Introduction A good example is the

Introduction A good example is the epiretinal prosthesis, a device that is able to electrically stimulate surviving retinal U0126 of patients suffering from diseases such as retinitis pigmentosa (RP) and age-related macular degeneration (AMD). A large percentage of patients with AMD retain good peripheral vision. In contrast, many patients with advanced RP retain their central vision. Thus, implantation of a retinal prosthesis would be justified for such patients only if it provided a substantial improvement in visual acuity; otherwise they would not benefit from it. Studies show that a retinal prosthesis must have about 1000pixels/electrodes to restore functions such as face recognition, reading and unaided mobility [1]. Unfortunately, most of the epiretinal prostheses currently under development comprise arrays of as few as 60electrodes, each with diameters of 100micrometers or more [2–5]. These implants provide very limited vision, allowing patients to only see spots of light and high-contrast edges. The design of high density microelectrode arrays presents several engineering and biological challenges. For instance, having 1000electrodes confined in an area of 30mm2 (area of the macula) leads to two major issues: (1) at least 10 conducting lines would need to pass between electrodes, which would produce large capacitive coupling and be very difficult to fabricate and (2) the center-to-center distance between electrodes cannot exceed 150μm, thus the electrode diameter has to be made small enough in order to avoid cross-talk, and most electrodes with a diameter that small (usually <100μm) cannot deliver enough charge to exceed the stimulation threshold of nerve cells without conflicting with the electrochemical safety requirements. The first challenging issue, which is the routing of signals from the current sources to the stimulating electrodes, has been addressed by patterning the conducting lines on different planes and using vias to connect the planes to each other [6,7]. However this increases the thickness of the device (leading to large stiffness) and greatly complicates the fabrication process flow. Some researchers have overcome the interconnect limitation by designing novel MEA systems and making use of multi-microchip architectures, which consist of multiple chips, each comprising several electrodes and a control circuit [8,9]. This approach offers the possibility of connecting several microchips via a bus system, which enables a decrease in the number of connection lines. However this does not result in a significant increase in electrode density as some space has to be dedicated to each control circuit and to the spacing between microchips. While fabricating small electrodes is technologically possible, using them to safely and efficiently stimulate neurons is the second challenging issue. Nevertheless, a great number of metals and metal alloys have been fabricated and used as microelectrodes for neural stimulation. Iridium oxide (IrOx) is considered to be one of the best neural electrode materials because of its very high charge injection capacity and its reversible faradic reaction. However, because IrOx delaminates under high current pulsing, it leaves traces in the tissue which would lead to harmful effects in the long run [10].
Materials and methods
Results and discussion
Conclusion There is an increasing need for high electrode density implants. While the electrode size continues to scale down, the stimulating neural microelectrode array has to remain efficient and safe. Furthermore, while the number of electrodes increases, new ways of routing the electrodes to current sources has to be found. This is not an easy task and arising problems can be solved or at least alleviated through intelligent design of the microelectrode chip. In this study, a new class of microelectrode chip was designed, fabricated and tested. Carbon Nanotubes have been used as the electrode material, with the objective of satisfying the size, efficiency and safety requirements; and TSV poly-Si interconnects were employed to solve the problem of routing for large microelectrode arrays. We are the first to combine these two technologies into a single chip. With the use of CNTs, TSV and Flip-Chip technology, the next generation of neural implants could easily incorporate at least 1000microelectrodes.