The middle components of the auditory evoked response in young children were investigated during awake and asleep. The results were as follows. (1) The wave form, latency, and appearance of the middle components depend on the test condition, that is, awake or asleep. Although the detectability of Po-peak was about the same for both of the test conditions, that of Na- and Pa- peak were lower in the sleeping state than in waking state and this difference was prominent in the Pa-peak. (2) The later peaks, such as Nb- and Pb-peak, were scarcely detected in the childring both waking and sleep. (3) It is surmised that the middle components may be divided into four groups comprising Po, Na, Pa and peaks later than Nb and that the each group has the different origins or pathways of the response. (4) It was found that there was a large discrepancy between the appearance of Pa or the later peaks such as Nb and Pb of Mendel's group and that of ours. A reason of this discrepancy may be due to the difference of filter slope.
Middle latency responses were recorded from 6 adults and 8 young children of normal hearing. Their EEG activities were stored in the data recorder, and played back in offline and averaged through the input filters with 9 combinations of pass-band and rejection slope; 1-200Hz, 10-200Hz and 30-200Hz for pass-band, and 6, 36, and 84dB/oct for rejection slope. Distortion of MLR wave forms were marked for MLRs obtained through the filters with 36dB/oct slope and 30Hz at low-frequency cutoff, and 84dB/oct slope and 10 and 30Hz at low-frequency cutoff. Later components appeared and their amplitudes increased as the slope of filter was set steeper and low-frequency cutoff higher. Only Po and Na were obtained from the children under 4 years old through the filter with 6dB/oct and 1Hz at low-frequency cutoff, whereas the MLRs of adults showed clear Pa and later components through the same filter. However, MLRs of children obtained through the filter with 36dB/oct and 10Hz at low-frequency cutoff showed Pa component. The filters with steeper slope elicited the later components more clearly, which lead to the conclusion that the later components in young children were artificially made by the filter.
The MLR components were approximated by half cycle sinusoids. The synthetic and original MLR were passed through the filters with various slope and pass-band, and compared with each other. The effect of filters on latency and amplitude of the original and synthetic wave were almost the same between them. However, some portion of MLR wave showed a little difference between original and synthetic wave. When MLRs passed through the filters with steeper slope and higher frequency at low-frequency cutoff, this difference moved to the later components. The effects of filter were studied on single half cycle sinusoid of 10 to 100Hz. The filters with 6dB/oct slope showed little change in the wave form of sinusoid, but marked wave form distortion was observed in those sinusoid passed through the filters with 36dB/oct or more steeper slope, and 10Hz or higher at low-frequency cutoff. Damped oscillation were observed, and their effects ranged to 3 or 4 later components.
In order to obtain a frequency-specific wholenerve response in man, the special frequencymodulated tone was developed and the responses to this tone were studied. A downward shift of 20% in the frequency of a 0.5kHz, 1kHz or 2kHz tone with an amplitude increase elicited the wholenerve response, and this response was recorded by the intratympanic needle electrode. The latency is greater for lower-frequency tones and this latency increase agrees well with the travelling delay for the frequency range used in this study. Thus, it is strongly suggested that the frequency-modulated tone in downward direction with increased amplitude yields the frequency specific whole-nerve response.