- Email: [email protected]
Auditory gap detection in the early blind
Hearing Research 211 (2006) 1–6 www.elsevier.com/locate/heares
Research paper
Auditory gap detection in the early blind Kurt E. Weaver a, Alexander A. Stevens a b
a,b,*
Department Of Behavioral Neuroscience, Oregon Health and Science University, Portland OR, USA Department of Psychiatry, Oregon Health and Science University, 3181 SW Sam Jackson Park Rd., CR139, Portland, OR 97239, USA Received 17 April 2005; received in revised form 17 April 2005; accepted 19 August 2005 Available online 26 October 2005
Abstract For blind individuals, audition provides critical information for interacting with the environment. Individuals blinded early in life (EB) typically show enhanced auditory abilities relative to sighted controls as measured by tasks requiring complex discrimination, attention and memory. In contrast, few deficits have been reported on tasks involving auditory sensory thresholds (e.g., Yates, J.T., Johnson, R.M., Starz, W.J., 1972. Loudness perception of the blind. Audiology 11(5), 368–376; Starlinger, I., Niemeyer, W., 1981. Do the blind hear better? Investigations on auditory processing in congenital or early acquired blindness. I. Peripheral functions. Audiology 20(6), 503–509). A study of gap detection stands at odds with this distinction [Muchnik, C., Efrati, M., Nemeth, E., Malin, M., Hildesheimer, M., 1991. Central auditory skills in blind and sighted subjects. Scand. Audiol. 20(1), 19–23]. In the current investigation we re-examined gap detection abilities in the EB using a single-interval, yes/no method. A group of younger sighted control individuals (SCy) was included in the analysis in addition to EB and sighted age matched control individuals (SCm) in order to examine the effect of age on gap detection performance. Estimates of gap detection thresholds for EB subjects were nearly identical to SCm subjects and slightly poorer relative to the SCy subjects. These results suggest some limits on the extent of auditory temporal advantages in the EB. Ó 2005 Elsevier B.V. All rights reserved. Keywords: Early blindness; Auditory temporal discrimination; Aging
1. Introduction Individuals blinded from birth critically depend upon audition as a primary source of information. A substantial amount of research from animals studies has suggested that in the absence of visual stimulation, auditory advantages develop that exceeds those of sighted individuals (For reviews, see Rauschecker, 1995; Bavelier and Neville, 2002, for a review). It has been argued however that advantages do not develop at all levels of auditory function; rather the early blind (EB) typically outperform
Abbreviations: EB, early blindness; SCm, age matched sighted controls; SCy, young sighted controls; MGT, minimum gap threshold; 2AFC, two alternative forced choice paradigm; ANOVA, analysis of variance * Corresponding author. Tel.: +503 494 0358; fax: +503 494 1209. E-mail address: [email protected] (A.A. Stevens). 0378-5955/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.heares.2005.08.002
sighted counterparts on auditory tasks that require complex perceptual discriminations. For instance, the EB consistently show better performance than SC individuals on tasks that examine auditory memory (Roder and Rosler, 2003), verbal memory (Amedi et al., 2003) and auditory attention (Kujala et al., 1997; Hugdahl et al., 2004), but have comparable auditory detection thresholds and intensity discrimination (Starlinger and Niemeyer, 1981; Yates et al., 1972). One area of particular interest underlying the development of auditory advantages in the blind is around the issue of temporal processing. Neville and Bavelier (2001) and Roder and colleagues (1999) have advanced the hypothesis that many of the auditory advantages possessed by the blind may stem from faster auditory temporal processing. Auditory temporal processing has different meanings at different levels within the auditory system.
2
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
For instance, gap detection tasks is thought to measure the shortest silent interval that can be detected within an auditory signal, a reflection of perceived within-channel sampling resolution (Plomp, 1964; Phillips, 1999; Roberts and Lister, 2004). Typically, trained subjects have thresholds for within-channel gaps on the order of 2– 4 ms in length, when sandwiched between broadband noise (Plomp, 1964; Green, 1985; Irwin and Purdy, 1982). Below this minimum gap threshold (MGT) the silent gap is masked by the post gap stimulus (Plomp, 1964; Gou and Burkard, 2002). In this within-channel paradigm, pre and post-gap noise bursts share identical spectral properties and therefore stimulate the same set of auditory neurons (in contrast, between or across-channel gap detection varies the spectral content of the noise markers so that leading and trailing markers do not overlap and thus activate different perceptual channels; see Phillips et al., 1997; Buus and Florentine, 1985). Converging evidence from single unit and auditory brain stem recordings suggests that, because listeners must simply detect the absence of afferent activity, within-channel gap detection resolves the discharge rates of peripheral relays, including auditory brainstem nuclei that feed into the cortex (Werner et al., 2001; Giraudi et al., 1980; Frisina, 2001; Zhang et al., 1990; Phillips et al., 1997). However, in animal models damage to primary auditory cortex disrupts gap detection, requiring longer duration gaps for accurate responding (Kelly et al., 1996). In EB subjects, MGTÕs estimates have been reported to be significantly shorter than SC, suggesting that loss of vision early in life results in superior auditory temporal resolution (Muchnik et al., 1991). In this task, listeners were required to detect a gap placed within one of two periods of noise using an adaptive 2 alternative forced choice paradigm (2AFC). This evidence suggests that early visual deprivation enhances the sampling resolution of the auditory pathway. However, the best MGT for the SC group reported in the gap detection task was 13.8 ms when noise period durations were set at 85 ms while the early blind threshold was 4.8 ms (Muchnik et al., 1991). The SC groupÕs threshold was substantially longer than MGT thresholds reported in the literature (typically 2–4 ms; Plomp, 1964; Green, 1985; Irwin and Purdy, 1982). Indeed the control groupÕs threshold in Muchnik et al. (1991) was approximately 10 ms longer than that of sighted controls performance on the same task reported previously by the same investigator (Muchnik et al., 1985). This raises the possibility that the sighted control group in the later study was not representative of normal sighted performance on the task. Given the inconsistencies in these earlier reports and the importance of understanding the different levels at which auditory temporal resolution may influence EB auditory perception, the goal of the present study was to reexamine auditory temporal resolution in EB subjects using a withinchannel gap detection task. We determined MGTÕs using a single-burst, psychometric yes/no paradigm (Florentine
