{"id":2080,"date":"2019-01-09T13:25:57","date_gmt":"2019-01-09T18:25:57","guid":{"rendered":"https:\/\/creatureandcreator.ca\/?page_id=2080"},"modified":"2019-03-07T15:25:30","modified_gmt":"2019-03-07T20:25:30","slug":"hearing","status":"publish","type":"page","link":"https:\/\/creatureandcreator.ca\/?page_id=2080","title":{"rendered":"Objective Audiometry"},"content":{"rendered":"\n

This page derives from a presentation on The Objective Evaluation of Human Hearing<\/em> in March 2019 at the meeting of the Deutsche Gesellschaft f\u00fcr Audiologie in Heidelberg. The notes<\/a> for the presentation and the actual PowerPoint file<\/a> are available to download. The sound samples are not directly accessible through this page but are available in the PowerPoint file. Some general references for this talk are Picton (2011) and Hoth et al. (2014). <\/p>\n\n\n\n

I. Audiometry<\/strong><\/p>\n\n\n\n

Subjective audiometry has two main parts: the pure tone audiogram and the word recognition score. Both are “subjective” in that they require the subject to make a conscious response \u2013 to press a button when a tone has been heard or to repeat a word that was just presented. We can often predict the word recognition score from the pure tone thresholds. This has led many to disregard the results of speech perception studies. Hearing aids are usually fit mainly on the basis of the pure tone audiogram. Speech perception is given little attention. One of the goals of this presentation is to promote the testing of speech. <\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

a) Objective Audiometry<\/strong><\/p>\n\n\n\n

Subjective audiometry requires both a conscious response from the person being tested, and an interpretation of this response by the audiologist performing the test. “Objective” audiometry does not require a conscious response from the patient. I shall propose that fully objective audiometry should also not require any conscious interpretation of the results by the audiologist. Present techniques of objective audiometry can provide good estimates of the pure tone audiogram. However, there is as yet no objective measurement of speech perception.<\/p>\n\n\n\n

Several objective responses to sound can be recorded without the subject having to make a decision and provide a behavioral response:<\/p>\n\n\n\n

  1. changes in the heart rate<\/li>
  2. reflex movements in the body<\/li>
  3. middle ear muscle reflexes<\/li>
  4. otoacoustic emissions<\/li>
  5. electrical responses from the\near or brain<\/li>
  6. magnetic responses from the\nbrain<\/li>
  7. changes in cerebral blood flow\nas measured with magnetic resonance imaging (fMRI) or event-related optical\nsignals (near-infra-red spectroscopy NIRS)<\/li><\/ol>\n\n\n\n

    I\nshall limit myself to the electrical measurements:<\/p>\n\n\n\n

    \"\"<\/a><\/figure>\n\n\n\n

    The\nillustration (Picton, 2013) shows the electrical responses to sound categorized\nas transient or following responses. The upper half of the figure shows the\ntransient auditory evoked potentials plotted on three different time scales \u2013\nfast, middle and slow. The waveforms represent the typical response to 70 dB\nnHL clicks presented at a rate of 1\/s. The lower half of the figure represents\nthe following responses recorded to amplitude-modulated noise presented at 60\ndB SPL when the rate of modulation is varied. These following responses are\nplotted on three different frequency scales with the slower responses on the\nright. Though initially counter-intuitive, this arrangement of the axis allows\ncomparison between homologous transient and following responses. The peaks of\nthe slow following responses near 4, 10 and 20 Hz have been named \u03b8, \u03b1, and \u03b2\nafter the frequency-bands of the human EEG. Until recently most of the studies\nof the auditory following responses concentrated on the 40 Hz response (also\nknown as the \u03b3 response), the 80 Hz response and the fast Frequency Following\nResponse. Recently the slow following response (labelled \u03b8) has be shown to be\nimportant in following the speech envelope. <\/p>\n\n\n\n

    Objective hearing tests are useful in many situations:<\/p>\n\n\n\n

    1. subjects too young to provide reliable behavioral responses.<\/li>
    2. patients who are too emotionally disturbed or cognitively impaired to\nprovide reliable behavioral responses.<\/li>
    3. unconscious or anesthetized patients<\/li>
    4. patients who may be consciously or unconsciously exaggerating their\nhearing impairment.<\/li>
    5. evaluating patients with auditory neuropathy or auditory processing\ndisorders. <\/li>
    6. assessing the function of cochlear implants<\/li><\/ol>\n\n\n\n

      The most important of these applications, and the one\nI shall focus on, is the evaluation of hearing in infants and young children. Different\ncountries have different procedures for screening and diagnosis in the newborn\nperiod. This slide shows typical protocols:<\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      Well babies are screened using otoacoustic emissions. Because of the possibility of auditory neuropathy, babies in an intensive care unit are screened using auditory brainstem responses (ABRs). Babies who fail the screening tests or who may have developed hearing loss after birth are referred for diagnostic evaluation \u2013 the estimation of thresholds at different frequencies. The usual approach records auditory brainstem responses to brief tones, or more recently to narrowband chirps.  <\/p>\n\n\n\n

      b) Chirps<\/strong><\/p>\n\n\n\n

      The\nmajor change in objective audiometry in the last decade has been the use of\nchirps. CE chirps designed by Claus Elberling are constructed to compensate for\nthe frequency-related delays that occur in the cochlea (Elberling & Don,\n2010; Elberling & Crone-Esmann, 2017; G\u00f8tsche-Rasmussen\net al., 2012). The\ntravelling wave distributes the different frequencies of sound along the\nbasilar membrane with the high frequencies near the stapes occurring a few\nmilliseconds before the low frequencies near the apex. Chirps can be broad-band\n(like a click) or limited to an octave of frequencies (like a brief tone).<\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      A\ntransient stimulus contains multiple different frequencies. If the timings of\nthese different frequencies are adjusted so that the slower frequencies begin\nearlier than the faster frequencies, the neural responses will all occur\nsimultaneously rather than sequentially and the combined response will be\nlarger. Larger responses shorten the time to recognize whether a response is\npresent or not. <\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      Tonepips and narrow-band chirps sound similar and have\na similar acoustic specificity. However, as the preceding slide shows, the ABR\nevoked by a chirp is significantly larger than that evoked by a simple tonepip.\nIn general, the wave V amplitude is about 50% larger when using chirps. This\nmakes it faster to determine whether a response is present (Cobb & Stuart,\n2016abc; Ferm et al., 2013; Ferm et al., 2015, Rodrigues et al., 2013). <\/p>\n\n\n\n

      A\ndifferently designed low-frequency chirp might provide more accurate assessment\nof low-frequency hearing, particularly if used with masking (Balji\u0107 et al., 2017; Frank et\nal., 2017), but this stimulus is not\nwidely available:<\/p>\n\n\n\n

      c) Speech Perception<\/strong><\/p>\n\n\n\n

      In order to facilitate the development of speech and language, once a baby has been identified as hearing-impaired, a full audiological evaluation must be performed and treatment initiated (Scollie & Bagatto, 2010).<\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      These procedures work well to provide the baby with amplified sounds. However, they do not guarantee that such sounds can be discriminated one from another or that they are perceived as speech. Even when fitting hearing aids in adults, one usually does not assess how well speech is perceived through the aids. One simply assumes that the amplified sound is adequate for communication. The problem is that the measurement of speech perception is time-consuming even when behavioral methods are used. <\/p>\n\n\n\n

