Comparison of absolute thresholds derived from an adaptive forced-choice procedure and from reaction probabilities and reaction times in a simple reaction time paradigm

Towards a unifying basis of auditory thresholds: Thresholds for multicomponent stimuli

Sounds consisting of multiple simultaneous or consecutive components can be detected by listeners when the stimulus levels of the components are lower than those needed to detect the individual components alone. The mechanisms underlying such spectral, spectrotemporal, temporal, or across-ear integration are not completely understood. Here, we report threshold measurements from human subjects for multicomponent stimuli (tone complexes, tone sequences, diotic or dichotic tones) and for their individual sinusoidal components in quiet. We examine whether the data are compatible with the detection model developed by Heil, Matysiak, and Neubauer (HMN model) to account for temporal integration (Heil et al. 2017), and we compare its performance to that of the statistical summation model (Green 1958), the model commonly used to account for spectral and spectrotemporal integration. In addition, we compare the performance of both models with respect to previously published thresholds for sequences of identical tones and for diotic tones. The HMN model is similar to the statistical summation model but is based on the assumption that the decision variable is a number of sensory events generated by the components via independent Poisson point processes. The rate of events is low without stimulation and increases with stimulation. The increase is proportional to the time-varying amplitude envelope of the bandpass-filtered component(s) raised to an exponent of 3. For an ideal observer, the decision variable is the sum of the events from all channels carrying information, for as long as they carry information. We find that the HMN model provides a better account of the thresholds for multicomponent stimuli than the statistical summation model, and it offers a unifying account of spectral, spectrotemporal, temporal, and across-ear integration at threshold.

Modulation in cortical excitability disrupts information transfer in perceptual-level stimulus processing

Despite significant interest in the neural underpinnings of behavioral variability, little light has been shed on the cortical mechanism underlying the failure to respond to perceptual-level stimuli. We hypothesized that cortical activity resulting from perceptual-level stimuli is sensitive to the moment-to-moment fluctuations in cortical excitability, and thus may not suffice to produce a behavioral response. We tested this hypothesis using electrocorticographic recordings to follow the propagation of cortical activity in six human subjects that responded to perceptual-level auditory stimuli. Here we show that for presentations that did not result in a behavioral response, the likelihood of cortical activity decreased from auditory cortex to motor cortex, and was related to reduced local cortical excitability. Cortical excitability was quantified using instantaneous voltage during a short window prior to cortical activity onset. Therefore, when humans are presented with an auditory stimulus close to perceptual-level threshold, moment-by-moment fluctuations in cortical excitability determine whether cortical responses to sensory stimulation successfully connect auditory input to a resultant behavioral response.

Comparing and modeling absolute auditory thresholds in an alternative-forced-choice and a yes-no procedure

Detecting sounds in quiet is arguably the simplest task performed by an auditory system, but the underlying mechanisms are still a matter of debate. Threshold stimulus levels depend not only on the physical properties of the sounds to be detected but also on the experimental procedure used to measure them. Here, thresholds of human subjects were measured for sounds consisting of different numbers of bursts using both an alternative-forced-choice and a yes-no procedure in the same experimental sessions. Thresholds measured with the yes-no procedure were typically higher than thresholds measured with the alternative-forced choice procedure. The difference between the two thresholds decreased as stimulus duration increased. It also varied between subjects and varied with the probability of false alarms in the yes-no procedure. It is shown that a previously proposed model of detection (Heil et al., Hear Res 2017) can account for these findings better than other models. It can also account for the shapes of the psychometric functions. The model is consistent with basic concepts of signal detection theory but is based on a decision variable that follows Poisson statistics. It also differs from other models of detection with respect to the transformation of the stimulus into the decision variable. The findings in this study further support the model.

