Decrement in Noise Test (DeNT): A Clinical Measure of Partially Filled Gap Detection Performance

by Julianne M. Ceruti, Au.D., Ph.D., Leslie Bernstein, Ph.D., and Frank Musiek, Ph.D.

The Decrement in Noise Test (DeNT) is a clinically oriented procedure modeled after the GIN test that employs both partially filled gaps (i.e., decrements) and full gaps. This test was developed to improve clinical assessment of temporal resolution that addresses the intensity resolution confound when only full gaps are employed. The DeNT utilizes four decrement depths (i.e., 25%, 50%, 75%, 100%) and 10 decrement durations that are randomly represented six times within 150 noise trials. The decrement durations vary slightly from the GIN test in order to sample at more equal intervals between two and 20 ms.

The DeNT uses 2, 3, 5, 6, 8, 10, 12, 15, 18 and 20 ms durations while the GIN uses 2, 3, 4, 5, 6, 6, 10, 12, 15, 20 ms durations. The equal sampling is important due to the additional decrement depths, as the duration thresholds are typically longer for smaller decrement depths (Forrest & Green, 1987). As there are multiple decrement depths, four psychometric functions (i.e., one for each decrement depth) can be obtained that relate the hit rate to the duration and depth of the decrement. The DeNT also employs catch trials (i.e., false alarms), which allow the clinician to measure the false alarm rate. Results from the DeNT procedure were validated and assessed against the gold-standard psychoacoustic technique for measurement of gap detection performance, a two interval, two alternative forced choice (2I/2AFC) paradigm, which used an adaptive technique to measure gap detection thresholds.

Participants consisted of normal-hearing individuals aged 18 to 31 years with normal peripheral auditory sensitivity and function, bilaterally, as determined by pure tone audiometry, tympanometry and distortion-product optoacoustic emissions. All participants scored at least 90% in each ear on the Dichotic Digits test presented at 50 dB HL (Musiek et al, 1991). Participants were all female and had no significant self-reported history of otologic disorders, neurologic disorders or learning disability. Participants were screened for normal IQ and cognition.

The DeNT was based on the method of constant stimuli. Each participant was presented with 165 six-second-long broadband noise stimuli and instructed to listen for any brief interruptions (i.e., silence gap) that may or may not occur within each noise. Each six-second long, broadband noise was followed by 50 ms of silence, after which the participant could begin the next trial by pressing the space bar. The DeNT took approximately 20 minutes to complete per ear. A recorded response was considered to be a “hit” if it occurred within 1000 ms of gap onset. Any responses that occurred within 1000 ms of the designated 0% decrement depths were considered to be false alarms. The performance measures were hit-rate and d’.

Thirty-five participants completed the DeNT procedure and a subset of the participants completed the 2I/2AFC paradigm. There was no significant ear effect observed. Responses were tabulated and an average hit rate was calculated across individuals for each combination of decrement duration and depth. A psychometric function was formed by plotting hit-rate as a function of gap duration.  For the values of k of 0.50, 0.75, and 1.00, logistic functions were fitted to the data using a least-squares criterion and the fits were characterized by values of r² of about 0.99. There was good test-retest reliability with no observed learning effects for the 2I/2AFC task. There was a statistically significant difference between decrement duration thresholds for the 75% and 100% decrement depths (paired-samples t-test, p<0.01) when comparing thresholds obtained using the adaptive 2I/2AFC to clinical auditory threshold obtained from the DeNT (lowest decrement duration with at least 67% performance).