et al., 1999, 2001; He et al., 1999). Further, because both the EB and SC groups in the current investigation were much older than those reported by Muchnik et al., a younger group of sighted individuals was included in the analysis to examine the influence of age. 2. Materials and methods 2.1. Subjects Ten early blind (EB, 6 female; mean age 51.5, range 42–57), 10 sighted, age-matched controls (SCm, 6 female, mean age 52.2, range of 44–64) and 10 young sighted controls (SCy, 4 females, mean age 27.9, range 24–33) were examined. Exclusion criteria for the blind included: age of blindness onset occurring after the first year of life, and light sensitivity defined as form or color vision. Additional exclusion criteria for all subjects included previously diagnosed hearing damage or loss, comorbid psychiatric or neurological disease, and drug or alcohol abuse within the past five years. Cause of blindness in all EB listeners stemmed from peripheral damage due to complications of retinopathy of prematurity. One EB participant had diffuse light sensitivity. Subjects reported normal hearing and all signed an informed consent prior to the testing session. The EB and SCm subjects also served as subjects in a study of auditory backward masking, reported elsewhere and were screened for frequency discrimination (Stevens and Weaver, in press). All experimental procedures were approved by the Institutional Review Board of Oregon Health & Science University.
2.2. Gap detection Subjects were presented with a single interval of band-pass noise (center frequency of 2500 Hz. with a band width of 2000 Hz) with an overall duration of 400 ms. Noise bursts were presented at 75 dB SPL with a 10 ms rise/fall time at the initial and final periods of the noise period, and a rectangular silent gap. Silent, gaps with a duration of 1–7 ms (inclusive of rise/fall time) were randomly presented in some trials. The onset of a gap occurred randomly between the first and last 80 ms of the stimulus period. Subjects indicated on a button box whether they detected a gap and no feedback was given. Thirty trials were presented at each gap duration as well as the no-gap condition (catch trials), resulting in a total of 240 trials across the session. The experiment was divided up into blocks of 100 trials and subjects were allowed to rest between each block if needed. A practice session was run prior to the experimental session where subjects were presented with 5 trials of the no gap and 5 trials of the 7 ms condition.
2.3. Data analysis To measure sensitivity differences and control for differences in false positives rates (i.e., a yes response on a no gap trial) between groups, psychometric curves were fitted to the individual subject data using a logistic regression function described by Green (1993) that incorporates false positive frequency,
P ðyesÞ ¼ a þ ð1 aÞ=ð1 þ ekðxmÞ Þ;
ð1Þ
where the probability of a ‘‘yes’’ response, P(yes), at a gap length x, reflect the rate of false positive responses a, and k is a free parameter that describes the slope of the fitted logistic curve with an estimated midpoint, m. For each listener, k and m were initially determined using a curve-estimation procedure that plots the best-fit logistic function against the measured data (cf., He et al., 1999). Midpoint values were used as an estimate of MGT (Buus, 2002). Because detection estimates are influenced by listenersÕ response criteria (i.e., false positives, see Gu and Green, 1994), Eq. (1) was used to solve for the midpoint parameter m, taking into account differences of individual false-alarm rates.
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
3
2.4. Materials Stimuli were generated using Tucker Davis Technologies (Gainsville, FL) Real-Time Processor System 3 and presented using Microsoft Visual Basic Software (Seattle, WA). Intensity of stimuli was controlled using System 3 attenuators and all stimuli were presented through Senheisser HD265 dynamic range headphones (Old Lyme, CT) which were calibrated before each session using a UEI DSM 100 digital sound meter (Beaverton, OR). The task was conducted in a 2-walled sound-proof booth.
3. Results 3.1. Estimating gap detection thresholds Fig. 1 shows logistic curve estimates for a representative SCm, EB and SCy subject. The open squares represent the measured data with the line showing the best-fit logistic function. Curve fits were significant for all subjects (P values <0.01; analysis of variance F test for Regression Relation) and showed high correlation coefficients (r2 > 0.90, see Fig. 1). We next recast individual logistic functions according to Eq. (1) in order to control for differences in false positive rates. This analysis resulted in similar performance curves averaged across individual groups (Fig. 2(a)) and are similar to previous estimates of MGT using similar protocols on audiologically normal individuals (cf., He et al., 1999, Experiment 1). Within all three groups, detection improved as the gap length increased. We compared group performance across gap conditions using a repeated measures analysis of variance (ANOVA) and a Geiser–Greenhouse correction for violations of sphericity (uncorrected degrees of freedom are reported for clarity). This analysis resulted in a main effect of group [F2,27 = 6.127, P = 0.001], a main effect of gap condition, [F1.51,40.88 = 1319.629, P < 0.001] and a significant interaction of group and gap [F3.028,40.88 = 4.141, P = 0.012]. Post hoc t tests using a Sˇida´k correction for multiple corrections (P-values <0.05) were used to assess group differences at each gap duration.
