|Year : 2020 | Volume
| Issue : 1 | Page : 32-37
Outcome dependency on the level of task: Gap detection threshold and temporal modulation transfer function
Farooq Hussam, Udit Saxena, Venkata Damarla
Department of Audiology, MAA Institute of Speech and Hearing, Hyderabad, Telangana, India
|Date of Submission||27-Apr-2019|
|Date of Decision||23-Sep-2019|
|Date of Acceptance||20-Oct-2019|
|Date of Web Publication||19-Feb-2020|
Mr. Farooq Hussam
Department of Audiology, MAA Institute of Speech and Hearing, Hyderabad, Telangana
Source of Support: None, Conflict of Interest: None
Introduction: It has been demonstrated that a listener's temporal-processing capability is predictive of his or her performance on speech recognition, especially in noisy and complex environments. Aims and Objectives: The objective of the study is to check the task-related dependency of gap detection threshold (GDT) and temporal modulation transfer function (TMTF) tests in adults and children. Methods and Materials: The study was conducted on 40 normal-hearing participants, in which 20 children and 20 adults. All the participants in the study were subjected to the basic audiological evaluation. Results: It was observed that there is a significant difference (P < 0.001) between children and adults for GDT and TMTF. In the present study, the relation between GDT and TMTF and their outcome dependency has been measured to show that these two measures correlate in terms of measuring temporal resolution. Conclusions: The results shows that the average value for TMTF (8–9 ms) is comparatively more than GDT (2–3 ms), which indicates that the task involved in GDT is simpler than TMTF.
Keywords: Gap detection threshold, temporal modulation transfer function, temporal processing
|How to cite this article:|
Hussam F, Saxena U, Damarla V. Outcome dependency on the level of task: Gap detection threshold and temporal modulation transfer function. Indian J Otol 2020;26:32-7
|How to cite this URL:|
Hussam F, Saxena U, Damarla V. Outcome dependency on the level of task: Gap detection threshold and temporal modulation transfer function. Indian J Otol [serial online] 2020 [cited 2020 Jul 14];26:32-7. Available from: http://www.indianjotol.org/text.asp?2020/26/1/32/278736
| Introduction|| |
An excellent temporal resolution eases the detection of a target sound in masker sounds with temporally fluctuating envelopes, sound localization, and the understanding of speech. It has been demonstrated that a listener's temporal-processing capability is predictive of his or her performance on speech recognition, especially in noisy and complex environments (George et al., 2006).,, Therefore, behavioral techniques to estimate auditory temporal acuity have been an important topic in psychoacoustics, and efforts have been made to implement these techniques into clinical practice.,
Temporal resolution refers to the ability of the auditory system to follow rapid changes in the envelope of sound. The temporal resolution is measured in various ways, including detection threshold for amplitude modulation, forward masking and backward masking, and temporal order discrimination. Among these techniques that have been developed to estimate auditory temporal acuity, gap detection threshold (GDT), and temporal modulation transfer function (TMTF) are commonly adopted in clinical research and arguably are the most studied. Although both GDT and TMTF are believed to probe temporal processing, there is a lack of data directly comparing the results obtained from these two methods.
In a gap detection experiment, listeners detect the presence of a silent gap in a carrier sound. The GDT corresponds to the shortest gap to be detectable. Both pure-tone and noise carriers have been used in gap detection experiments previously. When a noise carrier is used, the GDT decreases as the carrier bandwidth and stimulus level increases. When broadband noise carriers are used, listeners with hearing impairment usually exhibit higher GDTs compared with those with normal hearing (e.g., Florentine, Buus and Geng 1999).,,, For narrowband carriers, the GDTs are much worse for listeners with hearing impairment than for those with normal hearing, which has been explained due to the loss of cochlear nonlinearity. In addition, a number of previous studies have found that the GDTs measured from older listeners are higher than those of young listeners, even when the age-related hearing loss is controlled., Moreover, when the gap detection task is made cognitively demanding by randomizing the temporal location of the gap within the carrier duration on a trial-by-trial basis, older adults show an increased age-related deficit in gap detection.,
Another approach to estimate temporal acuity is TMTF. Listeners detect the presence of sinusoidal amplitude modulation imposed on a carrier sound in the TMTF experiment. TMTF is typically a function relating to the modulation detection threshold and the modulation rate. For broadband noise carriers, the modulation detection thresholds are low and constant at low modulation rates. As the modulation rate exceeds approximately 60 Hz, the thresholds increase with increasing modulation rate at about 3–4 dB per octave. The shape of the TMTF resembles a first-order low-pass filter (or equivalently, a leaky integrator in the time domain), and it can be described by two parameters: the sensitivity (S), which corresponds to the plateau performance at low modulation rates, and the cutoff frequency (fcutoff) of the low-pass shape, which is thought to reflect the sluggishness of the auditory system.
