Sandra C. Champagne

Frequency-specific Audiometry Using Steady-state Responses

Objective: To evaluate the audiometric usefulness of steady‐state responses to multiple simultaneous tones, amplitude‐modulated at 75 to 110 Hz. Design: Steady‐state responses to multiple tones amplitude‐modulated at different rates between 75 and 110 Hz and presented simultaneously were recorded at different intensities in normal adults, well babies, normal adults with simulated hearing loss, and adolescents with known hearing losses. Response thresholds were compared with behavioral thresholds. Results: In normal adults the thresholds for steady‐state responses to tones of 0.5, 1, 2, and 4 kHz were 14 ± 11, 12 ± 11, 11 ± 8, and 13 ± 11 dB, respectively, above behavioral thresholds for air‐conducted stimuli, and 11 ± 5, 14 ± 8, 9 ± 8, and 10 ± 10 dB above behavioral thresholds for bone‐conducted stimuli. In well babies tested in a quiet environment, the thresholds were 45 ± 13, 29 ± 10, 26± 8, and 29 ± 10 dB SPL. In adolescents with known hearing losses, the steady‐state responses thresholds predict behavioral thresholds with correlation coefficients (r) of 0.72, 0.70, 0.76, and 0.91 at 0.5, 1, 2, and 4 kHz, respectively. Conclusion: Steady‐state responses to tones amplitude‐modulated at 75 to 110 Hz can be used for frequency‐specific objective audiometry. The multiple‐stimulus technique allows thresholds to be estimated for eight different stimuli at the same time.

Potentials evoked by the sinusoidal modulation of the amplitude or frequency of a tone

Steady state responses to the sinusoidal modulation of the amplitude or frequency of a tone were recorded from the human scalp. For both amplitude modulation (AM) and frequency modulation (FM), the responses were most consistent at modulation frequencies between 30 and 50 Hz. However, reliable responses could also be recorded at lower frequencies, particularly at 2-5 Hz for AM and at 3-7 Hz for FM. With increasing modulation depth at 40 Hz, both the AM and FM response increased in amplitude, but the AM response tended to saturate at large modulation depths. Neither response showed any significant change in phase with changes in modulation depth. Both responses increased in amplitude and decreased in phase delay with increasing intensity of the carrier tone, the FM response showing some saturation of amplitude at high intensities. Both responses could be recorded at modulation depths close to the subjective threshold for detecting the modulation and at intensities close to the subjective threshold for hearing the stimulus. The responses were variable but did not consistently adapt over periods of 10 min. The 40-Hz AM and FM responses appear to originate in the same generator, this generator being activated by separate auditory systems that detect changes in either amplitude or frequency.

Auditory steady‐state responses to tones amplitude‐modulated at 80–110 Hz

Steady-state responses can be recorded from the human scalp in response to tones that are sinusoidally modulated in amplitude at rates between 60 and 120 Hz. For 60 dB SPL 1000-Hz tones the maximum baseline-to-peak amplitude of about 0.06 microV occurs for modulation rates between 80 and 95 Hz. The phase of the response does not change with modulation depths greater than 25% and the amplitude saturates at modulation depths greater than 50%. The presence or absence of a response can be accurately determined by frequency-domain statistics and the response becomes clearly recognizable at intensities that are 16 +/- 8 dB above behavioral thresholds. With increasing intensity the response increases in amplitude at 1.9 nV/dB until an intensity of 70 dB SPL. As the intensity increases above 70 dB SPL the response increases in amplitude more rapidly at 7.8 nV/dB (at 1000 Hz) and contains significant energy at harmonics of the modulation frequency. This second stage of the intensity function is more prominent for stimuli with lower carrier frequencies (500 more than 1000 more than 2000 Hz) and is attenuated by high-pass masking. These steady-state responses should be helpful in evaluating human auditory physiology and in objective audiometry.

Human auditory evoked potentials recorded using maximum length sequences

A maximum length sequence (MLS) is a specially constructed pseudorandom binary sequence that can be used to control the presentation of sensory stimuli. The evoked potentials to such a sequence of stimuli can be analyzed to give the response to one stimulus in the sequence. This procedure allows auditory evoked potentials to be recorded at stimulus rates that would cause a confusing overlap of responses with regular averaging. The MLS technique can be used with auditory evoked potentials at all latencies although it is most effective for the brain-stem and middle-latency responses. By demonstrating different refractory periods for different parts of the response, the technique may help delineate the component structure of the evoked potential. As well, an MLS analysis can disentangle the auditory brain-stem response from overlapping middle-latency responses during evoked potential audiometry.

Human evoked potentials to shifts in the lateralization of a noise

A continuous noise was generated by running a sequence of random numbers through a digital-analog converter and connecting the output through an amplifier and filter to an earphone. Two channels were programmed to generate identical noise stimuli with one channel delayed relative to the other. When these stimuli were presented through earphones, the subject lateralized the noise to the side receiving the leading stimulus. Changes in the relative timing of the two stimuli caused the noise to shift its lateralization. Since these shifts occurred without any detectable change in the ongoing monaural noise, any potentials they evoked were specifically related to binaural interaction. The response recorded from the vertex contained a positive-negative-positive complex with peak latencies of 75, 136 and 220 ms. This response was similar to that evoked by the onset of a monaural stimulus although it was slightly smaller and significantly later. Despite several attempts, we were unable to record any definite earlier evoked potentials.

