Search

Steady-state auditory evoked potentials (SSAEPs) in alert adults are most detectable at stimulus or modulation rates of about 40 Hz. Sedation reduces the detectability of 40-Hz SSAEPs and increases it for higher rate SSAEPs. This study examined whether rates higher than 40 Hz would be preferable for detecting responses to low-intensity tones in sedated adults. Fourteen normal adults listened to 640-Hz tones at modulation rates (and toneburst rates) of 20-160 Hz, in 10-Hz steps, at levels of 38 and 58 dB peak equivalent sound-pressure level (peSPL) (20 and 40 dB normal hearing level (nHL) for amplitude-modulated (AM) tones), both alert and sedated (1-2 g chloral hydrate). Sedation reduced both signal (SSAEP) power and noise power at all rates, but noise power reduction was greater for higher rates. Detectability in the alert condition was always greatest at 40 Hz. Under sedation, a second detectability peak was present at 90 Hz for 58-dB peSPL tones, approximately equal to that seen at 40 Hz. At 38 dB peSPL (sedated), peak detectability moved from 40 to 50 Hz. These results suggest that presentation/modulation rates around 40 Hz may be optimal for SSAEP detectability at low levels in adults, whether alert or sedated.

Sinusoids in background noise can conveniently be detected using unsegmented power spectra, comparing power at the signal frequency to average power at several neighbor frequencies. In this case, the F test is preferable to t tests based on rms or dB values, because of the skewed distributions of rms and dB when signal-to-noise ratio (SNR) = 0. F-test performance improves as the number of frequencies increases, to about 15, but can be degraded if the background noise is not white, with a slope exceeding about 10 dB for the range of frequencies sampled. Segment analysis, using magnitude-squared coherence (MSC) or related statistics, has equivalent statistical power; MSC and F each yield unbiased SNR estimates that have identical distributions when SNR = 0. Selection of F or MSC for detection of sinusoids will usually be a matter of convenience.

The frequency-following response (FFR) and the envelope-following response (EFR) were recorded in 1-month-old infants and in adults to examine the development of temporal coding. The stimuli were amplitude-modulated (AM) tones. A modulation frequency of 80 Hz was used in infants; modulation frequencies of 40 and 80 Hz were used in adults. The effects of intensity, carrier frequency, and modulation frequency on these responses were studied. Responses were analyzed using magnitude-squared coherence. The effect of intensity on the growth of FFR- and EFR-coherence were similar in infants and adults. In addition, the growth functions were not affected by the carrier frequency or the modulation frequency of the stimulus. FFR thresholds did not differ across age groups. 'Best frequency' (i.e., infant 80 Hz and adult 40 Hz) EFR thresholds were the same for infants and adults at 500 and 1000 Hz, but infant EFR thresholds were poorer than adult thresholds at 2000 Hz. Thus, although FFRs and EFRs are primarily adult-like at 1 month of age, there are some age differences in the EFR that deserve further study.

Objective: Objective threshold estimation is possible using statistics such as magnitude-squared coherence (MSC) and responses such as the 40-Hz auditory evoked potential (AEP). The purpose of this paper is to compare three general methods of threshold estimation using MSC., Design: Using 500-Hz amplitude-modulated tones and 40-Hz AEPs from 10 paid-volunteer young adults with normal hearing, we compared three MSC-based threshold estimation methods--extrapolation, intensity series, and adaptive., Results: The extrapolation method yielded low thresholds but unacceptable variability. The intensity series method was insensitive and time-consuming. Two adaptive methods displayed good performance in terms of precision and sensitivity. The first employed an MSC criterion with a 13.5% false positive rate (per trial), but achieved overall false positive rates under 5% after 5 to 7 runs, requiring 30 to 40 minutes test time (a run is a single set of ascending or descending trials, terminated by a reversal). The second, which performed better overall, included only a single run, but collected more data at intensities near threshold than at higher intensities; test time could be further reduced (from 22 to 18 minutes) by using a stopping rule based on initial MSC and noise power estimates at each test intensity. If speed is paramount, a simple single-run method achieved threshold estimates in about 9 minutes, starting at 40 dB nHL, with acceptable precision but less sensitivity., Conclusions: A single-run adaptive method, with or without collection of additional data near threshold, yielded the best results, in terms of statistical performance and data collection time.

Magnitude-squared coherence (MSC) was used to detect 40-Hz auditory-evoked potentials (AEPs) in a two-alternative forced-choice experiment. MSC performance was found to be better than that of ten human observers. Experienced observers did not perform better than naive observers. In a "yes-no" experiment, observers demonstrated large differences in criterion, with false-positive rates of 2%-20%. Observers also performed better when AEPs were presented in a 50-ms window (two response cycles) than when presented in a 205-ms window (eight response cycles), although total test time was the same; MSC performed equally well in both cases. MSC provides a useful and objective adjunct to visual detection of 40-Hz AEPs.

