sensitive rods and cones in the retina.
According to Professor Wilfried Mokwa of RWTH Aachen University, some 30% of retinal ganglion cells of RP patients are still working even after several years of blindness. The function of these cells is to transmit image-forming and non-image forming visual information from the retina to several regions of the brain.
“Electrical stimulation of the cells has been shown to cause a visual sensation,” says Professor Mokwa. “MEMS technology is being developed and employed to create stimulation electrodes that can be placed (1) in the space between sclera (the white of the eye) and the retinal nerve cells to replace degenerated retinal rods and cones [subretinal], (2) on the inner retina [epiretinal] or (3) onto the outer surface of the sclera where they can also penetrate partly or completely through the sclera [transscleral]. The transscleral technology is similar to the other two.”
Subretinal stimulation techniques
“Several research organizations are working on developing subretinal stimulation techniques,” Professor Mokwa continues. “An example is Stanford University’s approach to replace degenerated rods and cones with a high-density MPDAs, or microphotodiode arrays.”
According to Dr. Rostam Dinyari of Stanford’s Electrical Engineering Department, the work is an outgrowth of a project to construct curved monolithic silicon structures that can be produced using standard silicon processing prior to curving. “Originally targeting photographic applications, the development is now being studied as a solution to subretinal stimulation,” he says.
As explained by Dr. Dinyari, retinal implants with 230 μm, 115 μm and 57.5 μm pixels in a 2 mm by 2 mm array were fabricated. These contain 64, 256 and 1024 pixels respectively. The silicon flexures can be deformed allowing the implant to conform to the natural curvature of the eye.
“Each pixel contains two electrodes -- an active electrode at the center and a return electrode in the perimeter,” Dr. Dinyari says. “When the photovoltaic pixel is illuminated, a voltage difference is generated between the two electrodes. This causes a current to flow between them passing through the retinal tissue. If the current is sufficiently high (larger than a stimulation threshold), it activates the retinal neurons. These stimulations are sent via the optic nerve to the brain which perceives them as visual information.”
Single and triple-diode pixels were fabricated and tested. In single-diode pixels, the area of the pixel contains one whole photovoltaic cell. In triple-diode pixels (see figure below) the area of the pixel is divided into three equal photovoltaic sub-pixels that are connected together in series using a MEMS structure.
Dr. Dinyari says that for single-diode and triple-diode pixels of the same size, the open-circuit voltage of a triple-diode pixel is three times that of the single-diode pixel (its short-circuit-current is one third that of a single-diode pixel). “The electrochemical impedance at the electrode tissue interface is a nonlinear function of the voltage,” he says. “Due to this nonlinear dependence, the charge injected by a triple-diode pixel is 5 times that of a single-diode pixel of the same area. This enables stimulating the neurons with smaller pixel sizes and therefore making higher resolution implants.”
Initial experiments suggest that implants 115 μm triple-diode pixels (256 x 256 pixels) can inject currents beyond stimulation threshold at safe optical intensities. Experiments on animal models are in progress.
Another figure below shows how these implants can be introduced into the subretinal space by the classical transvitreal approach to the back of the eye and placed under the retina through a small incision (top), or through a scleral flap under the retina and in the subretinal space to the regiomacularis guided by a custom-made plastic foil (bottom).
Two techniques for introducing the photodiode array. Courtesy: Ophthalmologica.
MEMS and epiretinal stimulation developments
RWTH Aachen University’s Professor Mokwa says that, like Stanford’s subretinal system, epiretinal stimulation systems also require an extraocular and an intraocular component. But, until now, epiretinal stimulation has required surgery to connect the externally generated information to the implanted device. This has changed with the development of the EPIRET3 system at Aachen University’s Department of Ophthalmology that positions the implant completely inside the eye. It does not require a cable connection through the sclera and functions as a wireless intraocular retinal implant.
