MEMS may help the blind to see
12/01/2002
A technical team from five national labs, a private company, and two universities have set out to create a MEMS to help some who are blind to see. The concept is to place a MEMS on the retina — within the eye's vitreous humor — so its electrodes replace damaged rods and cones, directly stimulating neural connections to the brain (Fig. 1). In diseases such as macular degeneration and retinitis pigmentosa, inputs from rods and cones cease, but 70–90% of related nerve structures remain intact.
This project has been funded by a $9 million, three-year grant from the Department of Energy's Office of Biological and Environmental Research. Sandia Lab project leader Kurt Wessendorf, says, "The aim is to bring a blind person to the point where he or she can read, move around objects, and do basic chores. They won't be able to drive cars because instead of millions of pixels, they'll see approximately a thousand. Images will come in slowly and appear yellow. But people who are blind will see."
The overall plan uses a tiny camera and an RF transmitter placed in eyeglasses to transmit information and power to the MEMS placed within the eye (Fig. 1). Project director, Dean Cole of the Office of Biological and Environmental Research, says, "We felt that blindness is a devastating problem and that the modern conjunction of materials science with micro- and nano- technologies in our multidisciplinary national labs offers possibilities for advances, where before people had hit brick walls."
"Compared to the elegance of the original biological design, what we're doing is extremely crude," says Wessendorf. The size of cones and rods, as well as nerve connections, are in the micron range — a difficult but doable realm for scientists used to working with MEMS. Sandia manager, Mike Daily, adds, "We'll use a crude, shotgun approach that fires groups of nerves. In the long run, of course, we'd like to stimulate each individual nerve."
null
The project started with work by physician-researcher Mark Humayun at Johns Hopkins University. When Humayun began the Intraocular Retinal Prosthesis Group at Doheny Retina Institute at the University of Southern California (USC), the project moved with him. Teaming with Eli Greenbaum at Oak Ridge, they visited national labs and ultimately arranged to have each lab work on a different aspect of the electrode array-retina interface.
Goals of the project increase from 10-by-10 electrode arrays for fiscal year 2002 to 33-by-33 arrays for 2004:
- Oak Ridge National Lab will manage the multilaboratory effort and test the various components developed by other labs.
- Argonne National Lab will investigate the viability of diamond-based electrode arrays and biocompatible coatings.
- Lawrence Livermore National Lab will experiment with rubberized electrode arrays.
- Los Alamos National Laboratory will model and simulate neural paths of and from the retina to the brain.
- USC personnel will implant devices and test their medical effectiveness.
- Second Sight, Santa Clarita, CA, will commercially produce the finished system.
- North Carolina State University leads the development of the in situ medical electronics.
Humayun says, "There is a considerable amount of advanced technology literally on the shelf or already being used for defense purposes that we can use."
Daily notes that integrating microdevices into the human eye is extremely challenging because of the need for high-reliability in a saline environment. BioMEMs interfaces and biocompatibility issues drive much of the effort, particularly in the packaging of the microsystem (i.e., sealing and securing a microdevice in place).
The rods and cones of the retina lie beneath nerves, not above them, which makes it slightly easier to connect directly to nerves. "The tissue housing the nerves is relatively clear. We're investigating which electronic waveforms will best stimulate these nerves," Wessendorf says. "If we excite a nerve with electrons, we don't know exactly how that compares to the electrochemical response of light on a healthy retina."
To address the retina's inability to handle pressure, the project is favoring spring-loaded electrodes that ensure good electrode contact with minimal force. The researchers are also looking for solutions to "protein fouling" that interrupts delicate interfaces intended to transmit electrical impulses, biocompatibility, and long-term reliability.
 Figure 2. A prototype of a MEMS-based array that eventually may be inserted onto the retina of a blindperson. (Photo by Randy Montoya) |
null
Over a five-year period, the project will begin with goggles and end with corneal implants for five patients. After that, the FDA will determine whether they want a 100-patient, long-term study; DOE will leave the project in the hands of industry.
"Million dollar man" emerging in Flanders
Europe's Flemish research organization IMEC, Leuven, Belgium, kicked off its 2002 Annual Research Review Meeting (ARRM) — billed as "Seeds for Tomorrow's World" — with the launching of two programs that provide an intriguing view of future wireless networks and the quality of human life.