Onset-Duration Matching of Acoustic Stimuli Revisited: Conventional Arithmetic vs. Proposed Geometric Measures of Accuracy and Precision

Onsets of acoustic stimuli are salient transients and are relevant in humans for the perception of music and speech. Previous studies of onset-duration discrimination and matching focused on whether onsets are perceived categorically. In this study, we address two issues. First, we revisit onset-duration matching and measure, for 79 conditions, how accurately and precisely human listeners can adjust the onset duration of a comparison stimulus to subjectively match that of a standard stimulus. Second, we explore measures for quantifying performance in this and other matching tasks. The conventional measures of accuracy and precision are defined by arithmetic descriptive statistics and the Euclidean distance function on the real numbers. We propose novel measures based on geometric descriptive statistics and the log-ratio distance function, the Euclidean distance function on the positive-real numbers. Only these properly account for the fact that the magnitude of onset durations, like the magnitudes of most physical quantities, can attain only positive real values. The conventional (arithmetic) measures possess a convexity bias that yields errors that grow with the width of the distribution of matches. This convexity bias leads to misrepresentations of the constant error and could even imply the existence of perceptual illusions where none exist. This is not so for the proposed (geometric) measures. We collected up to 68 matches from a given listener for each condition (about 34,000 matches in total) and examined inter-listener variability and the effects of onset duration, plateau duration, sound level, carrier, and restriction of the range of adjustable comparison stimuli on measures of accuracy and precision. Results obtained with the conventional measures generally agree with those reported in the literature. The variance across listeners is highly heterogeneous for the conventional measures but is homogeneous for the proposed measures. Furthermore, the proposed measures show that listeners tend to under- rather than to overestimate the onset duration of the comparison stimuli. They further reveal effects of the stimulus carrier on accuracy and precision which are missed by the conventional measures. Our results have broad implications for psychophysical studies that use arithmetic measures to quantify performance when geometric measures should instead be used.

Linking the sender to the receiver: vocal adjustments by bats to maintain signal detection in noise

Short-term adjustments of signal characteristics allow animals to maintain reliable communication in noise. Noise-dependent vocal plasticity often involves simultaneous changes in multiple parameters. Here, we quantified for the first time the relative contributions of signal amplitude, duration, and redundancy for improving signal detectability in noise. To this end, we used a combination of behavioural experiments on pale spear-nosed bats (Phyllostomus discolor) and signal detection models. In response to increasing noise levels, all bats raised the amplitude of their echolocation calls by 1.8-7.9 dB (the Lombard effect). Bats also increased signal duration by 13%-85%, corresponding to an increase in detectability of 1.0-5.3 dB. Finally, in some noise conditions, bats increased signal redundancy by producing more call groups. Assuming optimal cognitive integration, this could result in a further detectability improvement by up to 4 dB. Our data show that while the main improvement in signal detectability was due to the Lombard effect, increasing signal duration and redundancy can also contribute markedly to improving signal detectability. Overall, our findings demonstrate that the observed adjustments of signal parameters in noise are matched to how these parameters are processed in the receiver's sensory system, thereby facilitating signal transmission in fluctuating environments.

Simple reaction time to the onset of time-varying sounds

Although auditory simple reaction time (RT) is usually defined as the time elapsing between the onset of a stimulus and a recorded reaction, a sound cannot be specified by a single point in time. Therefore, the present work investigates how the period of time immediately after onset affects RT. By varying the stimulus duration between 10 and 500 msec, this critical duration was determined to fall between 32 and 40 milliseconds for a 1-kHz pure tone at 70 dB SPL. In a second experiment, the role of the buildup was further investigated by varying the rise time and its shape. The increment in RT for extending the rise time by a factor of ten was about 7 to 8 msec. There was no statistically significant difference in RT between a Gaussian and linear rise shape. A third experiment varied the modulation frequency and point of onset of amplitude-modulated tones, producing onsets at different initial levels with differently rapid increase or decrease immediately afterwards. The results of all three experiments results were explained very well by a straightforward extension of the parallel grains model (Miller and Ulrich Cogn. Psychol. 46, 101-151, 2003), a probabilistic race model employing many parallel channels. The extension of the model to time-varying sounds made the activation of such a grain depend on intensity as a function of time rather than a constant level. A second approach by mechanisms known from loudness produced less accurate predictions.