      This is unfortunate. Patients come to the audiologist\nbecause they cannot understand speech. They do not usually complain of\ndifficulties in hearing faint tones. We need to pay more attention to speech\nperception. <\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      This slide shows the acoustic characteristics of the recently developed International Speech Test Signal (Holube et al., 2010). This was constructed by piecing together snippets of speech in 6 different languages: English, Arabic, Chinese, French, German and Spanish. The signal does not make any sense. Played in an endless loop it is a good cure for insomnia or a demonstration of how speech sounds when one is aphasic. The ISTS signal and supporting documents are available at the website of the European Hearing Instrument Manufacturers Association<\/a>. <\/p>\n\n\n\n

      The spectrogram (left) demonstrates some of the temporal characteristics of speech \u2013 phonemes that last between tens of milliseconds, syllables that recur a rate of 3-5 per second, words that last around a second and phrases that last several seconds. The average speech spectrum (right) shows that most of the energy of speech is in the lower frequencies (Byrne et al., 1994). Not clear from these acoustic characteristics is the fact that many of the acoustic features that distinguish phonemes involve the middle and high frequencies.  <\/p>\n\n\n\n

      Response\nto tones and chirps can tell us about hearing thresholds but they do not tell\nus whether the subject can perceive speech. How might we determine objectively\nwhether a patient can discriminate between different sounds? We need to show\nthat the brain can discriminate changes in the auditory stimulus: <\/p>\n\n\n\n

      \"\"<\/a><\/figure>\n\n\n\n

      This\nslide (Figure 11-2 in Picton, 2011) shows the human cortical response to an\nauditory stimulus in different contexts. An occasional stimulus evokes a\nP1-N1-P2 complex. An occasional pause in an ongoing tone evokes a response to\nboth the offset of the tone and its re-onset. A brief change in the frequency\nof a continuous tone evokes a similar response to the onset of a brief tone. These\nare simple responses to simple changes. <\/p>\n\n\n\n

      Things\nget more complicated when multiple stimuli are involved (right side of figure).\nIf an occasional tone differs from the preceding tones, we record both a response\nto the onset and a later negative wave in response to the stimulus deviance \u2013\nthe mismatch negativity. The difference between this response and the response\nto an ongoing standard stimulus shows both the N1 of the onset-response and the\nMMN to the change. Also visible in these results is the fact that at rapid\nrates the N1 response to the standard is small whereas the P1 persists. As well\nas the N1 response to a standard being smaller than the response to the\ndeviant, and the response to the standard that follows the deviant is larger\nthan to a standard that follows another standard. The illustration is from <\/p>\n\n\n\n

      The\nmismatch negativity (MMN) does not require any subjective response on the part\nof the subject. The subject does not even have to attend to the stimuli. Furthermore\nthe MMN can be recorded in response to phonetic changes as well as simple\nacoustic changes. The major problem with using the MMN as an objective test is\nthat it is small and takes a long time to record. <\/p>\n\n\n\n

      \"\"<\/figure>\n\n\n\n

      A Mismatch Response can be recorded in response to a change in the phonetic characteristic of a stimulus. This slide from Angela Friederici et al. (2002) shows the response of young infants to a change in phoneme duration. In German, the duration of a vowel is meaningful. The mismatch response in the awake infant is similar to that in adults \u2013 a small negative wave, the mismatch negativity (MMN). However, in the sleeping infant the response is quite different \u2013 a large positive wave rather than a small negative wave. In this slide (as opposed to most of the other slides in this presentation) negative is plotted upward. We who study evoked potentials have often been accused of not knowing which way is up.<\/p>\n\n\n\n

      <\/p>\n\n\n\n

      Audiometry \u2013 Conclusions<\/strong><\/p>\n\n\n\n

      Chirps are good.<\/strong><\/p>\n\n\n\n

      We do not yet have an objective assessment of speech perception.<\/strong><\/p>\n\n\n\n

      <\/p>\n\n\n\n

      <\/p>\n\n\n\n

      II. Multiplicity<\/strong><\/p>\n\n\n\n

      Now that we have considered the need for objective\naudiometry we shall look at three different aspects of this process \u2013 the\nprinciple of multiplicity, methods for identifying thresholds, and evaluating\nspeech perception. <\/p>\n\n\n\n

      The principle of multiplicity entails that <\/p>\n\n\n\n

      1. we should record responses to multiple stimuli simultaneously, and <\/li>
      2. we should record multiple different responses to these stimuli.<\/li><\/ol>\n\n\n\n

        a) Multiple Simultaneous\nStimuli to Assess Pure Tone Thresholds<\/strong><\/p>\n\n\n\n

        Many\nyears ago, Otavio Lins and I (Lins & Picton, 1994; Lins et al., 1996) showed\nthat we could record brainstem responses to multiple amplitude-modulated tones\npresented simultaneously. An amplitude modulated tone evokes a steady state\nresponse at the frequency of the modulation. This can be identified as a peak in\nthe spectra of the responses on the right of the figure. Steady-state responses\nwere evoked by different tones, each modulated at a unique frequency Each response\nis recorded separately.<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        The\nleft side of the figure shows what can be recorded when we combine the stimuli\nand present them simultaneously. If you pay attention to the sound you can\nrecognize the individual stimuli in the mixture. The cochlea processes each of\nthe modulated stimuli in a separate region of the basilar membrane. <\/p>\n\n\n\n

        The\nlower left shows the complete spectrum of the response to the combined stimuli.\nAt the specific modulation-frequencies of each stimulus we can see a\nresponse.  The region around these\nfrequencies is amplified 2X in the inset. Each response can also be viewed on\nindividual polar plots (upper left) which show the phase as well as the\namplitude. <\/p>\n\n\n\n

        There\nis no difference in response-amplitude when the stimuli are presented\nsimultaneously from when they stimuli are presented singly. It is therefore\nmuch more efficient to record the responses to multiple stimuli presented concurrently\nthan to record each response separately. We can present four stimuli to the\nleft ear and four stimuli to the right ear. In the best of all possible worlds\nwe can get 8 thresholds in the same time that it takes to record one.<\/p>\n\n\n\n

        A major recent advance in the multiple stimulus technique has been to use chirps rather than amplitude modulated tones. The use of multiple chirp-trains instead of multiple amplitude modulated tone was initially proposed by Cebulla et al. (2012). These stimuli evoke larger responses and are more quickly detected. They can be recorded close to threshold in infants (upper right from Rodriguez et al., 2014) and the estimated thresholds correlate well with those obtained using simple tonepip-ABRs when each stimulus is presented singly (lower illustration from Sininger et al., 2018). Michel and J\u00f8rgensen (2017) have reported similar results.<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        The advantage is that multiple chirp ASSRs over\nrecording tone-pip ABRs is that the ASSRs can assess thresholds in much less\ntime (20 as opposed to 32 minutes according to Sininger et al., 2018). <\/p>\n\n\n\n

        My recommendation for the objective determination of\nhearing thresholds is therefore to record the steady-state responses to\nmultiple chirps presented simultaneously. This may best solve our need for\nestimating the pure tone audiogram. However it does not tell us that the\nsubject can discriminate the sounds or perceive speech. Remember audiometry has\ntwo parts \u2013 pure tone thresholds and speech perception. <\/p>\n\n\n\n

        b) Multiple-Deviant Paradigm to Record Mismatch Responses<\/strong><\/p>\n\n\n\n

        The idea of multiple simultaneous stimuli can be used to decrease the recording time for the mismatch negativity (MMN). The conventional MMN paradigm presents one type of deviant in a train of standards with a probability of 1\/10. To record another type of MMN we need to repeat the whole recording. A \u201cmultiple-deviant\u201d paradigm presents each deviant at the same temporal probability but presents five deviants \u2013 e.g., frequency, intensity, duration, frequency-glide and gap \u2013 concurrently (N\u00e4\u00e4t\u00e4nen et al., 2004). <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Five mismatch negativities for the price of one. The sound samples do not follow the illustration. Rather they give you a sense of what the stimuli sound like. We shall return to this paradigm later when we consider speech stimuli. <\/p>\n\n\n\n

