Correlation between rCBF and speech perception in cochlear implant users

Abstract

Objective: Although cochlear implants (CIs) have provided the opportunity for bilaterally deaf individuals to recover their hearing abilities, the speech perception performances of the CI users varies considerably. To elucidate the cortical mechanisms of processing speech signals coded by CIs, we evaluated the correlation between the brain activity during speech activation and speech perception in CI users by PET. Methods: Fourteen postlingually deaf CI users were examined. CI used in the patients was a 22-channel system and its speech-coding strategy was the Nucleus spectral peak (SPEAK) strategy. To evaluate the speech perception performances, we examined vowel perception, consonant perception and speech tracking performances in the Japanese language. Regional cerebral blood flow (rCBF) was measured during no sound stimulation and speech sound stimulation. PET data of the silent condition was subtracted from that of speech stimulation to determine changes in rCBF. In the search for changes in rCBF in the areas for auditory processing, three regions of interest (ROI) were selected; primary auditory area, auditory association area and Broca’s area. The correlation between the rCBF changes in the ROIs and the speech perception performances was analyzed using Pearson’s correlation coefficient. Results: The patient’s speech perception performances ranged widely. Although there were no significant correlations between the speech perception and the rCBF increases in the primary auditory area and Broca’s area, there were positive correlations in the auditory association area. In the left auditory association area, the correlation coefficient of the vowel perception performance was 0.546 ( P<0.05) and that of the speech-tracking test was 0.657 ( P<0.05). Regarding the consonant perception performance, the correlation coefficient was 0.743 ( P<0.01). There was a positive correlation only between the consonant perception performance and the rCBF increase ( R=0.576, P<0.05) in the right auditory association area. These correlations are stronger in the left hemisphere than in the right hemisphere. Conclusions: It is suggested that the improvement of the auditory processing of speech in CI users with SPEAK strategy is accompanied by the recruitment of more neurons in the auditory association areas. The adult auditory cortices may still have plasticity or ability of reorganization. In addition, the left hemisphere may be more involved in speech perception than the right hemisphere.