No abstract

No abstract

In order to better understand the reasons for success or failure of a cochlear implant system for various patients, it appears necessary to analyze the patients' basic psychophysical capacities in relation to speech perception. Five patients with intracochlear multichannel Ineraid implants were studied in terms of their performance on temporal analysis in relation to their perception of consonants. For temporal analysis we measured the detection of a silent gap in noise and of an interval between two clicks. For consonant perception we established a confusion matrix based on 12 consonants presented in a vowel-consonant-vowel context using the vowel /a/. The results showed a correlation between temporal resolution for two successive clicks at the most basal cochlear electrode used, and the perception of place of articulation of consonants. This finding indicates that delivering fine temporal coding can be crucial for the success of an implant and that for a given subject, optimal capacity for temporal resolution may serve as a criterion for choosing a basal electrode.

Two identical multichannel intracochlear prostheses were implanted in the same patient. The first prosthesis, implanted in the congenitally-deaf right ear, elicited clear sound perception but no speech recognition. After 2 years, a second prosthesis, implanted in the acquired-deaf left ear, enabled the patient to understand speech without lip-reading. Brainstem and middle-latency evoked potentials were similar with electrical stimulation of both ears and resembled those evoked by acoustic stimuli in subjects with normal hearing. Cortical electric and magnetic responses differed for right- and left-sided electrical stimulation suggesting that stimulation of the congenitally-deaf ear elicited an abnormal activation of the auditory cortex. These results suggest that only cortical responses were affected by the different histories of deafness of the ears.

Electrically evoked short latency vestibular potentials were recorded in 9 patients during vestibular neurectomy. Patients were operated on because of intractable Meniere's disease. The VIIIth cranial nerve was exposed through a limited retrosigmoidal approach; the vestibular nerve was contacted in the cerebello-pontine angle with a bipolar platinum-iridium electrode and stimulated with biphasic current pulses (100 microseconds/phase, 0.75-1 mA p-p, 20/s). The responses were recorded over 12.8 ms between a forehead and an ipsilateral ear lobe electrode. Each recording consisted of 2 x 1,000 averaged responses. A systematically reproducible vertex-negative potential occurring at a latency of approximately 2 ms and having an amplitude of approximately 0.5 microV was recorded in all patients. This vertex-negative potential disappeared after selective vestibular neurectomy proximal to the stimulation site. Simultaneous continuous acoustic masking did not affect the response and no facial nerve response was observed on the facial nerve monitoring. These features strongly suggest that the characteristic vertex-negative potential constitutes a specifically evoked response of the vestibular system. Electrophysiological monitoring of the sectioning of the vestibular nerve during operation is one possible clinical application of intraoperative recording of electrically evoked vestibular potentials.