        So now\nwe have two protocols for presenting multiple stimuli simultaneously – auditory\nsteady state responses to multiple chirps and mismatch responses to multiple\ndeviances.<\/p>\n\n\n\n

        c) Multiple Responses<\/strong><\/p>\n\n\n\n

        We\nshould also consider the idea of recording multiple responses to the same stimuli.\nWhere possible, we should record multiple responses \u2013 transient, sustained and\nsteady-state \u2013 evoked by the same stimuli. The following illustration (Figure 10.12 of Picton, 2011) shows the responses to the intermittent amplitude modulation of a continuous\ntone \u2013 at 40 Hz and at 80 Hz. The modulations alternate in their phase. So\naveraging the responses to the different modulations together cancels out the\nsteady-state response and leaves the transient and sustained potentials. The\nonset of the modulation evokes an onset response that consists mainly of a P1\nwave. Since the stimuli are presented rapidly, the N1 is very small. A\nsustained potential lasts through the duration of the modulation.  <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        If we\ntake the difference between the responses to the out-of-phase stimuli, we\ncancel the transient and sustained potentials and are left with the\nsteady-state response \u2013 which can be filtered and amplified. These recordings\nshow the build-up of the steady-state response over time. The 80 Hz response\noriginating in the brainstem builds up quickly. There is a slight glitch in the\nbuildup \u2013 probably related to the onset response to the start of the\nmodulation. The cortical 40 Hz response takes about 200 ms to reach steady\nstate.  <\/p>\n\n\n\n

        Frequency-following responses have long been evoked\nusing pure tones. Nina Kraus and her colleagues have been using a special\nspeech stimulus to evoke both transient and following responses (Johnson et\nal., 2005, 2008; Anderson et al., 2015; Bin Khamis et al., 2019). The syllable\n\/da\/ is truncated to just the first 40 ms and then repeated at 20\/s. This\nstimulus elicits both the auditory brainstem response to the onset of the stop\nconsonant and a frequency-following response to the beginning of the vowel. The\nlarge waves of the frequency-following response (D, E, F) occur about 7 ms\nafter the onset of each of the glottal pulses. They represent the pitch\nenvelope. The small waves in each glottal cycle represent the formant\nfrequencies. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Another\nrecently reported paradigm allows the recording of both the brainstem frequency\nfollowing response and the cortical transient response (Bidelman, 2015;\nBidelman et al., 2018). The stimulus is the vowel \/a\/. This is repeated rapidly\nto allow recording the brainstem frequency-following response, and then more\nslowly for recording the cortical onset response:<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        These last three illustrations are just examples.\nWhere possible we should record multiple responses to the same stimuli. That\nway we might see how the sound is processed in different parts of the auditory\nsystem.<\/p>\n\n\n\n

        <\/p>\n\n\n\n

        Conclusions – Multiplicity<\/strong><\/p>\n\n\n\n

        The principle of multiplicity is established for the\nstimuli.<\/strong><\/p>\n\n\n\n

        We still need to record multiple different responses to the same stimulus.<\/strong><\/p>\n\n\n\n

        <\/p>\n\n\n\n

        <\/p>\n\n\n\n

        III. Threshold Detection<\/strong><\/p>\n\n\n\n

        Objectivity means that the patient need not make a\nsubjective response. Another aspect of objectivity involves the interpretation\nof the response and the determination of physiological threshold. My proposal\nis that assessing the thresholds for auditory evoked potentials should be as\nobjective for the examiner as it is for the patient. <\/p>\n\n\n\n

        a) Statistical Tests<\/strong><\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Ronald Fisher was a brilliant mathematician who almost single-handedly established modern scientific statistics. The \u201cF-test\u201d is named after him. When he was right he was really right. And if you use the correct statistical tests you will be as confident in your interpretation as Fisher is in his photograph. You will be unbiased \u2013 \u201cobjective.\u201d The Fsp test for the auditory brainstem response (ABR) is described in Elberling & Don (1984), and the F-test in steady-state responses is discussed in Dobie and Wilson (1996). <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        This figure illustrates the use of an F-test to detect\na steady-state response. In the frequency domain the steady-state response shows\nup as a line at the frequency at which the stimuli were presented. The activity\nat the other frequencies in the spectrum represents the unaveraged background\nEEG. The F-test compares the amplitude at the stimulus frequency to that at\nadjacent frequencies. If the stimulus amplitude is significantly larger than\nwhat would be expected by chance (left example) we can conclude that there is a\nresponse. If not (right example), we can only conclude that the response is not\nyet greater than the background EEG. We need another criterion \u2013 a low level of\nthe background noise \u2013 to decide that that response is not present. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        In the frequency domain, responses to chirps occur both at the rate of stimulation and at harmonics of this rate. Depending on the subject and the stimulus-rate, some harmonics may be more prominent than others. Combining the statistical tests for the responses at each of the multiple harmonics can give a more reliable test of whether or not a response is present in the recording than simply looking at the response at the fundamental frequency (Cebulla et al., 2006). <\/p>\n\n\n\n

        b) Threshold-Seeking\nAlgorithms<\/strong><\/p>\n\n\n\n

        Once we have a reliable technique for determining whether a response is\npresent or not, we can then determine the threshold for the response. Thresholds\ncan be estimated using fixed or adaptive methods. The fixed method records\nresponses at multiple levels above and below threshold and records the probability\nof detection as a psychometric function. In an adaptive technique the intensity\npresented on any one trial depends on the responses to preceding trials (Taylor\n& Creelman, 1967). For example the intensity can be decreased by 10 dB\nafter a positive response and increased by 5 dB after a negative response:<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        The above diagram represents the usual approach to\nestimating thresholds in behavioral audiometry (American Speech-Language-Hearing Association, 2005). Initially the subject is familiarized with the stimulus so that he or\nshe knows what to listen for. The tester then starts below threshold and\nascends in 5 dB steps until the subject hears the tone. The stimulus is then\ndecreased by 10 dB and the ascent begun again. This down-10-up-5-dB sequence is\nrepeated until a positive response occurs at one level in two out of three trials.\nThis level is then considered threshold. <\/p>\n\n\n\n

        This protocol works well when recording behavioral\nresponses. There is little difference in time between a \u201cyes\u201d response and a \u201cno\u201d\nresponse. In recording evoked potentials this approach is not efficient. The\nproblem is that it takes much longer to decide that an electrophysiological\nresponse is absent than it takes to decide that the response is present. In the\nexample provided, threshold is determined after 3 positive responses and 6\nnegative responses. To decide that a response is not present we must continue\nthe averaging until the background EEG noise is below some low criterion. Testing\ncan take a long time if the protocol requires more negative than positive\nresponses. Therefore when estimating thresholds for evoked potentials we\nusually use a descending technique. Then we only need to spend a long recording\ntime at the end when we have gone below threshold and have to determine that no\nresponses are present. The following set of data (Figure 10-14 from Picton,\n2011) illustrates the processes of recognizing responses (arrows) and\ndetermining threshold.  <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        When determining thresholds we need to consider the\npossibility of false positive responses. Using a test criterion of p=0.05,\nabout 1 in 20 responses could be \u201cdetected\u201d even if there were actually no\nresponses present. This is the nature of the statistical test. In the present\nset of recordings two responses out of 61 were considered to be false positives\n\u2013 designated by the ?<\/strong> <\/p>\n\n\n\n