Binaural loudness gain measured by simple reaction time

In order to yield equal loudness, different studies using scaling or matching methods have found binaural level differences between monaural and diotic presentations ranging from less than 2 dB to as much as 10 dB. In the present study, a reaction time methodology was employed to measure the binaural level difference producing equal reaction time (BLDERT). Participants had to respond to the onset of 1-kHz pure tones with sound pressure levels ranging from 45 to 85 dB, and being presented to the right, the left, or both ears. Equal RTs for monaural and diotic presentation (BLDERTs) were obtained with a level difference of approximately 5 dB. A second experiment showed that different results obtained for the left and right ear are largely due to the responding hand, with ipsilateral responses being faster than contralateral ones. A third experiment investigated the BLDERT for dichotic stimuli, tracing the transition between binaural and monaural stimulation. The results of all three RT experiments are consistent with current models of binaural loudness and contradict earlier claims of perfect binaural summation.

Equal latency contours and auditory weighting functions for the harbour porpoise (Phocoena phocoena)

Loudness perception by human infants and animals can be studied under the assumption that sounds of equal loudness elicit equal reaction times (RTs). Simple RTs of a harbour porpoise to narrowband frequency-modulated signals were measured using a behavioural method and an RT sensor based on infrared light. Equal latency contours, which connect equal RTs across frequencies, for reference values of 150-200 ms (10 ms intervals) were derived from median RTs to 1 s signals with sound pressure levels (SPLs) of 59-168 dB re. 1 μPa and centre frequencies of 0.5, 1, 2, 4, 16, 31.5, 63, 80 and 125 kHz. The higher the signal level was above the hearing threshold of the harbour porpoise, the quicker the animal responded to the stimulus (median RT 98-522 ms). Equal latency contours roughly paralleled the hearing threshold at relatively low sensation levels (higher RTs). The difference in shape between the hearing threshold and the equal latency contours was more pronounced at higher levels (lower RTs); a flattening of the contours occurred for frequencies below 63 kHz. Relationships of the equal latency contour levels with the hearing threshold were used to create smoothed functions assumed to be representative of equal loudness contours. Auditory weighting functions were derived from these smoothed functions that may be used to predict perceived levels and correlated noise effects in the harbour porpoise, at least until actual equal loudness contours become available.

Towards a unifying basis of auditory thresholds: binaural summation

Absolute auditory threshold decreases with increasing sound duration, a phenomenon explainable by the assumptions that the sound evokes neural events whose probabilities of occurrence are proportional to the sound's amplitude raised to an exponent of about 3 and that a constant number of events are required for threshold (Heil and Neubauer, Proc Natl Acad Sci USA 100:6151-6156, 2003). Based on this probabilistic model and on the assumption of perfect binaural summation, an equation is derived here that provides an explicit expression of the binaural threshold as a function of the two monaural thresholds, irrespective of whether they are equal or unequal, and of the exponent in the model. For exponents >0, the predicted binaural advantage is largest when the two monaural thresholds are equal and decreases towards zero as the monaural threshold difference increases. This equation is tested and the exponent derived by comparing binaural thresholds with those predicted on the basis of the two monaural thresholds for different values of the exponent. The thresholds, measured in a large sample of human subjects with equal and unequal monaural thresholds and for stimuli with different temporal envelopes, are compatible only with an exponent close to 3. An exponent of 3 predicts a binaural advantage of 2 dB when the two ears are equally sensitive. Thus, listening with two (equally sensitive) ears rather than one has the same effect on absolute threshold as doubling duration. The data suggest that perfect binaural summation occurs at threshold and that peripheral neural signals are governed by an exponent close to 3. They might also shed new light on mechanisms underlying binaural summation of loudness.

