N.R. Srivastava

Estimating Phosphene Maps for Psychophysical Experiments used in Testing a Cortical Visual Prosthesis Device

Visual prosthesis devices are being developed to restore vision for those with blindness. Researchers working in the field of visual prosthesis are taking different approaches to develop a practical device. Some are targeting the retina for stimulation, whereas at least one group is targeting the optical nerve, and our laboratory is developing a system for the visual cortex. To estimate the kind of response they might expect from a typical user, researchers are conducting psychophysical experiments on normally-sighted persons. The device being developed in our laboratory is a first generation visual prosthesis system, designed to test the limits of artificial visual pattern recognition. Targeting the visual cortex area with our first generation device has limitations including limitations in lateral cortical surface area for electrode implantation, surgical difficulties and the lack of understanding as to how to use an artificial interface for communication with the visual cortex. Here, we discuss the uncertainties related to visotopic mapping of the lateral surface of the occipital lobe in humans.

Some Solutions to Technical Hurdles for Developing a Practical Intracortical Visual Prosthesis Device

The goal of cortical neuroprosthesis researchers of last four decades is to develop a practical intracortical visual prosthesis device. Although the concept of such a system seems straightforward, details of its configuration remain undefined. Knowledge of how the human visual system will respond to artificially-induced visually-based input is sparse. Combined with technological limitations, these have hindered the progress in developing a practical intracortical visual prosthesis device. The long-term objective of this research is to develop a continuously wearable intracortical visual prosthesis device. Earlier studies have used relatively small numbers of cortical electrodes, and these have been insufficient to generate an integrated visual perception. Surgical difficulties also complicate problem. A prototype visual prosthesis system needs to be adaptable to varying stimulation, image processing, and user interface needs. It also has the obvious requirement portability, implying extremely low power consumption and low weight so that the system can be used outside the confinement of a lab. We feel that available technology has sufficiently advanced to develop a first-generation intracortical visual prosthesis device. In this paper we propose some solutions to the challenges for developing this visual prosthesis device using existing technologies

A proposed intracortical visual prosthesis image processing system

It has been a goal of neuroprosthesis researchers to develop a system, which could provide artificial vision to a large population of individuals with blindness. It has been demonstrated by earlier researches that stimulating the visual cortex area electrically can evoke spatial visual percepts, i.e. phosphenes. The goal of visual cortex prosthesis is to stimulate the visual cortex area and generate a visual perception in real time to restore vision. Even though the normal working of the visual system is not been completely understood, the existing knowledge has inspired research groups to develop strategies to develop visual cortex prosthesis which can help blind patients in their daily activities. A major limitation in this work is the development of an image processing system for converting an electronic image, as captured by a camera, into a real-time data stream for stimulation of the implanted electrodes. This paper proposes a system, which will capture the image using a camera and use a dedicated hardware real time image processor to deliver electrical pulses to intracortical electrodes. This system has to be flexible enough to adapt to individual patients and to various strategies of image reconstruction. Here we consider a preliminary architecture for this system

A laboratory testing and driving system for AIROF microelectrodes

The charge-injection currents of AIROF (activated iridium oxide film) microelectrodes, which are subjected to charge-balanced biphasic pulsing or monophasic current pulsing, have to be limited such that the anodic and cathodic voltage excursions are kept within safe limits of operation. In earlier studies it has been shown that when using anodic bias asymmetry in the magnitude of the balanced biphasic waveform can be used to increase the charge injection capacity of AIROF electrodes. We present the design of a single-channel testing and driving system for laboratory testing and driving of AIROF microelectrodes within safe charge-injection limits.

Simulations of Cortical Prosthetic Vision

Cortical stimulation for restoring vision presents researchers with many challenges and questions. The extent of the human visual cortex varies up to 50% from one individual to another, cortical folding and sulci limit the area of implantation, and surgical difficulties make it difficult to implant electrodes to produce phosphenes in the whole visual space. Researchers are faced with question such as: which electrodes to use – surface electrodes that are easy to implant or intracortical fine-metal electrodes that have lower current requirements and have five times better resolution? How many phosphenes will be enough to give limited, but useful vision? How will cortical physiology affect phosphene maps? Will percepts be distinct dots or complex in nature? What will be the long term response to stimulation? Will the brain adapt to seeing through dotted images? Some of these questions can be answered by conducting human psychophysical tests.

Test Setup for Supporting Human Implantation of Intracortical Visual Prosthesis Device

Earlier experiments in the field of cortical visual prosthesis have shown the possibility of generation of phosphenes. Experiments have been performed with different types of electrodes, researchers have found the stimulation parameters required to elicit a phosphene and they have shown the possibility of targeting different areas of visual cortex to elicit phosphenes. Experiments have not been conducted in which an image was captured and processed in real time, and an array of electrodes stimulated, corresponding to the image, to generate a sense of vision. Development of a prosthetic device faces the crucial question whether a practical number of cortical stimulating electrodes can provide a useful sense of vision. We aim to answer this question by designing a wearable cortical prosthesis device and testing it on blind human volunteers. Before we implant this device in human volunteers, we want to estimate the performance we might expect from a human implantation. We are planning to conduct psychophysical tests on normally-sighted humans and stimulation tests on non-human primates. Results from these experiments will help us understand what we should expect from implantation in a human volunteer.

Radix 4 SRT Division with Quotient Prediction and Operand Scaling

SRT division is an efficient method for implementing high radix division circuits. However, as the radix increases the size of a quotient digit selection table increases exponentially. To overcome the limitations of quotient prediction, a method in which a quotient digit is speculated has been proposed. The speculated quotient digit is utilised to update the possible partial remainders while the speculated quotient is corrected. In this paper, instead of using a huge quotient selection table an estimation and correction scheme is used for prediction of quotient digit. The prediction is done in parallel with the calculation of the partial remainder for the quotient predicted earlier thus improving the latency. In addition, since this method tends to consume less area as the radix increases compared to previous methods, it has the ability to improve higher radix implementations for SRT division

Testing and monitoring system for a first generation intracortical visual prosthesis device using Java, Labview and VGA / NTSC output

Hardware designed for a specific application presents a unique problem of testing it for accuracy and behavior. The visual prosthesis device designed in the Laboratory of Neural Prosthetic Research at IIT, (Chicago) uses an FPGA to capture images from a camera, performs image processing, and generates an instruction set which is sent on a radio frequency magnetic link to an implanted module in the visual cortex producing electrical currents in a group of intracortical electrodes. The design of this visual prosthesis device presents the problem of testing it both quantitatively and qualitatively, prior to implantation in a human. We can test the individual components of a visual prosthesis device using software simulations but the problem is to test it end to end. The problem is further complicated by the fact that the system is processing 50 frames/ second and generating data at a rate of 1.25 Mbits/sec. To evaluate all possible inputs and do an exhaustive test becomes a daunting task. In this paper we describe a method developed in our laboratory to easily and effectively test, and monitor, our system. Along with the test modules for our dedicated hardware, we have developed a secondary module as a requirement for visually monitoring the behavior of the system during non-human / human primate experiments, providing a means for estimating the cortical response and resultant visual perception. This system will also be used for human psychophysical experiments.