        We also need to consider the possibility that a\nresponse might be missed. Perhaps the noise in one recording was particularly\nhigh. Thus in the illustration we might consider the 1 kHz  response at 30 dB in the 1.6 minute recording\nto have been inadvertently missed because there is a response at 20 dB. Or\nperhaps the response at 20 dB is another false positive. Decisions! Decisions!<\/p>\n\n\n\n

        A major question is whether we should continue to recording to a set time \u2013 say 10 minutes \u2013 or whether we can stop as soon as all responses are recognized \u2013 e.g. after 1.6 minutes. At 70 dB SPL all responses are recognized after 1.6 minutes of averaging. There would be no need to continue longer at this intensity. <\/p>\n\n\n\n

        To save time, can we just stop the recording once a\nresponse is detected? Can we watch the ongoing response, monitor the\nsignificance level and then stop when criterion is reached? The problem with\nthis approach is that the more times we test for significance the greater the\nprobability that we might reach criterion level by chance. A significance level\nof p<0.05 means that we would judge a response significant 1 in 20 times\neven though there was no response present. If we check the responses more than\n20 times we should expect some false detections.<\/p>\n\n\n\n

        We can compensate for this using various statistical corrections. A simple Bonferroni correction divides the probability criterion by the number of tests. The following illustration (Figure 6-14 from Picton, 2011) shows the corrected criterion when monitoring the significance level every minute or every 4 seconds. The correction is more severe when the monitoring is more frequent. The response is to a 1 kHz amplitude modulated tone at low intensity (25 dB SPL) in two different subjects. In the upper subject the response becomes significant in 2 minutes using 1 minute monitoring and in 5.2 minutes using 4-second monitoring. In the lower subject the response never reaches significance using the 4-second monitoring protocol. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Cebulla and St\u00fcrzebecher (2015) have recently proposed protocols wherein the interval between the statistical tests increases over time. If the response is not significant the protocol waits longer before testing again.<\/p>\n\n\n\n

        c) Physiological and Behavioral Thresholds<\/strong><\/p>\n\n\n\n

        The lowest intensity at which an evoked potential is\ndetected is the \u201cphysiological threshold.\u201d Since this is higher than the\nbehavioral threshold, we normally subtract a \u201ccorrection factor\u201d \u2013 the mean\nphysiological-behavioral difference \u2013 to obtain an \u201cestimated behavioral\nthreshold.\u201d<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        It is important to note the standard deviation (SD) of\nthe correction factor since this will determine the range for the actual\nbehavioral threshold. One normally assumes that the range is \u00b12SD \u2013 this would\ninclude 95% of the possible values. The SD of the correction factors has\nusually been around 10 dB. This means that we could be out in our estimated\nbehavioral threshold by up to 20 dB in either direction. In the above example\nthe hypothetical patient could actually have either a flat audiogram or a\nlimited high-frequency loss (Figure 6-15 from Picton, 2011). The most recent\nreports for multiple auditory steady state responses with chirps (e.g.,\nSininger et al, 2018) give SDs of 5-8 dB. These are much better.  <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Correction factors decrease as the threshold intensity\nincreases. This is due to the response amplitude (like the perceived intensity)\nincreasing more quickly above threshold when there is a hearing loss (upper\nhalf of the above illustration). The lower table gives sample correction\nfactors for the 40 Hz steady state responses to chirps: 10 to 15 dB at low\nintensities and 0 dB at high. <\/p>\n\n\n\n

        At present most systems using multiple steady state\nresponses present all the stimuli at one intensity. In some systems one can\nseparately adjust the intensity of each stimulus. Once a significant response\nhas been detected, the intensity can be lowered. Or if no response has been\ndetected and the background noise has reached a criterion level, the test may\nbe stopped or the intensity increased. These ideas were initially examined by\nRoland Muehler and his colleagues (2012). The only caution in this approach is\nthat the difference in intensities between stimuli should likely not exceed 30\ndB so that there is no masking of one stimulus by the others (John et al.,\n2002). The following illustration suggests how this might occur in assessing a\nhigh-frequency hearing loss:<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        What is needed is for this procedure to be made\ncompletely automatic \u2013 for the system\u2019s algorithms (rather than the\naudiologist) to decide when to change the intensity as well as to decide\nwhether a response is present or not. Then we would have completely objective\naudiometry \u2013 objective on the part of the patient and objective on the part of\nthe examiner. <\/p>\n\n\n\n

        d) Sweep Techniques<\/strong><\/p>\n\n\n\n

        The following\nfigure illustrates the idea of using sweeps of intensity to evaluate thresholds\n(Picton et al., 2007). It shows shows the following response to a 2 kHz stimulus amplitude modulated at\n92 Hz that over 16 seconds ascends in intensity from 25 dB SPL to 75 dB and then\ndecreases back to 25 dB. In the upper left is shown the spectrogram \u2013 the\nresponse can be seen coming out of the background EEG noise near the beginning\nof the sweep, increasing in amplitude and then decreasing until it is no longer\nrecognizable. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        The\ngraphs demonstrate the effects of intensity on the amplitude and phase of the\nresponse. The advantage of the sweep technique is that you can use the recorded\namplitudes and phases to plot an amplitude-intensity function and extrapolate\nto threshold.<\/p>\n\n\n\n

        My dream is that the algorithms for assessing thresholds will sweep back and forth over and above threshold, narrowing the range of intensity until the threshold is accurately and automatically bracketed. One of the dreams of an old man. <\/p>\n\n\n\n

        <\/p>\n\n\n\n

        Conclusions \u2013\nThreshold Detection<\/strong><\/strong><\/p>\n\n\n\n

        We have good techniques to monitor the responses so\nthat we can recognize them quickly and accurately.<\/strong><\/p>\n\n\n\n

        What we still need are some automatic threshold-seeking algorithms that can vary the intensity among the different stimuli.<\/strong><\/p>\n\n\n\n

        <\/p>\n\n\n\n

        <\/p>\n\n\n\n

        IV. Speech Perception<\/strong><\/p>\n\n\n\n

        Now we turn to the final topic\u2013 the objective\nassessment of speech perception. In recent years there has been a revolution in\nauditory science (Hamilton & Huth, 2018). Rather than using clicks and\ntones to examine how the auditory system works, scientists are using speech\nstimuli. This makes sense. Human auditory systems have evolved to process\nspeech \u2013 not clicks and tones. <\/p>\n\n\n\n

        a) Assessing Hearing Impairment and Fitting Hearing Aids<\/strong><\/p>\n\n\n\n

        In audiology we should move \u201cbeyond the audiogram\u201d and\npay more attention to speech perception (Fabry 2015). Our patients come to us\ncomplaining that they cannot understand when others speak to them. They do not\ncomplain that they are unable to hear clicks and tones. We should evaluate how\nwell they perceive speech \u2013 in quiet and in noise (Dubno, 2018; Soli et al,\n2018ab). <\/p>\n\n\n\n

        Speech stimuli can be particularly important in\naudiology when fitting hearing aids to compensate for a hearing impairment. The\ngoal is to amplify the speech signal and compress it to fit within the range of\nwhat is comfortably audible:<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        In objective audiometry we can derive the unaided\nthresholds from chirp-evoked auditory steady state responses. We can also\nevaluate the responses of the brain to aided stimuli (Laugesen, et al., 2018;\nPicton et al., 1998). However, hearing aids were not designed to amplify chirps\nor tones. They are set up to amplify speech. They would be better evaluated\nusing speech sounds. <\/p>\n\n\n\n

        b) Frequency Following Responses<\/strong><\/p>\n\n\n\n

        A simple way to assess the brain\u2019s ability to process speech sounds is to record frequency- and envelope-following responses to simple speech stimuli like vowels (Aiken & Picton, 2008a; Krishnan, 2007, 2017; Ribas-Prats et al, 2019). If these speech stimuli are accurately reproduced by the brainstem we can be sure that the stimuli have been properly processed in the cochlea and are available to be discriminated.  <\/p>\n\n\n\n