Besides broadband stimuli, the TMTF is also measured using narrowband stimuli, although difficulties might arise because of the potential confound of spectral cues. Imposing amplitude modulation to a narrowband carrier creates spectral sidebands. This spectral change is comparable with the spectral resolution of the auditory periphery; listeners can use the spectral cue in performing the modulation detection task, leading to results that do not reflect temporal processing. One solution to this problem is to maintain the spectrum of the stimulus and manipulate the phase relationship among the spectral components to vary the modulation depth.
The outcome of any task depends on person's attention to sound stimuli as, how much they are motivated to respond, on their working memory, the way the instruction is given, underlying principle of tests, and the decision of the client to respond to stimuli. In the audiology field, there are some behavioral tests such as pure tone audiometry, speech audiometry, tests for adaptation like tone decay test and suprathreshold adaptation test and tests for recruitment, such as alternate binaural loudness balance test, monaural loudness balance, short increment sensitivity index, and binaural integration tests. It has been found that psychophysical behavioral measures are difficult to perform, and there is always an adult-child difference reflected in previous studies.
The objective of the study is to check the task-related dependency of GDT and TMTF tests in adults and children.
- To compare the GDT in adults and children
- To compare the TMTF in adults and children
- To see the relation between GDT and TMTF fest in adults
- To see the relation between GDT and TMTF test in children.
| Methods|| |
The present study consisted of 40 normal-hearing individuals, in which 20 children and 20 adults. All the participants have undergone basic hearing evaluation before the actual experiment. These participants have their pure-tone thresholds under 15 dBHL for frequencies 250–8000 Hz with speech recognition scores not <90% in both the ears and speech reception thresholds are appropriate with pure-tone thresholds. Immittance measures showed “A” type tympanogram with reflexes present in both ears, and Otoacoustic emissions (OAEs) were present.
The following equipment was used to conduct the study
- Otoscope: This was used to examine the ear canal and tympanic membrane of all participants
- Pure tone audiometer: PROTON SX3 clinical diagnostic audiometer with TDH 39P (Telephonics) circumaural headphones was used to estimate the hearing sensitivity for all participants
- Immittance audiometer: MAICO MI 34 with probe tone of 220 Hz was used to evaluate middle ear function for all participants
- Oto acoustic emission screener: Distortion product otoacoustic emissions were obtained using Interacoustics OAE screener for frequencies 1 KHz, 2 KHz, 3 KHz, 4 KHz, and 5 KHz
- Star Software V23: Standardized Star Software (Version 23, Cambridge University, United Kingdom) was installed on a laptop. The test was administered by presenting stimulus using Sony Premium In-Ear Extra Bass Headphones (MDR-XB55AP) connected to the laptop. The Internet Message Access protocol assesses a range of auditory and cognitive skills and tasks, which helps in assessing the psychoacoustic task for the temporal processing.
Pure tone thresholds were obtained using modified Hughson and Westlake procedure (Carhart and Jerger, 1959) across octave frequencies 250–8000 Hz for air conduction and 250–4000 Hz for bone conduction. Tympanometry and reflexometry were carried out to rule out any middle ear pathology. OAE was done to check for outer hair cell functioning.