Fig. 2. Gap detection performance in the EB, SCm and SCy. Proportion of gaps detected (a yes response) is plotted as a function of gap length for all three groups. (a) Psychometric curves were adjusted taking into consideration false positive rates according to Eq. (1) and averaged across each group. *Denotes significant differences at the 0.05 level between EB and SCy listeners. No significant performance differences appeared between EB and SCm subjects. Mean SCm scores were however significantly lower than SCy at all gap lengths except for the 7 ms condition. (b) Raw, unadjusted performance curves are shown for each group of listeners. Error bars represent standard error.
These detected no statistically reliable differences between SCm and EB but significant performance differences between the EB and SCy appeared on trials with no gap
Fig. 1. Logistic curve-estimation. The best-fit logistic function (lines) was determined for each subjectÕs measured results (open squares) based upon a least squares method. A representative SCm, EB and SCy subject are shown with corresponding r2, m and k values displayed in the inset. Values for the midpoint (m) and the slope (k) were estimated based upon the resulting logistic function.
4
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
and 1 ms gap, reflecting an elevated proportion of ‘‘yes’’ responses on these trials (Fig. 2), which on the no-gap trials, provide an estimate of false alarm rate. Observed proportion yes responses were reliably different between SCy and SCm listeners at all sampled gap lengths except the 7 ms condition. Fig. 2(b) (inset) show averaged group results uncorrected for false alarm rate. 3.2. Threshold estimates and slope parameters Estimates of the 50% threshold value using Eq. (1) to account for the false positive rate for each subject was calculated as a measure of gap threshold (MGT; Buus, 2002). The estimated MGT values for each individual and the mean scores are shown in Fig. 3(a) (mean values for: ms SCm: 3.39 ± 0.17; EB: 3.40 ± 0.14; SCy: 3.76 ± 0.09). A one-way analysis of variance comparing MGT values between groups failed to reach significance [F2,27 = 2.274, P = 0.122] despite a mildly larger mean MGT estimate for the SCm group relative to either group of listeners. A lack of a significant effect must be taken with some caution, however, due to the low sample size within each group. The slope (k) was estimated for each subject (Fig. 3(b)). The SCy group had a somewhat steeper slope estimate (EB: 1.37 ± 0.03; SCm: 1.40 ± 0.03; SCy: 1.56 ± 0.03) but in general, the k estimates of the SCm and SCy were similar to old/young differences reported in the literature (He et al., 1999). The omnibus F test did revealed a statistically reliable difference among the groups [F2,27 = 11.609, P = 0.001] and post hoc tests (P < 0.05) revealed that a reliable difference reflecting a steeper slope estimate for the SCy participants compared to both the EB and the SCm, who did not differ.
4. Discussion The current results suggest that EB individuals differ little from their age-matched sighted counterparts (SCm) in their ability to detect gaps embedded in noise. Additionally, although statistically significant performance differences appeared between the SCy and EB at the no gap and 1 ms gap conditions (Fig. 2) and slopes of the logistic curves (Fig. 3(b)), they appear to reflect a greater false alarm rate in the SCy group; higher rates of indicating the presence of a gap even in their absence. Additionally, no significant differences appeared between sighted listeners (young or old) and blind listeners in MGT estimates. These results must be interpreted with caution due to the relatively small sample sized of the groups, and potential differences in gap detection thresholds that may be associated with differences in the etiology of blindness in the current EB group (all 10 had retinopathy due to prematurity) and those tested in previous studies (Muchnik et al., 1991). It is also possible that variability within each group masked differences that would have appeared with more extensive training. Nevertheless, the thresholds of the EB, SCm and SCy groups are consistent with results from previous gap detection threshold studies using a yes/no detection paradigm (Florentine et al., 2001; He et al., 1999) suggesting limits in the range of auditory sensory alterations that develop as a consequence early visual deprivation. In addition to potential differences in the groups, there are other differences between the current experimental design and that employed in the earlier report of gap detection in the blind (Muchnik et al., 1991). However, comparisons across studies using normal listeners generally report similar MGT values regardless of the protocol employed
Fig. 3. Gap detection parameter estimates. Resulting psychometric midpoint (a) and slope (b) values are plotted for each individual listener with group mean values indicated by a black bar. Midpoint values were calculated by solving for m under Eq. (1). Slope values were determined by estimating k from the adjusted logistic function.
