The perceptual dimensions of single-electrode and nonsimultaneous dual-electrode stimuli in cochlear implantees.

Abstract

The perceptual dimensions evoked by dual-electrode stimulation were investigated in four cochlear implantees. The dual-electrode stimulation consisted of biphasic current pulse trains, such that two intracochlear electrodes each received one pulse in a 250-Hz cycle. The experiments tested the hypothesis that perceptual qualities would be altered when there was an increased likelihood of interactions occurring between the two electrode places. The parameters of the stimulation which were manipulated to test this hypothesis were the time delay between the pulses on the two electrodes (2 ms and a value between 0.62 and 0.92 ms), the distance between the component electrode rings of each bipolar pair (3.75 mm and either 2.25 or 1.5 mm), and the distance between the two bipolar pairs in the dual-electrode stimuli (from 0.75 to 12.0 mm). Five set of 15 loudness balanced stimuli were created, each set having different stimulation parameters. These stimuli consisted of five single-electrode stimuli (a 250-Hz pulse train on each of five electrodes) and the ten dual-electrode stimuli formed by the combinations of those five electrodes. Two perceptual dissimilarity matrices were obtained for each subject and each set, and were analyzed using repeated nonmetric multidimensional scaling techniques. The resultant "stimulus spaces" were then examined to see how many perceptual dimensions the stimuli were best described by, and to what perceptual or stimulus qualities the dimensions might correspond. The results showed that the percept evoked by dual-electrode stimulation contained two main dimensions. Increasing the width of the current path to create substantially overlapping stimulation areas, or altering the temporal delay between the two electrodes, had very little effect on this percept. The position of the stimuli in the two-dimensional space was related to the distances of the two component electrodes along the electrode array. These results are relevant to speech processing strategies and electrode design for cochlear implants, as they imply that the creation of discrete, nonoverlapping areas of stimulation may not cause the percepts to be more distinct, and therefore may not necessarily lead to better speech perception.