All the experiments were conducted in acoustically treated room without any environmental disturbances. Each participant asked to sit comfortably, instructed to remain calm and attentive to the stimuli being presented. The types of stimuli presented to them were explained verbally. It was explained that in gap detection task, they will be hearing three noises with an interstimulus interval of 500 ms following each noise. Each noise is represented by an animation. In those three noises, one noise consists of gap for which they have to identify and click on it. A similar procedure is followed in the modulation detection task where the participant has to identify the modulated noise from three noise presentations.
Task 1: The experimental trial in the modulation detection task consisted of three sound intervals, of which two signals are same, and one is different. The modulation frequency used was 8 Hz with intervals of 500 ms in duration, separated by inter-stimulus intervals of 500 ms, and each interval contained a band-limited noise. The listener is instructed to select which interval sounded different from the other two to discriminate the AM.
Task 2: The experimental trial in the gap detection task consisted of three sound intervals separated by interstimulus intervals of 500 ms. Each interval contained noise, 1 KHz and 2 KHz pure tones. A silent gap is introduced in the temporal center of the noise. The listener is instructed to select which interval contained the gap.
In the current version of STAR, there is only one method to implement an adaptive task “adaptive staircase.” It is a set of rules, which adjust the level of difficulty of the task in accordance with the participant's performance, i.e., when incorrect responses are given, the task becomes easier whereas the task becomes harder for correct responses. Typically, this staircase targets a performance level of a specific percentage correct at certain point on the psychometric function. The participant is instructed to hear the three different sprites which are embedded with the tone in it. When they listen to it, they respond by clicking on the screen with the pointer.
The level of difficulty of the trial is defined by either level delta, depending on the nature of the task. For detection tasks (e.g., tone in noise), level delta is in the intensity as that of the target. For discrimination tasks, level delta is the difference between the standardized and the target expressed as a percentage (usually the percentage of the standard).
The adaptive staircase is defined by the following information:
- Start point – The level delta to start the staircase from
- Reference level – The value from which level delta is calculated as a percentage (almost the standard)
- Base value – The value to which the level delta is added or subtracted to create the target tone
- Minimum value – The smallest level delta which the staircase is allowed to adapt to
- Maximum difficulty – The largest level delta which the staircase is allowed to adapt to
- Additive versus multiplicative adaption.
- Rule – The criteria for adaptation this consists of:
- Step size – The size of the increment that the level delta changes between trials
- “Up” criteria – Number of incorrect responses in a row required for the staircase to become easier
- “Down” criteria – Number of correct response in a row required for the staircase to become harder.
The same procedure was used for both GDT task and TMTF task.
| Results|| |
The present study investigated the task-related dependency of GDT and TMTF tests in adults and children. In the present study, GDT was compared in children and adults, and TMTF was compared in children and adults. GDT and TMTF were correlated in children and adults. Forty normal participants were selected in which 20 children 20 adults with an age range of 7–12 years and 18–30 years, respectively.
Gap detection threshold
Mean and standard deviation were calculated to evaluate the GDTs and compared in normal individuals (children and adults) using white noise [Table 1]. It was observed that there was a significant difference between children and adults GDT. Adult's GDTs were lesser in comparison to children. As shown in [Figure 1], the mean value of GDT in children is higher than that of adults. The Independent sample t-test was administered in children and adults. The results [Table 2] reveal that there was a statistically significant difference across the groups (P < 0.001).
|Table 1: The mean and standard deviation for both children and adults with respect to broadband (white) noise|
Click here to view
|Figure 1: Bar graph representing the mean and standard deviation of gap detection threshold for both children and adults|
Click here to view
|Table 2: Independent sample t-test scores of gap detection threshold for children and adults|
Click here to view
Temporal modulation transfer function
Mean and standard deviation were calculated to evaluate the TMTF and compared in normal individuals (children and adults). It was observed that there was a significant difference between children and adults. The adult's TMTF scores were lesser in comparison to children. As shown in [Figure 2], the mean of TMTF in children is higher than the mean of adult. The Independent sample t-test was administered to determine the significance of TMTF scores between children and adults. The results reveal that there is a statistically significant difference across the groups (P < 0.001) [Table 3] and [Table 4].