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
(Green and Forrest, 1989). For example, Fitzgibbons (1983) reported approximately 3 ms thresholds using a 2AFC technique, and He et al. (1999) reported a mean threshold of 3 ms using a yes/no protocol. Further, the sighted control MGT values presented by Muchnik and colleagues (1991, Experiment B) were higher than results obtained from another group of similarly aged sighted individuals using the same method (Muchnik et al., 1985), and are well above reliable threshold estimates observed from the current task, and throughout the literature for noise periods longer than 25 ms. Conversely, the EB participants mean threshold of 4.8 ms reported by Muchnik is consistent with threshold estimates in the extant literatures and the current study. This consistency of gap detection thresholds across studies, using a variety of assessment techniques suggests that early blindness does not significantly alter gap detection thresholds. It is also worth noting that while studies of auditory sound localization in EB have indicated greater acuity for sound localization, the advantage appears to be greatest under monaural listening condition (Lessard et al., 1998). Therefore, it appears that EB subjects do not have substantial differences from SC individuals in between-channel auditory abilities either. Psychophysical measures of Gap detection thresholds have been associated subcortical structures (Phillips et al., 1997; Phillips, 1999). Because electrophysiological recordings of peripheral auditory nerve fibers typically match psychophysical estimates of MGT, it has been suggested that within-channel gap detection reflects the refresh rate of the subcortical afferents that feed into auditory cortex (Zhang et al., 1990; Frisina, 2001). However, intact auditory cortex is also necessary to sustain normal gap detection thresholds (Kelly et al., 1996). Electrophysiological studies have routinely indicated that primary auditory cortex is sensitive to thresholds of within channel gap detection. Interestingly, in humans, dipole source modeling suggests both within- and between-channel gap detection have nearly identical sources located near primary auditory cortices suggesting some common timing mechanisms for intraand inter-aural temporal resolution (Heinrich et al., 2004). Clearly, the inter-play between cortical and sub-cortical systems in sustaining overt detection of gaps is necessary. 4.1. Age effects on gap detection The influence of age on gap thresholds in the present study suggested a small but significant increase in the average gap duration with increased age. SCy participants were, on average, more accurate than SCm individuals at detecting gaps at all gap lengths except the longest gap of 7 ms. Despite similar performance curves (curve estimates) between young and old sighted listeners, the SCmÕs average logistic function was shifted to the right (indicating poorer detection) of the SCy groupÕs curve. Some support for an age-related decline comes from He et al. (1999), who observed no differences between young (mean age of 32) and old listeners (mean age of 71) except when the gap
5
was placed within the first 5% or last 95% of the noise markers. Moore et al. (1992) found some minor differences between young and old individuals and their MGT estimates, but attributed the majority of the variance to a few of the oldest subjects who had markedly different scores (see also, Barsz et al., 2002). The majority of gap detection scores in that study for the older group fell within the same range as the younger subjects. Snell (1997) however, reported that older subjects (mean age of 70) had a larger variability in scores, although mean scores (MGTs) were significantly lower than that of younger observers (mean age of 26). The authors suggested that age has subtle effects on gap detection (a general ‘‘slowing’’) but that the increased variability within the group leads to a trend in poorer temporal resolution. Generalizations between gap detection abilities of older listeners reported in the literature and the current paradigm are confounded by differences within mean age scores, but nevertheless suggests that age effects are present while detecting amplitude decays within a steady state signal. Further work is needed to determine the degree to which gap detection abilities are altered in normal aging. The results of the current study also suggest that young listeners have a less stringent response criterion than older listeners regardless of visual status. This is indicated by a significantly increased number of false positives relative to both SCm and EB groups. When false positives were taken into consideration by adjusting curve estimates, detection differences between SCy and SCm subjects existed at all gap conditions except the longest gap duration. Speculatively, this suggests that early blindness may have protective consequences against the potential minor alterations that appear to occur as a result of aging. This suggestion is supported by the nearly identical MGT values between EB and young control participants and slightly higher but non-significant MGT values in older control listeners. 4.2. Conclusion The current finding that EB subjects performed comparably to age-matched sighted subjects suggest that gap detection is not substantially altered in our sample of EB subjects and suggests that alterations in auditory perception in the blind reflect changes involving more complex auditory mechanisms than those tapped by gap detection. In light of these results combined with the robust effects found with more complex auditory perceptual tasks (e.g., Stevens and Weaver, in press), the auditory perceptual enhancements observed in EB individuals appear to reflect changes in functions at later stages of auditory perception (Niemeyer and Starlinger, 1981; Bavelier and Neville, 2002; Stevens and Weaver, in press). Further studies will be necessary in order to better specify the systems that contribute critically to gap detection thresholds, and how these systems differ from those involved in complex auditory tasks in which the EB subjects do excel.
6
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