Interdependent effects of sound duration and amplitude on neuronal onset response in mice inferior colliculus

In this study, we adopted iso-frequency pure tone bursts to investigate the interdependent effects of sound amplitude/intensity and duration on mice inferior colliculus (IC) neuronal onset responses. On the majority of the sampled neurons (n=57, 89.1%), sound amplitude and duration had effects on the neuronal response to each other by showing complex changes of the rat-intensity function/duration selectivity types and/or best amplitudes (BAs)/durations (BDs), evaluated by spike counts. These results suggested that the balance between the excitatory and inhibitory inputs set by one acoustic parameter, amplitude or duration, affected the neuronal spike counts responses to the other. Neuronal duration selectivity types were altered easily by the low-amplitude sounds while the changes of rate-intensity function types had no obvious preferred stimulus durations. However, the first spike latencies (FSLs) of the onset response neurons were relative stable to iso-amplitude sound durations and changing systematically along with the sound levels. The superimposition of FSL and duration threshold (DT) as a function of stimulus amplitude after normalization indicated that the effects of the sound levels on FSLs are considered on DT actually.

Detection of Near-Threshold Sounds is Independent of EEG Phase in Common Frequency Bands

Low-frequency oscillations in the electroencephalogram (EEG) are thought to reflect periodic excitability changes of large neural networks. Consistent with this notion, detection probability of near-threshold somatosensory, visual, and auditory targets has been reported to co-vary with the phase of oscillations in the EEG. In audition, entrainment of δ-oscillations to the periodic occurrence of sounds has been suggested to function as a mechanism of attentional selection. Here, we examine in humans whether the detection of brief near-threshold sounds in quiet depends on the phase of EEG oscillations. When stimuli were presented at irregular intervals, we did not find a systematic relationship between detection probability and phase. When stimuli were presented at regular intervals (2-s), reaction times were significantly shorter and we observed phase entrainment of EEG oscillations corresponding to the frequency of stimulus presentation (0.5 Hz), revealing an adjustment of the system to the regular stimulation. The amplitude of the entrained oscillation was higher for hits than for misses, suggesting a link between entrainment and stimulus detection. However, detection was independent of phase at frequencies ≥1 Hz. Furthermore, we show that when the data are analyzed using acausal, though common, algorithms, an apparent "entrainment" of the δ-phase to presented stimuli emerges and detection probability appears to depend on δ-phase, similar to reports in the literature. We show that these effects are artifacts from phase distortion at stimulus onset by contamination with the event-related potential, which differs markedly for hits and misses. This highlights the need to carefully deal with this common problem, since otherwise it might bias and mislead this exciting field of research.

Why longer song elements are easier to detect: threshold level-duration functions in the Great Tit and comparison with human data

Our study estimates detection thresholds for tones of different durations and frequencies in Great Tits (Parus major) with operant procedures. We employ signals covering the duration and frequency range of communication signals of this species (40-1,010 ms; 2, 4, 6.3 kHz), and we measure threshold level-duration (TLD) function (relating threshold level to signal duration) in silence as well as under behaviorally relevant environmental noise conditions (urban noise, woodland noise). Detection thresholds decreased with increasing signal duration. Thresholds at any given duration were a function of signal frequency and were elevated in background noise, but the shape of Great Tit TLD functions was independent of signal frequency and background condition. To enable comparisons of our Great Tit data to those from other species, TLD functions were first fitted with a traditional leaky-integrator model. We then applied a probabilistic model to interpret the trade-off between signal amplitude and duration at threshold. Great Tit TLD functions exhibit features that are similar across species. The current results, however, cannot explain why Great Tits in noisy urban environments produce shorter song elements or faster songs than those in quieter woodland environments, as detection thresholds are lower for longer elements also under noisy conditions.

The psychophysics of absolute threshold and signal duration: a probabilistic approach

The absolute threshold for a tone depends on its duration; longer tones have lower thresholds. This effect has traditionally been explained in terms of "temporal integration" involving the summation of energy or perceptual information over time. An alternative probabilistic explanation of the process is formulated in terms of simple equations that predict not only the time ∕ duration dependence but also the shape of the psychometric function at absolute threshold. It also predicts a tight relationship between these two functions. Measurements made using listeners with either normal or impaired hearing show that the probabilistic equations adequately fit observed threshold-duration functions and psychometric functions. The mathematical formulation implies that absolute threshold can be construed as a two-valued function: (a) gain and (b) sensory threshold, and both parameters can be estimated from threshold-duration data. Sensorineural hearing impairment is sometimes associated with a smaller threshold ∕ duration effect and sometimes with steeper psychometric functions. The equations explain why these two effects are expected to be linked. The probabilistic approach has the potential to discriminate between hearing deficits involving gain reduction and those resulting from a raised sensory threshold.