In brief, an external transmitter coil sends data and energy from the camera to a receiver module coil encapsulated within a silicon material used in artificial lens fabrication. This assembly is positioned as an artificial lens mounted into the eye following surgical removal of the natural lens. It is connected via flexible micro-cable to the stimulator unit tacked to the retinal surface.
The intraocular device consists of a receiver coil, the electronics, and 25 three-dimensional stimulation electrodes that are 25 μm high and 100 μm in diameter as shown in the figure below. Stimulation data and control signals are sent via the wireless RF link to the receiver coil mounted in the artificial lens, which in turn sends them to a receiver microchip. Next the stimulator chip receives these pulses, generates the stimulation pulses and activates selected three-dimensional stimulation electrodes by means of bipolar current. A 10 μm thin polyimide foil 40 mm by 3 mm in size serves as the flexible micro cable delivering signals to the stimulator unit tacked to the retinal surface.
The EPIRET 3 receiver coil, metal wiring and the electrodes are formed by micro electroplating gold. Electrodes are covered with a thin film of iridium oxide. The entire implant is coated with parylene and the active surface of the stimulation electrodes is opened by plasma etching.
Professor Mokwa comments that data gathered from the EPIRET 3 trial on legally blind RP volunteers will lead to the development of a second generation wireless implant system. According to him, this implant will have a considerably higher number of electrodes and more signal processing power to provide useful artificial vision for blind people.
MEMS tool for intraocular surgery and drug delivery
At the Institute of Robotics and Intelligent Systems (IRIS) within the Swiss Federal Institute of Technology in Zurich (ETHZ), a sub-millimeter sized MEMS robot is under development to navigate the vitreous humor -- the clear gel-like substance in the eye -- to deliver drugs or perform retinal surgical procedures.
According to project leader Professor Bradley Nelson, the project combines micro and nano technology with macro-scale medical robotics. “We are working in two areas,” Professor Nelson says, “namely building the micro-robots by using MEMS, NEMS and robotic micro-assembly technology and developing medical imaging and magnetic steering to apply the device for in vivo applications.” The goal is to make the robot small enough to fit in a 23-gauge needle. At this size it can be injected into the eye using only a topical anesthetic and without the need of a suture.
According to Professor Nelson, a major challenge is achieving precise control of the fields and currents through the steering magnets. We are looking at two alternatives for generating external fields, namely electro-magnetic coils or permanent magnets. Once this is solved we will apply the robot to targeted delivery, treating diseases such as age-related macular degeneration or retinal vein occlusions. “In these instances,” he says, “our objective is to deliver drugs to specific locations on the retina.”
Currently, four different types of robots are being assembled from a combination of planar parts that are manufactured using a nickel electroplating process and bonded with UV activated glue. “The significant advantage of the hybrid design is that the individual parts of the assembly can be produced with standard MEMS manufacturing processes to create planar geometries,” Professor Nelson says. “This way, different sub-systems of the robot can be manufactured using the most suitable process for the purpose being targeted.” He says that a drug-dispensing robot can remain in the eye for several months and that the coating on the robot establishes its function. There are no internal components.
The “winged-ellipsoid” shape has an axis of symmetry along the long axis of the ellipsoid, according to Professor Nelson. An external magnetic field (see figure below) acts to align and pull the robot along this axis (i.e. magnetic torque and force) due to the shape anisotropy effect, much like a needle always becoming magnetized along its long axis. The winged shape also acts to reduce the sideway drift of the micro-robot by increasing the fluid drag along the axes perpendicular to the long axis.
The project is in the early stages of development with retinal surgery experiments on the eyes of dead animals. It will proceed to trials on live animals before human applications.
This article is a part of MEMS Investor Journal's ongoing market research project in the area of MEMS and non-MEMS implantable technologies for biomedical applications. If you would like to receive our comprehensive market research report on this topic, please contact John Williamson at firstname.lastname@example.org for more information about rates and report contents.
Copyright 2011 MEMS Investor Journal, Inc.