Acknowledgements The authors thank the anonymous reviewers for their excellent comments and the OHSU Tinnitus Clinic for use of their sound-proof booth. This work was supported by a a grant from the National Eye Institute (EY-13682). References Amedi, A., Raz, N., Pianka, P., Malach, R., Zohary, E., 2003. Early ÔvisualÕ cortex activation correlates with superior verbal memory performance in the blind. Nat. Neurosci. 6 (7), 758–766. Barsz, K., Ison, J.R., Snell, K.B., Walton, J.P., 2002. Behavioral and neural measures of auditory temporal acuity in aging humans and mice. Neurobiol. Aging 23 (4), 565–578. Bavelier, D., Neville, H.J., 2002. Cross-modal plasticity: where and how? Nat. Rev. Neurosci. 3 (6), 443–452. Buus, S., Florentine, M., 1985. Gap detection in normal and impaired listeners: the effect of level and frequency. In: Michelsen, A. (Ed.), Time Resolution in Auditory Systems. Springer-Verlag, London, pp. 159–179. Buus, S., 2002. Psychophysical methods and other factors that affect the outcome of psychoacoustic measurements. In: Tranebjaerg, L., et al. (Eds.), Genetics and the Function of the Auditory System. Proceedings of the 19th Danavox Symposium, pp. 183–225. Fitzgibbons, P.J., 1983. Temporal gap detection in noise as a function of frequency, bandwidth, and level. J. Acoust. Soc. Am. 74 (1), 67–72. Florentine, M., Buus, S., Geng, W., 1999. Psychometric functions for gap detection in a yes–no procedure. J. Acoust. Soc. Am. 106 (6), 3512– 3520. Florentine, M., Marvit, P., Buss, B., 2001. Maximum-likelihood yes–no procedure for gap-detection: effect of track length. J. Am. Acad. Audiol. 12, 113–120. Frisina, R.D., 2001. Subcortical neural coding mechanisms for auditory temporal processing. Hear. Res. 158 (1–2), 1–27. Giraudi, D., Salvi, R., Henderson, D., Hamernik, R., 1980. Gap detection by the chinchilla. J. Acoust. Soc. Am. 68 (3), 802–806. Green, D.M., 1993. A maximum-likelihood method for estimating thresholds in a yes–no task. J. Acoust. Soc. Am. 93 (4 Pt 1), 2096–2105. Green, D.M., 1985. Temporal factors in psychoacoustics. In: Michelson, A. (Ed.), Time Resolution in Auditory Systems. Springer, London, pp. 122–140. Green, D.M., Forrest, T.G., 1989. Temporal gaps in noise and sinusoids. J. Acoust. Soc. Am. 86 (3), 961–970. Gou, Y., Burkard, R., 2002. Onset and offset from the inferior colliculus and auditory paired noiseburts: inner hair cell loss. Hear. Res. 171, 158–166. Gu, X., Green, D.M., 1994. Further studies of a maximum-likelihood yes– no procedure. J. Acoust. Soc. Am. 96, 93–101. He, N.J., Horwitz, A.R., Dubno, J.R., Mills, J.H., 1999. Psychometric functions for gap detection in noise measured from young and aged subjects. J. Acoust. Soc. Am. 106 (2), 966–978. Heinrich, A., Alain, C., Schneider, B.A., 2004. Within- and betweenchannel gap detection in the human auditory cortex. Neuroreport 15 (13), 2051–2056. Hugdahl, K., Ek, M., Takio, F., Rintee, T., Tuomainen, J., Haarala, C., Hamalainen, H., 2004. Blind individuals show enhanced perceptual and attentional sensitivity for identification of speech sounds. Brain Res. Cogn. Brain Res. 19 (1), 28–32.
Irwin, R.J., Purdy, S.C., 1982. The minimum detectable duration of auditory signals for normal and hearing-impaired listeners. J. Acoust. Soc. Am. 71 (4), 967–974. Kelly, J.B., Rooney, B.J., Phillips, D.P., 1996. Effects of bilateral auditory cortical lesions on gap-detection thresholds in the ferret (Mustela putorius). Behav. Neurosci. 110 (3), 542–550. Kujala, T., Lehtokos, A., Alho, K., Kekoni, J., Naatanen, R., 1997. Faster reaction times in the blind than sighted during bimodal divided attention. Acta Psychol. 96, 75–82. Lessard, N., Pare, M., Lepore, F., Lassonde, M., 1998. Early-blind human subjects localize sound sources better than sighted subjects. Nature 395 (6699), 278–280. Moore, B.C., Peters, R.W., Glasberg, B.R., 1992. Detection of temporal gaps in sinusoids by elderly subjects with and without hearing loss. J. Acoust. Soc. Am. 92 (4 Pt 1), 1923–1932. Muchnik, C., Efrati, M., Nemeth, E., Malin, M., Hildesheimer, M., 1991. Central auditory skills in blind and sighted subjects. Scand. Audiol. 20 (1), 19–23. Muchnik, C., Hildesheimer, M., Rubinstein, M., Sadeh, M., Shegter, Y., Shibolet, B., 1985. Minimal time interval in auditory temporal resolution. J. Audit. Res. 25 (4), 239–246. Neville, H.J., Bavelier, D., 2001. Variability of developmental plasticity. In: McClelland, J.L., Seigler, R.S. (Eds.), Mechanisms of Cognitive Development: Behavioral and Ellon Symposia on Cognition, pp. 271–301. Niemeyer, W., Starlinger, I., 1981. Do the blind hear better. Investigations on auditory processing in congenital or early acquired blindness. II. Central functions. Audiology 20 (6), 510–515. Phillips, D.P., 1999. Auditory gap detection, perceptual channels, and temporal resolution in speech perception. J. Am. Acad. Audiol. 10 (6), 343–354. Phillips, D.P., Taylor, T.L., Hall, S.E., Carr, M.M., Mossop, J.E., 1997. Detection of silent intervals between noises activating different perceptual channels: some properties of ‘‘central’’ auditory gap detection. J. Acoust. Soc. Am. 101 (6), 3694–3705. Plomp, R., 1964. Rate of decay of auditory sensation. J. Acoust. Soc. Am. 36, 277–282. Rauschecker, J.P., 1995. Compensatory plasticity and sensory substitution in the cerebral cortex. Trends Neurosci. 18 (1), 36–43. Roder, B., Rosler, F., 2003. Memory for environmental sounds in sighted, congenitally blind and late blind adults: evidence for cross-modal compensation. Int. J. Psychophysiol. 50 (1–2), 27–39. Roder, B., Rosler, F., Neville, H.J., 1999. Effects of interstimulus interval on auditory event-related potentials in congenitally blind and normally sighted humans. Neurosci. Lett. 264 (1–3), 53–56. Roberts, R.A., Lister, J.J., 2004. Effects of age and hearing loss on gap detection and the precedence effect: broadband stimuli. J. Speech Lang. Hear. Res. 47, 965–978. Snell, K.B., 1997. Age-related changes in temporal gap detection. J. Acoust. Soc. Am. 101 (4), 2214–2220. Starlinger, I., Niemeyer, W., 1981. Do the blind hear better. Investigations on auditory processing in congenital or early acquired blindness. I. Peripheral functions. Audiology 20 (6), 503–509. Stevens A.A., Weaver, K., in press. Auditory perceptual consolidation in early onset-blindness. Neuropsychologia. Werner, L.A., Folsom, R.C., Mancl, L.R., Syapin, C.L., 2001. Human auditory brainstem response to temporal gaps in noise. J. Speech Lang. Hear. Res. 44 (4), 737–750. Yates, J.T., Johnson, R.M., Starz, W.J., 1972. Loudness perception of the blind. Audiology 11 (5), 368–376. Zhang, W., Salvi, R.J., Saunders, S.S., 1990. Neural correlates of gap detection in auditory nerve fibers of the chinchilla. Hear. Res. 46 (3), 181–200.