The Timing of the Processes Underlying Lateralization: Psychophysical and Evoked Potential Measures

This article describes a technique to measure binaural integration time. A binaural noise with an interaural time difference of 0.8 msec was presented in three conditions: alone, with intervening noise that was identical between the two ears, or with uncorrelated intervening noise. Both behavioral responses and evoked potentials were recorded. When the stimulus was presented in a quiet background, it was accurately detected and lateralized with durations as short as 2 msec. The N1 peak of the evoked potential occurred at approximately 90 msec. When the stimulus occurred as a brief change in an ongoing correlated binaural noise, a duration of 10 msec was necessary before the sound could be accurately lateralized or an evoked potential elicited. The N1 peak occurred at approximately 120 msec. When the stimulus occurred as a change in an ongoing uncorrelated binaural noise, a duration of 60 msec was necessary for the subject to lateralize the stimulus and for an evoked potential to be elicited. The N1 peak occurred at about 130 msec. These results suggest that a period of approximately 60 msec is required to detect the correlation of an ongoing binaural noise and that a somewhat shorter period is necessary to track changes in a sound source that has already been lateralized. The simplicity of this technique makes it an attractive tool for assessing central auditory function.

Human evoked potentials and the lateralization of a sound.

If an identical noise is presented to each ear with one ear receiving the noise slightly earlier than the other, the listener perceives the sound as originating from the side of the leading ear. If the interaural time-difference reverses, the subject perceives a shift in the lateralization of the sound to the other ear. This shift in lateralization evokes a late auditory potential with a negative wave at 135 ms and positive waves at 75 and 220 ms. This evoked potential specifically indexes central auditory processing since information about the timing of the auditory stimuli must be compared between the two ears. The response increases in amplitude with increasing interaural time-difference reaching maximum values between 0.3 and 1.5 ms. The response is evoked through acoustic frequencies below 2,000 Hz. In patients with multiple sclerosis the response is often abnormally delayed or small. The response may therefore be helpful in the clinical evaluation of patients with central auditory dysfunction.

Brainstem auditory steady‐state responses to multiple simultaneous stimuli

Steady-state responses can follow multiple simultaneous auditory stimuli. If the stimuli are modulated at different rates, responses specific to each stimulus can be assessed by measuring in the frequency domain response the spectral component corresponding to the rate of modulation. When each stimulus has a different carrier frequency or different ear of presentation, the responses when 8 stimuli are presented simultaneously are not significantly different than when each stimulus is presented alone. Since significant responses can be recognized down to intensities that average 14 dB above behavioral threshold, this technique may be useful in objective audiometry. It is also possible to record steady-state responses to multiple modulations of the same carrier frequency. In this case, the amplitude of the responses when the stimuli are combined is smaller than when the stimuli are presented alone. The decrease in amplitude depends upon the number of concomitant stimuli and their relative intensities. These effects are probably due to the compressive rectification occurring during cochlear transduction, and the data may be used to model cochlear processing of auditory stimuli.

Evoked potentials to shifts in the lateralization of a noise: Effect of stimulus frequency

Changing the interaural time difference of a continuous noise arriving at the two ears causes a shift in the perceived lateralization of the noise. This shift in lateralization evokes from human subjects a late evoked potential with a positive‐negative‐positive waveform between 80 and 290 ms. To determine which frequencies of the broadband noise contribute to the response, two experiments were performed. In the first, the noise was low‐pass filtered at 8000, 4000, 2000, 1000, and 500 Hz, and the late auditory‐evoked potentials were recorded. In the second experiment, the noise was bandpass filtered between 250–500, 500–1000, 1000–1500, and 1500–2000 Hz. The results of both experiments show that the evoked potential is mainly mediated by the low frequencies (250–1500 Hz) in the noise. Since it is possible that the response may be elicited through the high‐frequency region of the cochlea, the response in the presence of monaural high‐pass masking noise was evaluated. These results indicate that the response is elicited through the region of the cochlea specific to the frequencies of the sound that carry the lateralizing information.Changing the interaural time difference of a continuous noise arriving at the two ears causes a shift in the perceived lateralization of the noise. This shift in lateralization evokes from human subjects a late evoked potential with a positive‐negative‐positive waveform between 80 and 290 ms. To determine which frequencies of the broadband noise contribute to the response, two experiments were performed. In the first, the noise was low‐pass filtered at 8000, 4000, 2000, 1000, and 500 Hz, and the late auditory‐evoked potentials were recorded. In the second experiment, the noise was bandpass filtered between 250–500, 500–1000, 1000–1500, and 1500–2000 Hz. The results of both experiments show that the evoked potential is mainly mediated by the low frequencies (250–1500 Hz) in the noise. Since it is possible that the response may be elicited through the high‐frequency region of the cochlea, the response in the presence of monaural high‐pass masking noise was evaluated. These results indicate that the response...