        One\nfascinating aspect of the speech FFRs concern the ability of adults and infants\nto follow the pitch changes that are phonetically meaningful in tonal languages\nlike Chinese (Krishnan et al., 2004; Jeng et al., 2016).<\/p>\n\n\n\n

        Another\ninteresting aspect of using speech stimuli is that we can listen to the\nresponse played back into a speaker and recognize easily when the response is\nthere (Bidelman, 2018a). The examples illustrate early in the recording and\nlate in the recording. As the recording progresses you can hear the \/a\/ sound\nin the background noise: <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Viji Easwar and her colleagues have been considering a\nstimulus that uses several different speech sounds as a way to evaluate the\nability of a hearing aid to transfer speech information to the brain (Easwar et\nal., 2015ab). The frequencies in this stimulus cover the range of the long term\nspeech spectrum. The idea is to evaluate envelope- and frequency-following\nresponse to these amplified sounds to assess how well a hearing aid makes\nspeech audible. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        One\ndifficulty with these recordings is that several different regions of the\nauditory pathway from cochlea to cortex generate frequency-following responses\nand these have different phase relations to the stimulus (Bidelman, 2018b;\nEaswar et al., 2018; King et al., 2016). Sometimes the fields from different\ngenerators may be out of phase and thus cancel themselves. <\/p>\n\n\n\n

        c) Following the Speech Envelope<\/strong><\/p>\n\n\n\n

        The brainstem frequency following response tracks the\nenvelope of speech sounds, portraying how the pitch and the formants change over\ntime. This can give the phonetic structure of speech. Another aspect of speech\nis the overall speech envelope – how the amplitude changes over time at rates\nof 1-10 Hz. Shannon and his colleagues (1995) demonstrated how important this\nis to the recognition of ongoing speech. Magnetoencephalographic data indicated\nthat the human auditory cortex follows this speech envelope (Ahissar et al.\n2001). <\/p>\n\n\n\n

        Aiken and I (2008b) decided to record\nelectroencephalographic responses to the speech envelope of several sentences\nsuch as \u201cTo find the body they had to drain the (lake).\u201d The sentence comes\nfrom a detective story. The last word is missing since we had to ensure that\nthe subjects attended to the sentence when it was repeated. We ended the\nsentence either with the correct word or with an incongruous word. <\/p>\n\n\n\n

        If you take the envelope of a spoken sentence, and\nfill it with octave-band filtered versions of the original sentence, you can\nlikely still understand it even though it has no pitch and little frequency\ninformation. It sounds like whispering. If you fill the envelope with white\nnoise, the rhythm of the sentence is still preserved. The human brain likely\nuses the overall amplitude of the speech envelope to divide the speech up into\nrecognizable chunks \u2013 syllables and words.<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Basically we presented 6 different sentences, averaged\nthe response to each sentence over multiple repetitions, and derived source\nwaveforms from the left and right auditory cortex. We then correlated the\nsource waveform to the speech envelope, using variable delays. The correlation\nwas highest when the speech envelope occurred between 100 and 200 ms before the\nsource waveform (red regions). <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        We could then average the correlations over the period\nof the sentence to give a \u201ctransfer function.\u201d This was significantly greater\nthan what would occur by chance (shaded area). The human auditory cortex\nfollowed the speech envelope with a delay between 100 and 200\nmilliseconds.  As I speak your auditory\ncortex is following me.<\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        d) Development of Speech and Language<\/strong><\/p>\n\n\n\n

        However, following the speech envelope does not mean that we can\nunderstand what is being said. For that we need to discriminate the different\nspeech sounds. One way to see whether the brain recognizes the difference\nbetween speech sounds is to record the mismatch response to a phonetic\nchange.  We mentioned this earlier in the\ntalk \u2013 remember ba\u2026.ba\u2026.baa<\/em>\u2026.ba.  Unfortunately it takes a long time to record\nthe mismatch response to a single phonetic change.<\/p>\n\n\n\n

        Recently, Partanen and his colleagues (2013) have shown that many different mismatch responses may be recorded concurrently to different speech sounds in young infants. Five different mismatches were run concurrently. The actual sound sample includes only the consonant change \/te\/ to \/ti\/, the intensity change and the change in vowel duration. The recordings show significant mismatch responses to the phonetic changes. Again negativity is plotted upward in this illustration. For these infants the Mismatch Response was a slow positive wave. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        A major problem\nremains about using the mismatch response in infancy and childhood. We need to\nfigure out when the infant response is positive and when negative. How does\nthis depend on the state of arousal and the age of the child? And we need to\ntrack the development of the mismatch response over childhood. <\/p>\n\n\n\n

        Angela Friederici (2005) has proposed that infants\nshow a progressive development of language capabilities: the discrimination of\nphonemes in the language to which they have been exposed, the recognition of prosodic\nfeatures, the identification of words, and finally the structuring of syntax. <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        This development can be followed using the auditory\nevoked potentials. I have illustrated the mismatch negativity (MMN). The CPS \u2013\nclosure positive shift \u2013 occurs at the end of sentences or phrases. The N400\noccurs when the words don\u2019t make sense. The ELAN early left anterior negativity\n\u2013 and the P600 \u2013 occur with grammatical errors. We need to set up standard\nparadigms to assess the development of speech and language in infants and young\nchildren. <\/p>\n\n\n\n

        e) Multivariate\nTemporal Response Functions<\/strong><\/p>\n\n\n\n

        So now for another dream. Recently sophisticated\nmathematical techniques have been developed to determine the ability of the\nhuman brain to follow speech \u2013 most particularly \u201cmultivariate temporal\nresponse functions\u201d (mTRF). These techniques can demonstrate EEG entrainment to\nthe speech envelope, and also to the phonetic changes that are occurring during\nthe ongoing speech (Ding\n& Simon, 2014; Lalor & Foxe, 2010). The degree of\ncorrelation determined using these techniques can relate to the ability of the\nsubject to recognize speech (Vanthornhout et al., 2018). <\/p>\n\n\n\n

        \"\"<\/a><\/figure>\n\n\n\n

        Recent years have seen a plethora of papers using these new statistical techniques to measure how the brain follows various aspects of the speech signal (Brodbeck et al., 2018; Di Liberto et al., 2015, 2017; Kalashnikova et al., 2018;.Trng et al, 2017, 2018; W\u00f6stmann et al, 2017). With more research we might be able to present natural ongoing speech to a patient and obtain an objective measure of speech perception. <\/p>\n\n\n\n

        <\/p>\n\n\n\n

        Conclusions \u2013 Speech Perception<\/strong><\/strong><\/p>\n\n\n\n

        Following responses can evaluate how speech is being\nprocessed in the brain.<\/strong><\/p>\n\n\n\n

        What we still need is some objective test of speech discrimination. I have mentioned the MMN. We also need to study the development of other late evoked potentials that process speech and language.<\/strong><\/p>\n\n\n\n

        <\/p>\n\n\n\n

        <\/p>\n\n\n\n

        Summary<\/strong><\/p>\n\n\n\n

        I have briefly reviewed the need for objective\naudiometry \u2013 for both pure tone thresholds and for speech perception. I have\nproposed the principle of multiplicity \u2013 many stimuli at the same time, and\nmany different responses for each stimulus. I discussed how thresholds can be\nobjectively recognized, and proposed automatic algorithms to reach threshold \u2013\nour testing should be as objective for the examiner as for the patient. Finally\nI suggested some ways in which we can assess speech perception \u2013 using following\nresponses and the mismatch response. <\/p>\n\n\n\n