Research paper
Auditory gap detection in the early blind Kurt E. Weaver a, Alexander A. Stevens a b
a,b,*
Department Of Behavioral Neuroscience, Oregon Health and Science University, Portland OR, USA Department of Psychiatry, Oregon Health and Science University, 3181 SW Sam Jackson Park Rd., CR139, Portland, OR 97239, USA Received 17 April 2005; received in revised form 17 April 2005; accepted 19 August 2005 Available online 26 October 2005
Abstract For blind individuals, audition provides critical information for interacting with the environment. Individuals blinded early in life (EB) typically show enhanced auditory abilities relative to sighted controls as measured by tasks requiring complex discrimination, attention and memory. In contrast, few deficits have been reported on tasks involving auditory sensory thresholds (e.g., Yates, J.T., Johnson, R.M., Starz, W.J., 1972. Loudness perception of the blind. Audiology 11(5), 368–376; Starlinger, I., Niemeyer, W., 1981. Do the blind hear better? Investigations on auditory processing in congenital or early acquired blindness. I. Peripheral functions. Audiology 20(6), 503–509). A study of gap detection stands at odds with this distinction [Muchnik, C., Efrati, M., Nemeth, E., Malin, M., Hildesheimer, M., 1991. Central auditory skills in blind and sighted subjects. Scand. Audiol. 20(1), 19–23]. In the current investigation we re-examined gap detection abilities in the EB using a single-interval, yes/no method. A group of younger sighted control individuals (SCy) was included in the analysis in addition to EB and sighted age matched control individuals (SCm) in order to examine the effect of age on gap detection performance. Estimates of gap detection thresholds for EB subjects were nearly identical to SCm subjects and slightly poorer relative to the SCy subjects. These results suggest some limits on the extent of auditory temporal advantages in the EB. Ó 2005 Elsevier B.V. All rights reserved. Keywords: Early blindness; Auditory temporal discrimination; Aging
1. Introduction Individuals blinded from birth critically depend upon audition as a primary source of information. A substantial amount of research from animals studies has suggested that in the absence of visual stimulation, auditory advantages develop that exceeds those of sighted individuals (For reviews, see Rauschecker, 1995; Bavelier and Neville, 2002, for a review). It has been argued however that advantages do not develop at all levels of auditory function; rather the early blind (EB) typically outperform
Abbreviations: EB, early blindness; SCm, age matched sighted controls; SCy, young sighted controls; MGT, minimum gap threshold; 2AFC, two alternative forced choice paradigm; ANOVA, analysis of variance * Corresponding author. Tel.: +503 494 0358; fax: +503 494 1209. E-mail address: [email protected] (A.A. Stevens). 0378-5955/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.heares.2005.08.002
sighted counterparts on auditory tasks that require complex perceptual discriminations. For instance, the EB consistently show better performance than SC individuals on tasks that examine auditory memory (Roder and Rosler, 2003), verbal memory (Amedi et al., 2003) and auditory attention (Kujala et al., 1997; Hugdahl et al., 2004), but have comparable auditory detection thresholds and intensity discrimination (Starlinger and Niemeyer, 1981; Yates et al., 1972). One area of particular interest underlying the development of auditory advantages in the blind is around the issue of temporal processing. Neville and Bavelier (2001) and Roder and colleagues (1999) have advanced the hypothesis that many of the auditory advantages possessed by the blind may stem from faster auditory temporal processing. Auditory temporal processing has different meanings at different levels within the auditory system.