The relative performance of phase coherence (PC) and magnitude-squared coherence (MSC) for detection of steady-state evoked potentials was studied using 40 Hz auditory evoked potentials (AEPs) in 10 normal human subjects. In addition, simulation experiments were carried out to determine the effects of signal amplitude and phase variability on detection performance. All simulations showed MSC performance to be better than PC performance, with further improvements when MSC was supplemented with weighted averaging. However, human 40 Hz AEP data showed essentially identical detection performance for PC and MSC, with or without weighted averaging. These data support a "phase aggregation" model (at least near threshold) over the more usual model in which an AEP signal is added to a stationary noise. Human data collected under "no-stimulus" conditions agree well with theoretical distributions for both PC and MSC. For equal test time, long analysis periods (with less averaging) yielded equal performance to short analysis periods (with more averaging), for both PC and MSC.

Objective response detection statistics such as magnitude-squared coherence (MSC) reflect the degree to which subaverage phases are clustered or dispersed, but not their agreement with an expected, or target, phase. Using signal-plus-noise simulations and human 40-Hz auditory evoked potentials, we tested MSC performance with and without phase weighting, in which MSC values were multiplied by weights related to the phase error between measured phase and target phase. Phase weighting improved MSC performance for both simulated and 40-Hz auditory evoked potential data. However, the improvement was greater for the simulations, probably because target phase was precisely known.

We have previously described the usefulness of the magnitude-squared coherence function in analysis of auditory evoked potentials (AEPs) (Ear Hear., 1989, 20: 2-13). For each frequency of interest, the coherence value can be compared to a 'critical value' to determine whether a response is present. Coherence functions must be smoothed across either multiple subaverages or adjacent frequencies (or both) to be reliable, but there are trade-offs: increasing the degree of smoothing increases computational time and (in the case of frequency smoothing) reduces spectral resolution. Using AEPs to clicks and amplitude-modulated tones we investigated the effects of variable degrees of smoothing on threshold estimates in 10 normal human subjects. Thresholds were found to be lower for coherence estimates than for visual detection and are also lower for longer data collection periods. However, there appears to be little if any advantage to segment smoothing beyond 8-16 subaverages. The optimal degree of frequency smoothing is more difficult to specify, depending on the spectrum of the AEP being analyzed (especially the rate of change of phase), and the spectral resolution of the analysis system.

Auditory evoked potentials (AEPs) to 40 Hz clicks and amplitude-modulated 500 Hz tones in human subjects were digitally filtered using an optimal ('Wiener') filter uniquely determined for each AEP. Use of coherence functions to compute coefficients appropriate for filtering grand average AEPs or subsets such as split-half averages is described. Wiener-filtered AEPs correlated better than unfiltered AEPs with split-half replicates and with reference AEPs (obtained with long data collection periods). Visual detection thresholds were lower (more sensitive) for the Wiener-filtered AEPs, but not as low as objectively determined thresholds using coherence values.

Magnitude squared coherence is a method of auditory evoked potential (AEP) frequency-domain analysis, measuring the degree to which the AEP is determined by the stimulus as a function of frequency. In 10 normal-hearing human subjects, scalp responses to an amplitude-modulated (AM) tone (500 Hz carrier modulated with a 40 Hz envelope) were recorded. Critical value criteria were utilized in the statistical analysis of coherence-intensity functions for determination of threshold responses. When compared with subjective determination of AEP threshold from time-domain waveforms, coherence analysis provided a significantly (p less than 0.001) more sensitive threshold measure. Coherence analysis provides information on the spectral content of the response, and allows for the objective determination of threshold which may potentially be utilized in a more expeditious threshold measurement scheme.

The effects of click stimulus presentation rate, spectrum, and polarity on binaural interaction in auditory brain stem responses (BI/ABR) were observed in guinea pigs and in normal-hearing human adults. The effects of repetition rate on the BI/ABR in both species were similar to those seen in ABRs. Six of 16 human subjects showed an unusual "early BI" only at high click intensity and high repetition rates; this is probably due to acoustic reflex effects. Animal data from stimulus spectrum experiments suggest a mid-to-high frequency dependence of the BI/ABR, while human data are less clear. Results are discussed relative to suggested recording strategies for the BI/ABR.

