THE GOAL OF THIS RESEARCH IS TO DISCOVER A MORE EFFECTIVE SPEECH CODING TECHNIQUE FOR SINGLE-ELECTRODE COCHLEAR IMPLANT RECIPIENTS.

THE GOAL OF THIS RESEARCH IS TO DISCOVER A MORE EFFECTIVE SPEECH CODING TECHNIQUE FOR SINGLE-ELECTRODE COCHLEAR IMPLANT RECIPIENTS. SUBJECTS HAVE BEEN IDENTIFIED AND ARE PRESENTLY WEARING 3M DEVICES. A BATTERY OPERATED PULSE GENERATOR HAS BEEN DEVELOPED AND TESTED AND WILL BE USED TO ACTIVATE THE ELECTRODE. A STANDARD TEST TAPE IS BEING PREPARED WITH SEVERAL MAC BATTERY SUBTESTS, A DIGIT RECOGNITION TEST AND A STRESS AND WORD RECOGNITION TEST. THE TAPE WILL TAKE ABOUT TWENTY MINUTES TO RUN. THE METHODOLOGY WILL CONSIST OF PROCESSING THE STANDARD TAPE OFF-LINE USING A VAX 11/750 AND GENERATING TEST TAPES WITH VARIOUS SCHEMES USING PULSATILE CODES. THE TEST TAPES WILL BE PLAYED FOR SUBJECTS USING A BATTERY OPERATED RECORDER ELECTRICALLY COUPLED TO THE STIMULATING UNIT AND COUPLED THROUGH A STANDARD COIL TO THE ELECTRODE. THE MAJOR ASSUMPTION BEHIND THE CODING TECHNIQUES IS THAT THE ELECTRODE MUST BE DRIVEN PITCH SYNCHRONOUSLY FOR OPTIMAL PERFORMANCE. THE HYPOTHESIS TO BE TESTED IS THAT VOICE PITCH WILL BE CONVEYED THROUGH THE ELECTRODE BY THE REPETITION RATE OF CLUSTERS OF PULSES AND TIMBRE CAN BE VARIED BY MODIFYING THE CHARACTERISTICS AND NUMBER OF PULSES WITHIN THE CLUSTERS.