2
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
For instance, gap detection tasks is thought to measure the shortest silent interval that can be detected within an auditory signal, a reflection of perceived within-channel sampling resolution (Plomp, 1964; Phillips, 1999; Roberts and Lister, 2004). Typically, trained subjects have thresholds for within-channel gaps on the order of 2– 4 ms in length, when sandwiched between broadband noise (Plomp, 1964; Green, 1985; Irwin and Purdy, 1982). Below this minimum gap threshold (MGT) the silent gap is masked by the post gap stimulus (Plomp, 1964; Gou and Burkard, 2002). In this within-channel paradigm, pre and post-gap noise bursts share identical spectral properties and therefore stimulate the same set of auditory neurons (in contrast, between or across-channel gap detection varies the spectral content of the noise markers so that leading and trailing markers do not overlap and thus activate different perceptual channels; see Phillips et al., 1997; Buus and Florentine, 1985). Converging evidence from single unit and auditory brain stem recordings suggests that, because listeners must simply detect the absence of afferent activity, within-channel gap detection resolves the discharge rates of peripheral relays, including auditory brainstem nuclei that feed into the cortex (Werner et al., 2001; Giraudi et al., 1980; Frisina, 2001; Zhang et al., 1990; Phillips et al., 1997). However, in animal models damage to primary auditory cortex disrupts gap detection, requiring longer duration gaps for accurate responding (Kelly et al., 1996). In EB subjects, MGTÕs estimates have been reported to be significantly shorter than SC, suggesting that loss of vision early in life results in superior auditory temporal resolution (Muchnik et al., 1991). In this task, listeners were required to detect a gap placed within one of two periods of noise using an adaptive 2 alternative forced choice paradigm (2AFC). This evidence suggests that early visual deprivation enhances the sampling resolution of the auditory pathway. However, the best MGT for the SC group reported in the gap detection task was 13.8 ms when noise period durations were set at 85 ms while the early blind threshold was 4.8 ms (Muchnik et al., 1991). The SC groupÕs threshold was substantially longer than MGT thresholds reported in the literature (typically 2–4 ms; Plomp, 1964; Green, 1985; Irwin and Purdy, 1982). Indeed the control groupÕs threshold in Muchnik et al. (1991) was approximately 10 ms longer than that of sighted controls performance on the same task reported previously by the same investigator (Muchnik et al., 1985). This raises the possibility that the sighted control group in the later study was not representative of normal sighted performance on the task. Given the inconsistencies in these earlier reports and the importance of understanding the different levels at which auditory temporal resolution may influence EB auditory perception, the goal of the present study was to reexamine auditory temporal resolution in EB subjects using a withinchannel gap detection task. We determined MGTÕs using a single-burst, psychometric yes/no paradigm (Florentine
et al., 1999, 2001; He et al., 1999). Further, because both the EB and SC groups in the current investigation were much older than those reported by Muchnik et al., a younger group of sighted individuals was included in the analysis to examine the influence of age. 2. Materials and methods 2.1. Subjects Ten early blind (EB, 6 female; mean age 51.5, range 42–57), 10 sighted, age-matched controls (SCm, 6 female, mean age 52.2, range of 44–64) and 10 young sighted controls (SCy, 4 females, mean age 27.9, range 24–33) were examined. Exclusion criteria for the blind included: age of blindness onset occurring after the first year of life, and light sensitivity defined as form or color vision. Additional exclusion criteria for all subjects included previously diagnosed hearing damage or loss, comorbid psychiatric or neurological disease, and drug or alcohol abuse within the past five years. Cause of blindness in all EB listeners stemmed from peripheral damage due to complications of retinopathy of prematurity. One EB participant had diffuse light sensitivity. Subjects reported normal hearing and all signed an informed consent prior to the testing session. The EB and SCm subjects also served as subjects in a study of auditory backward masking, reported elsewhere and were screened for frequency discrimination (Stevens and Weaver, in press). All experimental procedures were approved by the Institutional Review Board of Oregon Health & Science University.
2.2. Gap detection Subjects were presented with a single interval of band-pass noise (center frequency of 2500 Hz. with a band width of 2000 Hz) with an overall duration of 400 ms. Noise bursts were presented at 75 dB SPL with a 10 ms rise/fall time at the initial and final periods of the noise period, and a rectangular silent gap. Silent, gaps with a duration of 1–7 ms (inclusive of rise/fall time) were randomly presented in some trials. The onset of a gap occurred randomly between the first and last 80 ms of the stimulus period. Subjects indicated on a button box whether they detected a gap and no feedback was given. Thirty trials were presented at each gap duration as well as the no-gap condition (catch trials), resulting in a total of 240 trials across the session. The experiment was divided up into blocks of 100 trials and subjects were allowed to rest between each block if needed. A practice session was run prior to the experimental session where subjects were presented with 5 trials of the no gap and 5 trials of the 7 ms condition.
2.3. Data analysis To measure sensitivity differences and control for differences in false positives rates (i.e., a yes response on a no gap trial) between groups, psychometric curves were fitted to the individual subject data using a logistic regression function described by Green (1993) that incorporates false positive frequency,
P ðyesÞ ¼ a þ ð1 aÞ=ð1 þ ekðxmÞ Þ;
ð1Þ
where the probability of a ‘‘yes’’ response, P(yes), at a gap length x, reflect the rate of false positive responses a, and k is a free parameter that describes the slope of the fitted logistic curve with an estimated midpoint, m. For each listener, k and m were initially determined using a curve-estimation procedure that plots the best-fit logistic function against the measured data (cf., He et al., 1999). Midpoint values were used as an estimate of MGT (Buus, 2002). Because detection estimates are influenced by listenersÕ response criteria (i.e., false positives, see Gu and Green, 1994), Eq. (1) was used to solve for the midpoint parameter m, taking into account differences of individual false-alarm rates.
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
3
2.4. Materials Stimuli were generated using Tucker Davis Technologies (Gainsville, FL) Real-Time Processor System 3 and presented using Microsoft Visual Basic Software (Seattle, WA). Intensity of stimuli was controlled using System 3 attenuators and all stimuli were presented through Senheisser HD265 dynamic range headphones (Old Lyme, CT) which were calibrated before each session using a UEI DSM 100 digital sound meter (Beaverton, OR). The task was conducted in a 2-walled sound-proof booth.
3. Results 3.1. Estimating gap detection thresholds Fig. 1 shows logistic curve estimates for a representative SCm, EB and SCy subject. The open squares represent the measured data with the line showing the best-fit logistic function. Curve fits were significant for all subjects (P values <0.01; analysis of variance F test for Regression Relation) and showed high correlation coefficients (r2 > 0.90, see Fig. 1). We next recast individual logistic functions according to Eq. (1) in order to control for differences in false positive rates. This analysis resulted in similar performance curves averaged across individual groups (Fig. 2(a)) and are similar to previous estimates of MGT using similar protocols on audiologically normal individuals (cf., He et al., 1999, Experiment 1). Within all three groups, detection improved as the gap length increased. We compared group performance across gap conditions using a repeated measures analysis of variance (ANOVA) and a Geiser–Greenhouse correction for violations of sphericity (uncorrected degrees of freedom are reported for clarity). This analysis resulted in a main effect of group [F2,27 = 6.127, P = 0.001], a main effect of gap condition, [F1.51,40.88 = 1319.629, P < 0.001] and a significant interaction of group and gap [F3.028,40.88 = 4.141, P = 0.012]. Post hoc t tests using a Sˇida´k correction for multiple corrections (P-values <0.05) were used to assess group differences at each gap duration.