|Figure 2: Bar graph representing the mean and standard deviation values for both children and adults with respect to the temporal modulation transfer function|
Click here to view
|Table 3: The mean and standard deviation of temporal modulation transfer function for both children and adults using 8Hz modulation|
Click here to view
|Table 4: Independent sample t-test results in children and adults the results for temporal modulation transfer function task|
Click here to view
The Pearson's correlation coefficient test was applied to see the relation between the two measures, i.e., GDT and TMTF in both children and adults. There was a significant correlation between GDT and TMTF in normal-hearing children with an age range of 7–12 years (P < 0.001), as well as in normal-hearing adults with an age range of 18–30 years (P < 0.001).
| Discussion|| |
From [Figure 1] and [Figure 2], there is a significant difference (P < 0.001) between children and adults in the GDT and TMTF. Further, it is evident from [Figure 3] and [Figure 4] that GDT and TMTF are correlating in both children and adults, and the mean of GDT and TMTF is varying among children and adults.
|Figure 3: Graphical representation shows the relation between gap detection threshold and temporal modulation transfer function in normal-hearing children|
Click here to view
|Figure 4: Graphical representation shows the relation between gap detection threshold and temporal modulation transfer function in normal-hearing adults|
Click here to view
In the present study, the relation between GDT and TMTF showed that these two measures correlate in terms of measuring temporal resolution. The results showed that the average value for TMTF (8–9 ms) is comparatively more than GDT (2–3 ms) which indicates that the task involved in GDT is simpler than TMTF. Although the TMTF and GDT can both be modeled using a leaky integrator model of auditory temporal processing, it is not clear that these two measures of temporal resolution are equivalent to one another. Formby and Muir reported a study where the bandwidths of the noise carriers were manipulated in both gap detection and TMTF measurements. This manipulation introduced within-subject variability in the GDT and the TMTF parameters. Studying within-subject variability, a negative correlation was observed between (sensitivity) S and GDT, whereas (cut-off rate) fc was found to be relatively invariant with GDT. These observations suggest that the GDT is predictive of S, but not fc. Therefore, it is likely that the TMTF provides more information regarding temporal processing than the GDT. Similar to this argument, it has been suggested that the GDT reflects both temporal and intensity resolution, whereas the TMTF paradigm allows separate assessments of temporal and intensity resolution. Therefore, in many cases, the TMTF paradigm might be preferred over the gap detection paradigm.
It was observed that the results of the present study are supported by Shen and Richards (2015) that these two procedures were developed for efficient assessment of the temporal acuity. Both procedures adjust stimuli adaptively to maximize the information gain regarding the TMTF parameters S and fc. One procedure (the Bayesian procedure) adopted the Bayesian adaptive algorithm described by Kontsevich and Tyler while the other procedure (the two-track procedure) utilized two up-down tracks, each of which targets a specific region in the stimulus space. The TMTF estimates were compared with results from the TMTF measurements using traditional procedure and from gap detection measurements. The results showed that although the Bayesian adaptive procedure is theoretically sound and demonstrates excellent efficiency in computational simulations, it underestimated S when human listeners were tested. The estimated TMTFs predicted higher modulation detection thresholds than the experimental data using the traditional procedure. In contrast to the Bayesian procedure, the estimated TMTFs using the two-track procedure predicted the modulation detection thresholds that were more consistent with the traditional procedure.
Thus, the following conclusions can be drawn from the present study; aging can alter temporal resolution. The response of children and adults varied accordingly in GDT and TMTF tasks with better responses observed in GDT than TMTF. Thus, it can be observed that the task involved in both measures is important when assessing the temporal resolution.
An excellent temporal resolution eases the detection of a target sound in masker sounds with temporally fluctuating envelopes, sound localization, and the understanding of speech. It has been demonstrated that a listener's temporal-processing capability is predictive of his or her performance on speech recognition, especially in noisy and complex environments. Therefore, behavioral techniques to estimate auditory temporal acuity have been an important topic in psychoacoustics, and efforts have been made to implement these techniques into clinical practice.
Especially while considering the auditory processing disorders, temporal processing is affected badly. The present study showed aging can alter temporal resolution. The response of children and adults varied accordingly in GDT and TMTF tasks with better responses observed in GDT than TMTF. Thus, it can be concluded that the task involvement in both measures is important when assessing the temporal resolution.