A systematic relationship between the effective amplitude change within a decrement and listener performance was observed. The relationship between performance and the effective decrease in amplitude that occurs within a partially filled gap (i.e.,. decrement) can be described in terms of a depth-duration trade-off. This trade-off is essentially how much longer or shorter a decrement’s duration has to be made, as a function of the decrement depth, in order to maintain criterion performance. Forrest and Green (1987) examined the relation between decrement depth and decrement duration and posited that detection occurs once the partially filled gap produces a criterion change (e.g. decrease) in the amplitude of a stimulus, which can be achieved by changing either the decrement duration or the decrement depth. In order to gain an intuitive understanding of the model, consider that the auditory system cannot, of course, respond instantaneously to changes in amplitude. Rather, it requires some finite time to respond over which the input is effectively averaged.  This suggests the presence of some sort of low-pass filtering of the input. Such filtering has been described as coming about through the operation of a “temporal window”. The longer the amount of time over which the system averages, the more sluggish the system is to respond to changes in amplitude.  Assuming a simple filter model of the temporal window, Forrest and Green (1987) derived a set of equations relating gap duration, gap depth, the time-constant of the filter, and the criterion change in amplitude required for detection.  They found that the simple temporal window model reasonably accounted for thresholds obtained for both modulation and partially filled gap detection tasks. The temporal model obtained from the DeNT and 2I/2AFC experimental data is consistent with those obtained by Forrest and Green (1987). It should be noted that substantially lower decrement duration thresholds are obtained when d’ is used in lieu of a hit-rate based criterion, which, along with the low overall false-alarm rate, suggests that participants are using a strict criterion that minimized false alarms when completing the DeNT procedure.

In summary, the depth-duration trade off is non-linear and can be described mathematically using the temporal window model of Forrest and Green (1987), with parameters that have been optimized for a best fit for the data collected from the 2I/2AFC procedure and thresholds derived via  d’ analysis from the DeNT procedure. The DeNT procedure can be used to accurately and reliably predict decrement detection thresholds that are comparable to decrement detection thresholds obtained using a 2I/2AFC procedure. As the DeNT procedure incorporates both a measure of hit rate and false alarm rate, d’ can and should be calculated in order to correct for a criterion response bias.

Previously presented as a poster at American Auditory Society 2016, Joint Defense Veterans Affairs Conference 2016 and Academy Research Conference 2016.

References & Suggested Readings

1. Forrest, T. G., & Green, D. M. (1987). Detection of partially filled gaps in noise and the temporal modulation transfer function. The Journal of the Acoustical Society of America, 82(6), 1933–1943.
2. Glasberg, B. R., Moore, B. C. J., & Peters, R. W. (2001). The influence of external and internal noise on the detection of increments and decrements in the level of sinusoids. Hearing Research, 155(1-2), 41–53.
3. Green, D. M., & Forrest, T. G. (1989). Temporal gaps in noise and sinusoids. The Journal of the Acoustical Society of America, 86(3), 961–970.
4. Hautus, M. J., Setchell, G. J., Waldie, K. E., & Kirk, I. J. (2003). Age-related improvements in auditory temporal resolution in reading-impaired children. Dyslexia, 9(1), 37–45.
5. Hoch, M., Blumsack, J., & Soles, L. (2014). Relationship between gap detection thresholds and performance on the Advanced Measures of Music Audiation Test. The Journal of the Acoustical Society of America, 136(4), 2307–2307.
6. Levitt, H. (1971). Transformed up-down methods in psychoacoustics. The Journal of the Acoustical Society of America, 49(2), Suppl 2:467+.
7. Musiek, F. E., Gollegly, K. M., Kibbe, K. S., & Verkest-Lenz, S. B. (1991). Proposed screening test for central auditory disorders: Follow-up on the dichotic digits test. The American Journal of Otology, 12(2), 109–113.
8. Oxenham, A. J. (1997). Increment and decrement detection in sinusoids as a measure of temporal resolution. The Journal of the Acoustical Society of America, 102(3), 1779–1790.
9. Plack, C. J., Gallun, F. J., Hafter, E. R., & Raimond, A. (2006). The detection of increments and decrements is not facilitated by abrupt onsets or offsets. The Journal of the Acoustical Society of America, 119(6), 3950–3959.
10. Plack, C. J., & Moore, B. C. (1991). Decrement detection in normal and impaired ears. The Journal of the Acoustical Society of America, 90(6), 3069–307