Interpretation of the human frequency-following response

Abstract

The human frequency-following response (FFR) is recorded (vertex to mastoid) by standard response-averaging methods. 500-Hz tone bursts with 4-msec rise time are excellent stimuli. The threshold is sharp at approximately 40 dB SL. FFR is weak in a subject with abrupt high-tone hearing loss. We believe that FFR is evoked primarily by stimulation of the basal turn of the cochlea. Each sound wave evokes a quasi-rhythmic sequence of waves, like the midbrain response to a click, and the sequences overlap. Experimental proof is difficult because the identity and latency of the first waves of FFR are uncertain and there are differences among subjects in the waveform of the midbrain responses and of their overlap when repeated at 500 Hz. Masking with high-pass (2000 Hz) noise sometimes reveals an apically generated FFR with a longer latency corresponding to the traveling-wave delay, but this result is clearer with brief low-frequency transients than with tone bursts. In a clinical setting FFR is not likely to yield information about the apical portion of the cochlea, and for the basal turn the midbrain responses at 6 to 9 msec evoked by high-frequency tone pips (4000 Hz) have lower threshold and are very reliable.