The study will add to the existing literature on the temporal processing test. Results will give clinician an idea on which test to select depending on the age of the participant. Results will help prorate the result of test which is not done based on the result of the test which is done. Depending on which test comes out to be more performable, it will be easy to decide on which should be used for training to improve the underlie acuity.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
George EL, Festen JM, Houtgast T. Factors affecting masking release for speech in modulated noise for normal-hearing and hearing-impaired listeners. J Acoust Soc Am 2006;120:2295-311.
Carhart R, Jerger J. Preferred method for clinical determination of pure-tone thresholds. Journal of Speech & Hearing Disorders 1959;24:330-45.
Snell KB, Mapes FM, Hickman ED, Frisina DR. Word recognition in competing babble and the effects of age, temporal processing, and absolute sensitivity. J Acoust Soc Am 2002;112:720-7.
Florentine M, Buus S, Geng W. Psychometric functions for gap detection in a yes-no procedure. J Acoust Soc Am 1999;106:3512-20.
Musiek FE, Shinn JB, Jirsa R, Bamiou DE, Baran JA, Zaida E. GIN (Gaps-in-noise) test performance in subjects with confirmed central auditory nervous system involvement. Ear Hear 2005;26:608-18.
Green DM, Forrest TG. Temporal gaps in noise and sinusoids. J Acoust Soc Am 1989;86:961-70.
Eddins DA, Hall JW 3rd
, Grose JH. The detection of temporal gaps as a function of frequency region and absolute noise bandwidth. J Acoust Soc Am 1992;91:1069-77.
Fitzgibbons PJ, Wightman FL. Gap detection in normal and hearing-impaired listeners. J Acoust Soc Am 1982;72:761-5.
Irwin RJ, Hinchcliff LK, Kemp S. Temporal acuity in normal and hearing-impaired listeners. Audiology 1981;20:234-43.
Irwin RJ, McAuley SF. Relations among temporal acuity, hearing loss, and the perception of speech distorted by noise and reverberation. J Acoust Soc Am 1987;81:1557-65.
Tyler RS, Summerfield Q, Wood EJ, Fernandes MA. Psychoacoustic and phonetic temporal processing in normal and hearing-impaired listeners. J Acoust Soc Am 1982;72:740-52.
Moore BC, Glasberg BR. Gap detection with sinusoids and noise in normal, impaired, and electrically stimulated ears. J Acoust Soc Am 1988;83:1093-101.
Grose JH, Hall JW 3rd
, Buss E, Hatch D. Gap detection for similar and dissimilar gap markers. J Acoust Soc Am 2001;109:1587-95.
He NJ, Horwitz AR, Dubno JR, Mills JH. Psychometric functions for gap detection in noise measured from young and aged subjects. J Acoust Soc Am 1999;106:966-78.
Harris KC, Eckert MA, Ahlstrom JB, Dubno JR. Age-related differences in gap detection: Effects of task difficulty and cognitive ability. Hear Res 2010;264:21-9.
Viemeister NF. Temporal modulation transfer functions based upon modulation thresholds. J Acoust Soc Am 1979;66:1364-80.
Kohlrausch A, Fassel R, Dau T. The influence of carrier level and frequency on modulation and beat-detection thresholds for sinusoidal carriers J Acoust Soc Am 2000;108:723-34.
Tabuchi H, Borucki E, Berg BG. Effects of randomizing phase on the discrimination between amplitude-modulated and quasi-frequency-modulated tones. Hear Res 2012;290:72-82.
Formby C, Muir K. Modulation and gap detection for broadband and filtered noise signals. J Acoust Soc Am 1988;84:545-50.
Strickland EA, Viemeister NF. The effects of frequency region and bandwidth on the temporal modulation transfer function. J Acoust Soc Am 1997;102:1799-810.
Shen Y, Richards VM. Temporal modulation transfer function for efficient assessment of auditory temporal resolution. J Speech Lang Hear Res 2015;58:2280-92.
Kontsevich LL, Tyler CW. Bayesian adaptive estimation of psychometric slope and threshold. Vision Res 1999;39:2729-37.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3], [Table 4]