The object of this proposal is to develop and implement the concept of a computer workstation for neurology. The system will have all the present functionality of the EEG machine (montage selection, control of raise and fall times, 60 Hz interference, gain adjustments, paper speed control), but will be enhanced with characteristics intrinsic to the computer. We want to stress the possibilities of the instrument for the clinical environment: to make precise measurements in the EEG (microvolt accuracy and millisecond resolution), receive and send EEG records through data networks (real-time teleconferencing), integrate the EEG signal with the patient clinical history in the data base, where images can also be stored (CAT scanners), upgrade the system with existing automated signal detection modules (epilepsy, sleep, etc.), create reports and interface with the hospital information system. What makes this instrument possible are the advances in computer hardware, the powerful floating point digital signal processors, sophisticated graphics processors, and new languages and paradigms such as SMALLTALK and HYPERBASES. We plan to develop a multiprocessor system based on a workstation and a special purpose frontend processor. This processor will be composed of a microcontroller for data acquisition and flow control, interfaced with a floating point DSP processor for fast arithmetic operation. Algorithms to implement the functions discussed above will be tested using simulators, and then translated into assembly language.