The ultimate goal is to produce new types of networks and a series of noninvasive sensors (e.g., distributed MEMS and bio-sensing transducers) that add up to "ambient intelligence" (AI). With AI, computing would be ubiquitous in our lives — seeing, listening, and feeling — and the associated wireless IT proactive but hidden (i.e., noconventional keyboard or menus).
It almost seems as if the advanced technology of the 70s television show, The Six Million Dollar Man, is emerging in Belgium.
The first applications envisioned for AI include support for athletes, diagnostics, and repair in health care, personal communications, and entertainment. In athletics, for example, realization of IMEC's goals behind the new programs could change today's common use of a wired heart rate monitor to a body area network (BAN).
In BAN, sensors for heart function, blood pressure and composition, leg and foot position, and force communicate wirelessly to a wearable digital assistant (WDA) linked to coaches and a medical team. It is not difficult to envision extending this technology to hazardous industrial and military situations.
Rudy Lauwereins, IMEC's VP of integrated information and communications systems, says, "The world is evolving into ubiquitous smart environments, where AI will allow electronic access to information and services any time and any place. But making these smart environments a reality requires a personal wireless network, hundreds of electronic devices residing in the background of our daily lives, and devices that are sensitive and adaptive to people."
Lauwereins adds that many of the services envisioned for the future demand high bandwidth and very low power 100µW operation without batteries. Power requirements could be met with an electrostatic generator or polymer based photovoltatics, both being developed at IMEC.
The specific IMEC programs are M4 and Human++. The former targets developing a multimedia multimode terminal, the first step in evolving to a WDA that communicates via wireless distributed transmissions. IMEC engineers envision the WDA fabricated in SoC and SiP technology and capable of IF inputs that see, hear, and feel; IF outputs that speak, show, and stimulate; and RF communications to satellites and various base stations, including GSM, GPRS, UNTS, and WLAN. One WDA/person would provide GPS computing, global connectivity, biometrics input, health monitoring, and ambient control, and many infotainment options. The latter targets a prototype BAN able to network more than 100 sensors/person. The sensors will be based on a generic platform that includes power supply, a wireless transceiver, network control, and packaging.
As revealed at ARRM 2002, IMEC is already making progress with specific sensors. For example, it has a Ta2O3 gate-dielectric pH sensor that is applicable to glucose measurements. IMEC engineers are also working on a magnetic detection genechip that can detect biomolecules related to genetic diseases (e.g., cystic fibrosis specific mutations on DNA and RNA).
Other work at IMEC is with specific MEMS that are capable of microcalorimetry for microphysiometry; a poly-SiGe based bolometer for thermal IR sensing; CMOS integrated FETs and amperometric sensors that provide blood-gas data and a silicon-based impedance sensor that provides affinity based immuno- and geno-sensing. —P.B.
 A prototype design for a Ta2O3 gate-dielectric pH sensor applicable to glucose measurements. (Source: IMEC) |
null
HP unveils molecular memory developments
Hewlett-Packard Co. (HP) has unveiled what it calls "breakthroughs" in molecular electronics by its scientists in HP Labs, the central research facility. R. Stanley Williams, director of quantum science research at HP Labs, said his group had done three things.
The group "created the highest density electronically addressable memory reported to date." The laboratory demonstration circuit, a 64-bit memory using molecular switches as active devices, fits inside a square micron. The device's bit density is more than 10 times greater than silicon memory chips.
The group also combined, for the first time, both memory and logic using rewritable, nonvolatile molecular-switch devices.
Lastly, they made circuits with an advanced manufacturing system called nano-imprint lithography, a printing method that stamps out an entire wafer of circuits quickly and inexpensively from a master. The circuits were made using HP's patented crossbar architecture, incorporating molecular switches.
First, researchers made a master mold of eight parallel lines, each only 40nm wide. Then, in a three-step process, researchers: pressed the mold into a polymer layer on a silicon wafer to make eight parallel "east-west" trenches, which they then filled with platinum metal to form wires; deposited a single layer of electronically switchable molecules on the surface; and repeated the first step, after rotating the mold 90° to make another eight wires, running "north-south," on top of the molecular layer.
At each of the 64 points where top and bottom wires crossed, the roughly 1000 molecules sandwiched between them became a bit of memory. A bit can be written by applying a voltage pulse to set the molecules' electrical resistance and read by measuring their resistance at a lower voltage. The memories proved to be both rewritable and nonvolatile. Researchers also put logic in the same circuit by configuring molecular-switch junctions to make a demultiplexer.