Sequential recording of three kinds of acoustically evokad vertexpotentials, ABR-MLR-SVR, was analysed by both linear time-scale and logarithmic time-scale, to elucidate the identification and duplication of every component of these resposes. The subjects was 8 young adults with normal hearing. The responeses were recorded from the vertex-ipsilateral ear, which were avaraged up to 50-400 times. Analysis time was 500msec. The stimuli were 80dBnHL clicks with 600meec. intervals. The results were as follows: 1. When the analysis time is 500mesc, the clear detection of fast components of ABR was impossible if the time axis was less than 200cm. with linear time scale recording. However, 10cm. time-base axis was enough for clear detection of ABR fast-components and succeeding responses with logarithmic time-scale. 2. Pl component of SVR was the combined potential of Pa and Pb components of MLR. Pl is not simply the same potential with Pb, as is generally described. This duplication was caused by the filtering maneuver in signal processing.
Derived ABR was obtained at four frequencies (0.5kHz, 1kHz, 2kHz and 4kHz) by using pure-tone masking. It was observed that the component, W6 showed significantly larger amplitude in 1kHz derived ABR than in 2kHz and 4kHz derived ABR's and W7 showed significantly larger amplitude in 0.5kHz derived ABR than in 2kHz and 4kHz derived ABR, suggesting the frequency specificity of these componets. From the results obtained, it was confirmed that waves W6 and W7 represent the activity initiated from low-frequency regions along the cochlear partition.
The masking levels measured by band noise (band noise masking level: BNML) and by pure tones (pure tone masking level: PTML) at the pitch of tinnitus were compared in 121 tinnitus ears without hearing loss or associated with cochlear deafness in those pitch was simulated by band noise. The results obtained were as follows; (1) The correlation coefficient between PTML and BNML was 0.81 when both masking levels were indicated by sensation level and it was 0.94 when they were indicated by audiometer level. (2) The correlation coefficient between loudness and masking level was 0.75 in BNML and was 0.61 in PTML when both loudness and masking level were indicated by sensation level. It was 0.94 in BNML and 0.88 in PTML when they were indicated by audiometer level. (3) The regression line between the loudness (x) and the masking level (y) was estimated. It was y=1.23x+6.43 in BNML and y=1.17x+8.44 in PTML when both loudness and masking level were indicated by sensation level. It was y=0.93x+10.53 in BNML and y=0.89x+19.25 in PTML when they were indicated by audiometer level. (4) The relationship between the band noise hearing level at the pitch of tinnitus and the loudness, PTML and BNML was analysed. The loudness, PTML and BNML indicated by sensation level became small abruptly when the band noise hearing level exceeded 30-40dB. However, PTML and BNML indicated by audiometer level were slightly influenced by the band noise hearing level at the pitch of tinnitus. From these results, it was concluded that the BNML indicated by audiometer level could be most useful as an indicator representing the intensity of tinnitus in which pitch was simulated by band noise.
The oto-acoustic emission (OAE) was recorded by air-conduction stimulation from the contralateral ear in 12 normal hearing ears with type A tympanogram, and also by air-conduction stimulation from the ipsilateral ear. The results obtained were as follows; 1) The OAE was evoked by air-conduction stimulation from the contralateral ear in all five frequencies between 1.0kHz and 4.0kHz in all 12 ears. 2) The pseudothreshold of OAE evoked by the contralateral stimulation was about 50dB higher than that of OAE evoked by the ipsilateral stimulation. 3) The averaged emission cochleogram between 1.0kHz and 4.0kHz showed the similar configurations both by the ipsilateral stimulation and by the contralateral stimulation. 4) The OAE elicited by air-conduction stimulation from the contralateral ear could be evoked by bone-conduction stimulation due to the similar mechasism as the shadow hearing in psychoacoustic audiometry.
The effects of temporal factors on the vowel perception of the sensorineural hearing-impaired listeners were examined. In the experiment, the ability to identify Japanese five vowels with various durations and two-vowel sequences with various inter-vowel intervals were examined. For isolated vowel perception, the shorter the vowel segment was, the lower the intelligibility was. Although the intelligibility of individual vowels of the sequences tends to be lower than that for isolated vowels, it tends to improve with a silent interval between the vowels becomes longer. These results indicate that the vowel perception of sensorineural hearing-impaired is easily affected by the temporal factors in the stimuli. This indicates that the temporal processing in auditory system of the hearing-impaired is reduced.
In order to applicate the phase spectral analysis of auditory brainstem response (ABR) to the clinical diagnositic test, the following basic studies were performed. The changes of component synchrony measure (CSM) values in different numbers of sweep for each set of group averages and in different frequency spacings were investigated in 20 normal adults. Moreover, the relationship between ABR wave form and phase spectrum was investigated in 20 normal adults and 384 patients (768 ears). The results were as follows; (1) CSM values without stimulation were almost same (mean value was about 0.1) in different numbers of sweep and in different frequency spacings. (2) In the presence of stimulation, CSM values for responsive frequency components were high in rough frequency spacing (97.7Hz), and low in fine frequency spacing (24.4Hz). (3) The greater the number of sweep, the higher the CSM values for responsive frequency components. (4) Component A (0-400Hz) of the phase spectrum related to a slow component of ABR and component C (800-1200Hz) related to a fast component. Component B (400-800Hz) was likely to refect both fast and slow components of ABR.
A consonant confusion study was performed in 15 patients with a 22-channel cochlear implant device (Nucleus). The data were analyzed and the information transfer of consonants was investigated based on the information theory. The results were as follows: 1) The information transfer in the condition of cochlear implant was greater than in the lipreading condition for voicing, nasality and semi-vowels. 2) The information transfer in the lipreading condition was greater than in the cochlear implant condition for the place of articulation. 3) The scores in the cochlear implant plus lipreading were roughly similar to the sums of scores of the cochlear implant alone condition and of the lipreading alone condition. 4) The cochlear implant and the lipreading were mutually complementary for recognition of monosyllable sounds.