Using guinea pig models to examine the relation between psychophysical and neural responses to electrical stimulation of cochlear implants

Animal models are used to study the physiology underlying cochlear implant function. In these models, various treatment protocols are used in attempts to represent the conditions of deafness in implanted patients. These protocols are known to affect psychophysical detection threshold levels and are likely to affect physiological data as well. This variation in methodology might affect the conclusions drawn from the neural data about the mechanisms underlying auditory perception. The purpose of this dissertation was to determine how the neural data were affected by treatment protocol. Data were collected from groups of animals using various treatment procedures, which included acute, short-term, and long-term deafening groups. One group underwent psychophysical testing in addition to physiological recording. Multiunit activity was recorded along the tonotopic axis of the inferior colliculus (IC). Strength-duration functions and spatial tuning curves (STC) were determined for bipolar (BP) and monopolar (MP) stimulation. Threshold levels were strongly affected by the treatment protocol. However, the slopes of the strength-duration functions and the STC widths did not differ systematically across treatment groups. Neural thresholds collected from the short-term deafened groups were generally higher and more variable than those collected for the long-term deafened groups. Neural thresholds for the psychophysically trained animals were lowest. Thresholds for BP stimulation showed more variance across all groups than thresholds for MP stimulation. Electrode configuration affected the psychophysical strength-duration-function slopes, but did not affect the neural strength-duration function slopes. Implant insertion methods had large and unexpected effects on psychophysical and neural thresholds. Of the groups tested, the psychophysically-tested group seemed to be best for studying mechanisms of psychophysical stimulus detection because the neural thresholds were closest to the psychophysical thresholds obtained from these same animals. In addition, because of the wide variability of thresholds observed in the acutely-tested groups, these groups might be good animal models for study of functionally relevant anatomical and physiological changes in the cochlea following deafening and/or implantation. Data from this study and future studies will be helpful in interpreting differences in thresholds observed across patients and across stimulation sites, as well as threshold changes over time.