Near-threshold equal-loudness contours for harbor seals (Phoca vitulina) derived from reaction times during underwater audiometry: a preliminary study

Equal-loudness functions describe relationships between the frequencies of sounds and their perceived loudness. This pilot study investigated the possibility of deriving equal-loudness contours based on the assumption that sounds of equal perceived loudness elicit equal reaction times (RTs). During a psychoacoustic underwater hearing study, the responses of two young female harbor seals to tonal signals between 0.125 and 100 kHz were filmed. Frame-by-frame analysis was used to quantify RT (the time between the onset of the sound stimulus and the onset of movement of the seal away from the listening station). Near-threshold equal-latency contours, as surrogates for equal-loudness contours, were estimated from RT-level functions fitted to mean RT data. The closer the received sound pressure level was to the 50% detection hearing threshold, the more slowly the animals reacted to the signal (RT range: 188-982 ms). Equal-latency contours were calculated relative to the RTs shown by each seal at sound levels of 0, 10, and 20 dB above the detection threshold at 1 kHz. Fifty percent detection thresholds are obtained with well-trained subjects actively listening for faint familiar sounds. When calculating audibility ranges of sounds for harbor seals in nature, it may be appropriate to consider levels 20 dB above this threshold.

Measuring sound detection and reaction time in infant and toddler cochlear implant recipients using an observer-based procedure: a first report

OBJECTIVES

First, to establish the feasibility of the observer-based psychophysical procedure (OPP) in measuring sound detection in infant and toddler cochlear implant (CI) recipients. Second, to measure the psychometric function for detection (PFD) from individual subjects. Third, to determine whether reaction time (RT) provides information about the auditory sensitivity of young CI users.

DESIGN

Twelve CI recipients, 11 to 32 mo old, participated in our study. Initially, tones were presented in sound field, and children learned to respond when they heard tones but not at other times. Once an 80% correct criterion was met in sound field, a novel stimulation paradigm was used to present stimuli to a single electrode while the child listened to acoustic input on most other electrodes using their usual map. The PFD and RT were measured using this single-electrode stimulation paradigm.

RESULTS

Eleven subjects met criterion, 6 within the minimum possible number of trials. For eight subjects, the asymptotic level of detecting single-electrode stimuli averaged 86% correct, similar to levels achieved by normal-hearing infants and toddlers detecting pure tones. The PFD slope of infant and toddler CI recipients was less than or equal to the slope for adult CI users reported in previous studies. RT decreased significantly with stimulus level in four children.

CONCLUSIONS

These preliminary results suggest that psychophysical detection data can be obtained from infant and toddler CI recipients using OPP. The PFD of young CI users may be shallower than that of adult CI users. Relatively good asymptotic detection performance implies that young CI users are more attentive to sound than has been suggested in previous studies. RT tended to be a less reliable measure of detection, but methodological changes could improve its utility.

Multiband product rule and consonant identification

The multiband product rule, also known as band-independence, is a basic assumption of articulation index and its extension, the speech intelligibility index. Previously Fletcher showed its validity for a balanced mix of 20% consonant-vowel (CV), 20% vowel-consonant (VC), and 60% consonant-vowel-consonant (CVC) sounds. This study repeats Miller and Nicely's version of the hi-/lo-pass experiment with minor changes to study band-independence for the 16 Miller-Nicely consonants. The cut-off frequencies are chosen such that the basilar membrane is evenly divided into 12 segments from 250 to 8000 Hz with the high-pass and low-pass filters sharing the same six cut-off frequencies in the middle. Results show that the multiband product rule is statistically valid for consonants on average. It also applies to subgroups of consonants, such as stops and fricatives, which are characterized by a flat distribution of speech cues along the frequency. It fails for individual consonants.