        In objective audiometry, I long for the day when pure\ntone thresholds are accurately and quickly assessed without any intervention on\nmy part, and when I can speak to a patient and see through his\nelectroencephalogram whether he or she is following what I am saying. I am getting old and it is the prerogative of the\nelderly to dream. <\/p>\n\n\n\n

        Acknowledgments<\/strong><\/p>\n\n\n\n

        My sincere thanks to colleagues who provided me with\nsounds, papers, ideas and advice: Steve Aiken, Gavin Bidelman, Mario Cebulla,\nAndrew Dmitrijevic, Andr\u00e9e Durieux-Smith, Viji Easwar, Claus Elberling, Angela\nFriederici, Sasha John, Nina Kraus, Otavio Lins, Roland Muehler, Eino Partanen,\nMarilyn Perez-Abalo, David Purcell, Yvonne Sininger, Susan Small, David\nStapells, Andy Stuart, Martin Walger, Jan Wouters. <\/p>\n\n\n\n

        References<\/strong><\/p>\n\n\n\n

        Ahissar, E., Nagarajan, S., Ahissar, M., Protopapas,\nA., Mahncke, H., Merzenich, M.M. (2001) Speech comprehension is correlated with\ntemporal response patterns recorded from auditory cortex. Proceedings of the\nNational Academy of the Sciences United States of America, <\/em>98, 13367-13372.\n<\/p>\n\n\n\n

        Aiken, S.J., & Picton, T.W. (2008a). Envelope and\nspectral frequency following responses to vowel sounds. Hearing Research<\/em>,\n245, 35-47.<\/p>\n\n\n\n

        Aiken, S.J., & Picton, T.W. (2008b). Human\ncortical responses to the speech envelope. Ear and Hearing<\/em>, 15, 139-157.<\/p>\n\n\n\n

        American Speech-Language-Hearing Association. (2005). Guidelines for manual pure-tone threshold audiometry<\/a>. <\/p>\n\n\n\n

        Anderson,\nS., Parbery-Clark, A., White-Schwoch, T., & Kraus, N.\n(2015). Development of subcortical speech representation in human infants. Journal of the Acoustical Society of America<\/em>. 137(6), 3346-3355.<\/p>\n\n\n\n

        Balji\u0107, I., E\u00dfer, D., Foerst, A. & Walger, M. (2017). Evaluation of\noptimal masking levels in\nplace-specific low-frequency chirp-evoked auditory brainstem responses. Journal of the Acoustical Society of America<\/em>,\n141, 197\u2013206. <\/p>\n\n\n\n

        Bidelman,\nG. M. (2015). Towards an optimal paradigm for simultaneously recording cortical\nand brainstem auditory evoked potentials. Journal of <\/em>Neuroscience Methods, <\/em>241, 94\u2013100.<\/p>\n\n\n\n

        Bidelman,\nG. M. (2018a). Sonification of scalp-recorded frequency-following\nresponses (FFRs) offers improved response detection over conventional\nstatistical metrics.  Journal of Neuroscience Methods, <\/em>293,\n59-66.<\/p>\n\n\n\n

        Bidelman,\nG. M. (2018b). Subcortical sources dominate the neuroelectric auditory\nfrequency-following response to speech. NeuroImage<\/em>, 175, 56-69.<\/p>\n\n\n\n

        Bidelman,\nG. M., Pousson, M., Dugas, C., & Fehrenbach, A. (2018). Test-retest\nreliability of dual-recorded brainstem vs. cortical auditory evoked potentials\nto speech. Journal of the American Academy of <\/em>Audiology, <\/em>29, 164\u2013174.<\/p>\n\n\n\n

        BinKhamis G, L\u00e9ger A, Bell SL, Prendergast G,\nO’Driscoll M, & Kluk K. (2019, in press<\/em>) Speech auditory brainstem responses: effects of\nbackground, stimulus duration, consonant\u2013vowel, and number of epochs. Ear & Hearing.<\/em><\/p>\n\n\n\n

        Brodbeck,\nC., Presacco, A., & Simon, J.Z. (2018). Neural source dynamics of\nbrain responses to continuous stimuli: Speech processing from acoustics to\ncomprehension. NeuroImage<\/em>, 172:162-174. <\/p>\n\n\n\n

        Byrne, D., Dillon, H., Tran, K., Arlinger, S.,\nWibraham, K., Cox, R., Hagerman, B., Hetu, R., Kei, J., Lui, C., Kiessling, J.,\nKotby, M., Nasser, N., El Kholy, W., Nakanishi, Y., Oyer, H., Powell, R.,\nStephens, D., Meredith, R., Sirimanna, T., Tavatkiladze, G., Frolenkov, G.,\nWesterman, S. & Ludvigsen, C (1994). An international comparison of\nlong-term average speech spectra. Journal of the Acoustical Society of\nAmerica<\/em>, 96 2108-2120. <\/p>\n\n\n\n

        Cebulla, M., & E. Sturzebecher, E. (2015).\nAutomated auditory response detection: Further improvement of the statistical\ntest strategy by using progressive test steps of iteration. International\njournal of audiology,<\/em> 54,<\/strong> 568-572.<\/p>\n\n\n\n

        Cebulla, M., St\u00fcrzebecher, E., Don, M.,\n& M\u00fcller-Mazzotta J. (2012). Auditory brainstem response recording to multiple interleaved broadband\nchirps. Ear & Hearing<\/em>, 33, 466-479.  <\/p>\n\n\n\n

        Cebulla,\nM., St\u00fcrzebecher, E., & Elberling, C. (2006). Objective detection of\nauditory steady-state responses: Comparison of one-sample and q-sample tests. Journal\nof the American Academy of Audiology<\/em>, 17<\/em>, 93\u2013103.<\/p>\n\n\n\n

        Cobb,\nK.M., & Stuart, A. (2016a). Neonate auditory brainstem responses to\nCE-chirp and CE-chirp octave band stimuli II: versus adult auditory brainstem\nresponses. Ear and Hearing, 37, <\/em>724-743. <\/p>\n\n\n\n

        Cobb,\nK.M., & Stuart, A. (2016b). Neonate auditory brainstem responses to\nCE-chirp and CE-chirp octave band stimuli I: versus click and tone burst\nstimuli. Ear and Hearing, 37, <\/em>710-723. <\/p>\n\n\n\n

        Cobb,\nK.M., & Stuart, A. (2016c). Auditory brainstem response thresholds to air-\nand bone-conducted CE-Chirps in neonates and adults. Journal of Speech,\nLanguage, and Hearing Research, <\/em>59, <\/em>853-859.<\/p>\n\n\n\n

        Ding, N., & Simon, J. Z. (2014). Cortical entrainment to continuous speech: functional roles and interpretations.<\/a> Frontiers in Human Neuroscience<\/em>, 8, 311. <\/p>\n\n\n\n

        Di Liberto, G.M., O’Sullivan, J.A., & Lalor, E.C.\n(2015). Low-frequency cortical entrainment to speech reflects phoneme-level processing. Current\nBiology<\/em>, 25, 2457\u20132465.<\/p>\n\n\n\n

        Di\nLiberto, G. & Lalor, E. (2017). Indexing cortical entrainment to natural\nspeech at the phonemic level: Methodological considerations for applied\nresearch. Hearing Research<\/em>, 348, 70-77. <\/p>\n\n\n\n

        Dobie, R. A., & Wilson, M. J. (1996). A comparison\nof t test, F test, and coherence methods of detecting steady-state\nauditory-evoked potentials, distortion-product otoacoustic emissions, or other\nsinusoids. Journal of the Acoustical Society of America,<\/em> 100<\/em>,\n2236-2246. <\/p>\n\n\n\n