Averaged scalp potentials evoked by continuous pseudorandom noise can be cross-correlated with the evoking stimulus, yielding a cross-correlation function (CCF) which reflects neural phase-locking and is quite sensitive for low-frequency stimulus components [M.J. Wilson and R.A. Dobie (1987) Electroencephalogr. Clin. Neurophysiol. 66, 529-538]. However, for higher frequency signals, replicable CCFs can only be obtained at moderate to high intensities. Since auditory neurons also respond to envelopes of complex sounds, even for high-frequency carriers, we compared scalp responses evoked by band-limited complex sounds to the envelopes of these sounds; the resultant envelope cross-correlation functions (ECCFs) contained replicable response components primarily below 1,000 Hz, regardless of the evoking stimulus spectrum. ECCF thresholds for three octave-band stimuli (830-1,562, 1,611-3,125, and 3,174-6,201 Hz) were more sensitive than CCF thresholds (P = 0.006), averaging 35 dB spectrum level for 10 normal subjects. When stimuli with only odd harmonics were used, replicable odd-component scalp responses were seen only in the spectral range of the stimuli, while even-component responses (presumably to stimulus envelope) were seen only in low-passed scalp responses.

Binaural interaction (BI) in auditory brain-stem responses (ABRs) can be shown by comparing a binaurally elicited ABR to the algebraic sum of monaurally elicited ABRs. Subtracting the summed monaural ABR from binaural ABR yields a wave form assumed to contain response components attributable to BI. It has been suggested that acoustic crosstalk accounts for some of the 'BI' seen with this technique and that contralateral masking should be used during monaural stimulation to eliminate crosstalk. However, this practice might in itself confound the results, even if the masking noise were not intense enough to affect the opposite ear, by 'central masking' of brain-stem neural activity. We studied the effects of contralateral wide-band masking on BI/ABR in 10 normal adult subjects. Clicks were presented at levels from 55 to 115 dB peSPL, at 10 dB intervals. Masking was presented at 73 dB SPL (47 dB effective masking level); based on pilot studies of interaural attenuation, this was a level expected to be at, or just below, the threshold of audibility in the contralateral ear. BI/ABR wave forms were not noticeably affected by the addition of contralateral noise in the monaural trials. In addition, BI/ABR was seen (as previously reported) at levels well below any possibility of crossover artifact. Thus, BI/ABR is not simply attributable to crossover. Contralateral masking is not necessary in recording BI/ABR, within the limits of the stimuli and transducers used in this study.

Evoked potentials are usually analyzed in the time domain (voltage versus time). The most familiar frequency-domain measure, the power spectral density function, displays power as a function of frequency but doesn't distinguish signal power from noise power. The coherence function estimates, for each frequency, the ratio of signal power to total (signal plus noise) power and, thus, indicates the degree to which system output (scalp potential) is determined by input (acoustic stimulus). Coherence ranges from 0 to 1; values above specified critical values can be accepted as demonstrating statistically significant system response. In this paper, we present coherence analysis of human scalp responses to clicks and amplitude-modulated tones. In both cases, this analytic method provides insight into the spectral character of the response (for example, assisting in specifying desirable filter characteristics). Threshold sensitivity is also improved: statistically significant responses can be detected at lower intensity by coherence analysis than by inspection of time-domain waveforms.

Sixteen normal-hearing adults were tested in both awake and sedated states for several evoked responses to low-frequency stimuli. Responses obtained by cross-correlation function, middle latency responses, and the 40-Hz response proved most sensitive; all had mean thresholds of less than 20 dB normal hearing level for the awake-alert state, but 40-Hz and middle latency response mean thresholds were shifted about 10 dB under sedation. The cross-correlation method seems to offer promise for pediatric auditory assessment.

Short-latency auditory responses were obtained by cross-correlation of continuous, pseudorandom noise stimuli with averaged scalp potentials from adults with normal hearing. Responses were recorded for spectrum levels of 14-74 dB for noise bandwidths from 800 to 6000 Hz. At the lowest intensity level of broadband noise, all 10 subjects showed replicable cross-correlation functions (CCFs), which were characterized by prominent positive peaks at delays (latencies) of 5-7 msec. Male subjects exhibited longer delays than females. Delay (latency) increased with decreasing stimulus intensity. Very early responses (less than 2 msec) attributable to cochlear microphonic, which were prominent in earlier work on guinea pigs, were not well seen in these human data. CCFs for responses to band-limited stimuli and off-line derivation of band-limited CCFs for responses evoked by broadband stimuli both showed that this technique is most sensitive to frequency-following behavior at low frequencies (less than 800 Hz). However, definite phase-locked responses to even the highest passband (3100-6200 Hz) were seen. These results support the use of the CCF technique as an efficient method of frequency-specific assessment of the auditory system.