Auditory Brainstem and Middle Latency Responses

Abstract

Auditory brainstem (ABR) and middle latency responses (MLR) were recorded for ten normal subjects. Changes in wave latency were recorded with high- and low-pass filtering using Butterworth filters with standard phase-shift characteristics. When an open filter (15–3,000 Hz) was used, waves IV, V and VI of the ABR were superimposed on wave Pø of the MLR. A positive (P) wave was recorded in the trough of wave Na as the high-frequency cutoff of the recording bandpass (15–100 Hz) was raised above 100 Hz. Wave P divided Na into two parts. The first trough was the slow-negative response and the second was Na2. The origin of the P wave is unclear but may represent a muscle potential. Because of previous inconsistencies in the waveform identification used with these recording techniques, the authors present a modified method of classification which accommodates changes in waveform appearance that occur with different response filtering. The thresholds for various waves of the auditory brainstem (ABR) or middle latency response (MLR) were determined to a 500-Hz tone pip in ten normal listeners in a variety of recording filter bandwidths. Wave V (recording filters 100–3,000 Hz) had a threhsold of 26 dB nHL. Wave SN (30–3,000 Hz) had a threshold of 18 dB nHL. Waves Na, Pa and Nb had thresholds of 10 to 11 dB nHL. The threshold for waves Na, Pa and Nb were unaffected by changing the recording bandpass from 15–100 Hz to 15–3,000 Hz. The ABR amplitude increased when the recording filter's low-frequency cutoff was lowered from 100 to 15 Hz (high frequency cutoff, 3,000 Hz). This augmentation is probably due to the inclusion of the lower frequency energy of wave Pø. Lowering the recording filters high-frequency cutoff from 3,000 to 100 Hz (low-frequency cutoff, 15 Hz) resulted in exclusion of the higher frequency ABR but did not elevate response threshold. In all listeners tested, the MLR had a lower threshold than the ABR to a 500-Hz tone pip. It is suggested that the MLR has good potential for use in evaluation of low-frequency threshold.