        Dubno, J. R. (2018). Beyond the audiogram: application of models of\nauditory fitness for duty to assess communication in the real world. Ear & <\/em>Hearing, 39(3)<\/em>, 434-435.<\/p>\n\n\n\n

        Easwar,\nV., Purcell, D. W., Aiken, S. J., Parsa, V., & Scollie, S. D. (2015a).\nEffect of stimulus level and bandwidth on speech-evoked envelope following\nresponses in adults with normal hearing. Ear and Hearing<\/em>, 36<\/em>(6),\n619-634<\/p>\n\n\n\n

        Easwar,\nV., Purcell, D. W., Aiken, S. J., Parsa, V., & Scollie, S. D. (2015b).\nEvaluation of speech-evoked envelope following responses as an objective aided\noutcome measure: effect of stimulus level, bandwidth, and amplification in\nadults with hearing loss. Ear and Hearing<\/em>, 36<\/em>(6),\n635-652. <\/p>\n\n\n\n

        Easwar, V., Banyard, A., Aiken, S.J., & Purcell,\nD.W. (2018). Phase-locked responses to the\nvowel envelope vary in scalp-recorded amplitude due to across-frequency\nresponse interactions. European Journal of\nNeuroscience, <\/em>48(10), 3126-3145.<\/p>\n\n\n\n

        Elberling,\nC., & Don, M. (1984). Quality estimation of averaged auditory brainstem responses. Scandinavian Audiology<\/em>, 13<\/em>, 187\u2013197.<\/p>\n\n\n\n

        Elberling,\nC., & Don, M. (2010). A direct approach for the design of chirp stimuli\nused for the recording of auditory brainstem responses. Journal of the\nAcoustical Society of America<\/em>, 128, 2955\u20132964.<\/p>\n\n\n\n

        Elberling,\nC., & Crone-Esmann, L. (2017). Calibration of\nbrief stimuli for the recording of evoked responses from the human\nauditory pathway. Journal of the Acoustical Society of America<\/em>, 141, 466-474. <\/p>\n\n\n\n

        Fabry, D. A. (2015). Moving beyond the audiogram<\/a>. Audiology Today<\/em>, 27 (3), 34-39. <\/p>\n\n\n\n

        Ferm, I., Lightfoot, G.,\n& Stevens J. (2013). Comparison of ABR response amplitude, test time, and estimation of\nhearing threshold using frequency specific chirp and tone pip stimuli in\nnewborns. International Journal of Audiology<\/em>, 52, 419-423.<\/p>\n\n\n\n

        Ferm,\nI., & Lightfoot, G. (2015). Further comparisons of ABR response amplitudes,\ntest time, and estimation of hearing threshold using frequency-specific chirp\nand tone pip stimuli in newborns: Findings at 0.5 and 2 kHz. International\nJournal of Audiology<\/em>, 54, 745\u2013750<\/p>\n\n\n\n

        Frank, J.,\nBalji\u0107, I., Hoth, S., E\u00dfer, D., & Guntinas-Lichius, O. (2017). The accuracy of objective\nthreshold determination at low frequencies: comparison of different auditory\nbrainstem response (ABR) and auditory steady state response (ASSR) methods. International\nJournal of Audiology, <\/em>56, 337-345. <\/p>\n\n\n\n

        Friederici,\nA. D. (2005). Neurophysiological markers of early language acquisition: from\nsyllables to sentences. Trends in Cognitive Sciences<\/em>, 9, 481-488<\/p>\n\n\n\n

        Friederici, A. D., Friedrich, M., & Weber,\nC. (2002). Neural manifestation of cognitive and precognitive\nmismatch detection in early infancy. NeuroReport, 13<\/em>, 1251-1254.<\/p>\n\n\n\n

        G\u00f8tsche-Rasmussen,\nK., Poulsen, T., & Elberling, C. (2012).  Reference hearing threshold levels for chirp signals delivered by an\nER-3A insert earphone. International Journal of Audiology, <\/em>51, 794-799. <\/p>\n\n\n\n

        Hamilton, L. S., & Huth, A. G. (2018): The revolution will not be controlled: natural stimuli in speech neuroscience. <\/a>Language, Cognition and Neuroscience<\/em>.<\/p>\n\n\n\n

        Holube, I., Fredelake, S., Vlaming, M., &\nKollmeier, B. (2010). Development and analysis of an International Speech Test Signal (ISTS). International\nJournal of Audiology<\/em>, 49(12), 891\u2013903.<\/p>\n\n\n\n

        Hoth, S., M\u00fchler, R., Neumann, K., & Walger, M.\n(2014). Objektive Audiometrie im Kindesalte<\/em>r. Berlin: Springer. <\/p>\n\n\n\n

        Jeng, F.C., Lin, C.D., & Wang, T.C. (2016).\nSubcortical neural representation to Mandarin pitch contours in American and Chinese\nnewborns. Journal of the Acoustical Society of America<\/em>, 139, 190-195. <\/p>\n\n\n\n

        John, M.S., Purcell, D.W.,\nDimitrijevic, A., & Picton, T. W. (2002). Advantages and caveats when\nrecording steady-state responses to multiple simultaneous stimuli. Journal of the American Academy of\nAudiology,<\/em> 13, 246-259.<\/p>\n\n\n\n

        Johnson, K. L., Nicol, T. G., & Kraus, N. (2005).\nBrain stem response to speech: a biological marker of auditory processing. Ear\n& Hearing, 26, <\/em>424-434.<\/p>\n\n\n\n

        Johnson,\nK. L., Nicol, T., Zecker, S. G., Bradlo, A. R.,\nSko, E., & Kraus, N. (2008). Brainstem encoding of voiced consonant\u2013vowel stop syllables. Clinical Neurophysiology<\/em>,\n119<\/em>, 2623\u20132635.<\/p>\n\n\n\n

        Kalashnikova, M., Peter, V., Di\nLiberto, G.M., Lalor, E.C., & Burnham, D. (2018). Infant-directed speech\nfacilitates seven-month-old infants\u2019 cortical tracking of speech. Scientific Reports, 8<\/em>,\nArticle number: 13745. <\/p>\n\n\n\n

        King, A., Hopkins, K., & Plack, C. J (2016). Differential group delay of\nthe frequency following response measured\nvertically and horizontally. Journal of\nthe Association for Research in Otolaryngology,<\/em> 17, 133\u2013143. <\/p>\n\n\n\n

        Krishnan,\nA. (2007). Human frequency following response. In R. F. Burkard, M. Don, &\nJ. J. Eggermont (Eds.), Auditory evoked potentials: Basic principles and\nclinical application. <\/em>(pp. 313\u2013335). Baltimore: Lippincott Williams &\nWilkins.<\/p>\n\n\n\n

        Krishnan,\nA. (2017). Shaping brainstem representation of pitch-relevant\ninformation by language experience. In N. Kraus, S. Anderson, T. White-Schwoch, R. R. Fay & A. N. Popper (Eds.). The frequency-following response: <\/em>a window into human\ncommunication<\/em>. (pp.\n45-73). Cham, Switzerland: Springer International. <\/p>\n\n\n\n

        Krishnan,\nA., Xu, Y., Gandour, J., & Cariani, P. A. (2004). Human frequency-following\nresponse: Representation of pitch contours in Chinese tones. Hearing\nResearch<\/em>, 189, 1\u201312.<\/p>\n\n\n\n

        Lalor,\nE. C., & Foxe, J.J. (2010). Neural responses to uninterrupted natural speech\ncan be extracted with precise temporal resolution. European Journal of\nNeuroscience<\/em>, 31, 189-193. <\/p>\n\n\n\n

