連続音声に対する下丘聴ニューロンの応答

抄録

The neuroral response of the collicular auditory neurons of the cat were investigated using the connected speech and the vowel /a/ segmented from the connected speech. Using a CAT B400 computer, PST (poststimulus time) histograms are made from the spike data obtained for quantitative measurements and the correspondence between the components of speech sound are discussed. In the PST histograms, the abscissa is time and the height along the ordinate represents the number of spikes which occurred in a particular time bin over the 64 repetitions of the speech sound. (1) The collicular auditory neurons responding to the voiced sound carry the information concerning the fundamental frequency of the voice by means of spike train time patterns regardlees of the response type, CF or stimulating intensity level (Fig. 2). (2) When the intensity level of the stimulus is increased, the neuroral activity increases and the duration of response to the sound becomes long due to the response of slow adaptation-type neurons. The response pattern on the histograms for on-type neurons is different from that for slow adaptation-type neurons (Fig. 3). (3) The relation between the frequency components of speech sound and the response patterns for on-type neurons with various CF is diverse and complex. (4) The neuroral responses for a vowel segmented from the connected speech differs from the responses for the same vowel in connected speech. In many cases, the neuroral activity for the segmented vowel increases (Figs. 6 and 7). There appears to be a particular inhibition having a delay extending over 100 msec in the auditory neural network, and thus neuroral response th the segmented vowel increases due to the reduction of such an inhibition which is produced by the preceding syllable.