Effects of the task of categorizing FM direction on auditory evoked magnetic fields in the human auditory cortex

We examined effects of the task of categorizing linear frequency-modulated (FM) sweeps into rising and falling on auditory evoked magnetic fields (AEFs) from the human auditory cortex, recorded by means of whole-head magnetoencephalography. AEFs in this task condition were compared with those in a passive condition where subjects had been asked to just passively listen to the same stimulus material. We found that the M100-peak latency was significantly shorter for the task condition than for the passive condition in the left but not in the right hemisphere. Furthermore, the M100-peak latency was significantly shorter in the right than in the left hemisphere for the passive and the task conditions. In contrast, the M100-peak amplitude did not differ significantly between conditions, nor between hemispheres. We also analyzed the activation strength derived from the integral of the absolute magnetic field over constant time windows between stimulus onset and 260 ms. We isolated an early, narrow time range between about 60 ms and 80 ms that showed larger values in the task condition, most prominently in the right hemisphere. These results add to other imaging and lesion studies which suggest a specific role of the right auditory cortex in identifying FM sweep direction and thus in categorizing FM sweeps into rising and falling.

First spike latency and spike count as functions of tone amplitude and frequency in the inferior colliculus of mice

Spike counts (SC) or, spike rate and first spike latency (FSL), are both used to evaluate the responses of neurons to amplitudes and frequencies of acoustic stimuli. However, it is unclear which one is more suitable as a parameter for evaluating the responses of neurons to acoustic amplitudes and frequencies, since systematic comparisons between SC and FSL tuned to different amplitudes and frequencies, are scarce. This study systematically compared the precision and stability (i.e., the resolution and the coefficient variation, CV) of SC- and FSL-function as frequencies and amplitudes in the inferior colliculus of mice. The results showed that: (1) the SC-amplitude functions were of diverse shape (monotonic, nonmonotonic and saturated) whereas the FSL-amplitude functions were in close registration, in which FSL decreased with the increase of amplitude and no paradoxical (an increase in FSL with increasing amplitude) or constant (an independence of FSL on amplitude) neuron was observed; (2) the discriminability (resolution) of differences in amplitude and frequency based on FSL are higher than those based on SC; (3) the CVs of FSL for low amplitude stimuli were smaller than those of SC; (4) the fraction of neurons for which BF=CF (within +/-500Hz) obtained from FSL was higher than that from SC at any amplitude of sound. Therefore, SC and FSL may vary, independent from each other and represent different parameters of an acoustic stimulus, but FSL with its precision and stability appears to be a better parameter than SC in evaluation of the response of a neuron to frequency and amplitude in mouse inferior colliculus.

Correcting for false alarms in a simple reaction time task

Simple reaction times (RTs) constitute an important source of information and tool in human and animal psychophysics, in cognitive neuroscience, and in the clinic. We measure simple RTs to auditory stimuli, in a high signal rate vigilance design, to examine the possibility that simple RT may be used as a tool to study mechanisms of temporal summation at absolute threshold. By means of catch trials, we monitor the subjects' tendencies to produce false alarms, that is reactions not controlled by the reaction stimulus. Here we examine the possibility that a model of a race between a stimulus-controlled reaction process and a false alarm process can account for the numbers as well as for the timing of early and late RTs on stimulus trials in our experiments. We show that the responses in both tails of our empirical RT distributions on stimulus trials are indeed correctly predicted by the race model and can be considered false alarms. This suggests that the race model might also provide a reasonable description of the way in which false alarms and stimulus-controlled reactions combine when they overlap in time. We examine the magnitudes of the estimated effects of false alarms on several parameters of the RT distributions by application of the race model. The analyses reveal that the effects not only vary with false alarm rate but also with stimulus parameters. Our data suggest that the race model may provide a theoretically reasonable and easy means of correcting for false alarms in simple RT paradigms and thus may constitute a useful alternative to the common practice of truncation.