        Laugesen,\nS., Rieck, J.E., Elberling, C., Dau, T., & Harte, J.M. (2018). On the cost of introducing\nspeech-like properties to a stimulus for auditory steady-state response\nmeasurements. Trends in Hearing, <\/em>22, 1-11. <\/p>\n\n\n\n

        Lins, O.G., & Picton, T.W.  (1995). Auditory steady-state responses to\nmultiple simultaneous stimuli.  Electroencephalography\nand clinical Neurophysiology,<\/em> 96, 420-432. <\/p>\n\n\n\n

        Lins, O.G., Picton, T.W., Boucher, B., Durieux-Smith,\nA., Champagne, S.C., Moran, L.M., Perez-Abalo, M.C., Martin, V., & Savio,\nG. (1996). Frequency-specific audiometry using steady-state responses.  Ear and Hearing,<\/em> 17, 81-96. <\/p>\n\n\n\n

        Michel,\nF., & J\u00f8rgensen, K. F. (2017). Comparison of threshold estimation in\ninfants with hearing loss or normal hearing using auditory steady-state\nresponse evoked by narrow band CE-chirps and auditory brainstem response evoked\nby tone pips. International Journal of Audiology<\/em>, 56<\/em>, 99\u2013105.<\/p>\n\n\n\n

        M\u00fchler, R., Mentzel, K., & Verhey, J. (2012) Fast\nhearing-threshold estimation\nusing multiple auditory steady-state responses with narrow-band chirps and adaptive stimulus patterns. Scientific World Journal<\/em>, Article ID\n192178<\/p>\n\n\n\n

        N\u00e4\u00e4t\u00e4nen, R., Pakarinen, S., Rinne, T., Takegata, R.,\n2004. The mismatch negativity (MMN)\u2014towards the optimal paradigm. Clinical Neurophysiology<\/em> 115,\n140\u2013144.<\/p>\n\n\n\n

        Partanen, E., Pakarinen, S., Kujala, T., &\nHuotilainen, M. (2013). Infants\u2019 brain responses for speech sound changes in fast multi-feature MMN\nparadigm. Clinical Neurophysiology<\/em>, 124(8), 1578\u20131585.<\/p>\n\n\n\n

        Picton T. W. (2011). Human Auditory Evoked\nPotentials<\/em>. San Diego: Plural Press.<\/p>\n\n\n\n

        Picton, T. W. (2013). Hearing in time: Evoked\npotential studies of temporal processing. Ear & Hearing<\/em>, 34, 385-40<\/p>\n\n\n\n

        Picton, T. W, Durieux-Smith, A., Champagne, S. C.,\nWhittingham, J., Moran, L.M., Gigu\u00e8re, C., & Beauregard Y. (1998).\nObjective evaluation of aided thresholds using auditory steady-state responses.\nJournal of the American Academy of Audiology<\/em>, 9:315-331.<\/p>\n\n\n\n

        Picton,\nT.W., Van Roon, P., & John, M.S. (2007). Human auditory steady-state\nresponses during sweeps of intensity.  Ear\n& Hearing<\/em>, 28, 542-557. <\/p>\n\n\n\n

        Ribas-Prats, T., Almeida, L., Costa-Faidella, J.,\nPlana, M., Corral M.J., .Gomez-Roig, M.D., & Escera, C. (2019). The\nfrequency-following response (FFR) to speech stimuli: A normative dataset in\nhealthy newborns. Hearing Research<\/em>, 371 28-39.<\/p>\n\n\n\n

        Rodrigues,\nG. R. I., Ramos, N., & Lewis, D. R. (2013). Comparing auditory brainstem responses (ABRs) to\ntoneburst and narrow band CE-chirp in young infants. International Journal\nof Pediatric Otorhinolaryngology<\/em>, 77<\/em>, 1555\u20131560.<\/p>\n\n\n\n

        Rodrigues,\nG. R., & Lewis, D. R. (2014). Establishing auditory steady-state response\nthresholds to narrow band CE-chirps(\u00ae) in full-term neonates. International\nJournal of Pediatric Otorhinolaryngology<\/em>, 78<\/em>,\n238\u2013243.<\/p>\n\n\n\n

        Scollie, S., & Bagatto, M. (2010). Fitting hearing aids to babies: three things you should know – 2010 update.<\/a> Audiology Today<\/em>. October 11, 2010<\/p>\n\n\n\n

        Shannon, R.V., Zeng, F.G., Kamath, V., Wygonski, J.,\nEkelid, M. (1995) Speech recognition with primarily temporal cues. Science, <\/em>270,\n303-304. <\/p>\n\n\n\n

        Sininger, Y. S., Hunter, L. L., Hayes, D., Roush, P.\nA., Uhler, K. M. (2018). Evaluation of speed and accuracy of next-generation auditory steady\nstate response and auditory brainstem response audiometry in children with\nnormal hearing and hearing loss. Ear & Hearing<\/em>, 39<\/em>,1207-1223.<\/p>\n\n\n\n

        Soli, S. D., Giguere, C., Laroche, C., et al. (2018a)\nEvidence-based occupational hearing Screening I: Modeling the effects of real-world noise\nenvironments on the likelihood of effective speech communication. Ear & <\/em>Hearing, 39(3)<\/em>, 436-448.<\/p>\n\n\n\n

        Soli, S. D., Amano-Kusumoto, A., Clavier, O., et al.\n(2018b). Evidence-based\noccupational hearing screening II: Validation of a screening methodology using\nmeasures of functional hearing ability. International\nJournal of Audiology<\/em>, 57(5)<\/em>, 323\u2013334.<\/p>\n\n\n\n

        Taylor, M. M. & Creelman, C. D. (1967). PEST:\nEfficient estimates on probability functions. Journal of the Acoustical Society of America, <\/em>41, <\/em>782-787.<\/p>\n\n\n\n

        Teng, X., Tian, X., Rowland, J., & Poeppel, D. (2017) Concurrent temporal channels for auditory processing: Oscillatory neural entrainment reveals segregation of function at different scales. <\/a>PLOS Biology<\/em>, 15(11): e2000812. <\/p>\n\n\n\n

        Teng,\nX., Tian, X., Rowland, J., & Poeppel, D. (2019). Theta and gamma bands\nencode acoustic dynamics over wide-ranging timescales. PLOS Biology<\/em> (in press)<\/p>\n\n\n\n

        Vanthornhout, J., Decruy,\nL., Wouters, J., Simon, J.Z., & Francart, T. (2018). Speech intelligibility\npredicted from neural entrainment of the speech envelope. Journal of the Association for Research in Otolaryngology<\/em>, 19(2), 181-191. <\/p>\n\n\n\n

        W\u00f6stmann, M., Fiedler, L. & Obleser, J.  (2017) Tracking the signal, cracking the code: speech and\nspeech comprehension in non-invasive human electrophysiology. Language,\nCognition and Neuroscience<\/em>, 32:7, 855-869.<\/p>\n","protected":false},"excerpt":{"rendered":"

        This page derives from a presentation on The Objective Evaluation of Human Hearing in March 2019 at the meeting of the Deutsche Gesellschaft f\u00fcr Audiologie in Heidelberg. The notes for the presentation and the actual PowerPoint file are available to download. The sound samples are<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":3,"comment_status":"closed","ping_status":"closed","template":"","meta":{"iawp_total_views":338,"footnotes":""},"class_list":["post-2080","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/pages\/2080","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2080"}],"version-history":[{"count":17,"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/pages\/2080\/revisions"}],"predecessor-version":[{"id":2224,"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=\/wp\/v2\/pages\/2080\/revisions\/2224"}],"wp:attachment":[{"href":"https:\/\/creatureandcreator.ca\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2080"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}