Fig. 2. Gap detection performance in the EB, SCm and SCy. Proportion of gaps detected (a yes response) is plotted as a function of gap length for all three groups. (a) Psychometric curves were adjusted taking into consideration false positive rates according to Eq. (1) and averaged across each group. *Denotes significant differences at the 0.05 level between EB and SCy listeners. No significant performance differences appeared between EB and SCm subjects. Mean SCm scores were however significantly lower than SCy at all gap lengths except for the 7 ms condition. (b) Raw, unadjusted performance curves are shown for each group of listeners. Error bars represent standard error.
These detected no statistically reliable differences between SCm and EB but significant performance differences between the EB and SCy appeared on trials with no gap
Fig. 1. Logistic curve-estimation. The best-fit logistic function (lines) was determined for each subjectÕs measured results (open squares) based upon a least squares method. A representative SCm, EB and SCy subject are shown with corresponding r2, m and k values displayed in the inset. Values for the midpoint (m) and the slope (k) were estimated based upon the resulting logistic function.
4
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
and 1 ms gap, reflecting an elevated proportion of ‘‘yes’’ responses on these trials (Fig. 2), which on the no-gap trials, provide an estimate of false alarm rate. Observed proportion yes responses were reliably different between SCy and SCm listeners at all sampled gap lengths except the 7 ms condition. Fig. 2(b) (inset) show averaged group results uncorrected for false alarm rate. 3.2. Threshold estimates and slope parameters Estimates of the 50% threshold value using Eq. (1) to account for the false positive rate for each subject was calculated as a measure of gap threshold (MGT; Buus, 2002). The estimated MGT values for each individual and the mean scores are shown in Fig. 3(a) (mean values for: ms SCm: 3.39 ± 0.17; EB: 3.40 ± 0.14; SCy: 3.76 ± 0.09). A one-way analysis of variance comparing MGT values between groups failed to reach significance [F2,27 = 2.274, P = 0.122] despite a mildly larger mean MGT estimate for the SCm group relative to either group of listeners. A lack of a significant effect must be taken with some caution, however, due to the low sample size within each group. The slope (k) was estimated for each subject (Fig. 3(b)). The SCy group had a somewhat steeper slope estimate (EB: 1.37 ± 0.03; SCm: 1.40 ± 0.03; SCy: 1.56 ± 0.03) but in general, the k estimates of the SCm and SCy were similar to old/young differences reported in the literature (He et al., 1999). The omnibus F test did revealed a statistically reliable difference among the groups [F2,27 = 11.609, P = 0.001] and post hoc tests (P < 0.05) revealed that a reliable difference reflecting a steeper slope estimate for the SCy participants compared to both the EB and the SCm, who did not differ.
4. Discussion The current results suggest that EB individuals differ little from their age-matched sighted counterparts (SCm) in their ability to detect gaps embedded in noise. Additionally, although statistically significant performance differences appeared between the SCy and EB at the no gap and 1 ms gap conditions (Fig. 2) and slopes of the logistic curves (Fig. 3(b)), they appear to reflect a greater false alarm rate in the SCy group; higher rates of indicating the presence of a gap even in their absence. Additionally, no significant differences appeared between sighted listeners (young or old) and blind listeners in MGT estimates. These results must be interpreted with caution due to the relatively small sample sized of the groups, and potential differences in gap detection thresholds that may be associated with differences in the etiology of blindness in the current EB group (all 10 had retinopathy due to prematurity) and those tested in previous studies (Muchnik et al., 1991). It is also possible that variability within each group masked differences that would have appeared with more extensive training. Nevertheless, the thresholds of the EB, SCm and SCy groups are consistent with results from previous gap detection threshold studies using a yes/no detection paradigm (Florentine et al., 2001; He et al., 1999) suggesting limits in the range of auditory sensory alterations that develop as a consequence early visual deprivation. In addition to potential differences in the groups, there are other differences between the current experimental design and that employed in the earlier report of gap detection in the blind (Muchnik et al., 1991). However, comparisons across studies using normal listeners generally report similar MGT values regardless of the protocol employed
Fig. 3. Gap detection parameter estimates. Resulting psychometric midpoint (a) and slope (b) values are plotted for each individual listener with group mean values indicated by a black bar. Midpoint values were calculated by solving for m under Eq. (1). Slope values were determined by estimating k from the adjusted logistic function.
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
(Green and Forrest, 1989). For example, Fitzgibbons (1983) reported approximately 3 ms thresholds using a 2AFC technique, and He et al. (1999) reported a mean threshold of 3 ms using a yes/no protocol. Further, the sighted control MGT values presented by Muchnik and colleagues (1991, Experiment B) were higher than results obtained from another group of similarly aged sighted individuals using the same method (Muchnik et al., 1985), and are well above reliable threshold estimates observed from the current task, and throughout the literature for noise periods longer than 25 ms. Conversely, the EB participants mean threshold of 4.8 ms reported by Muchnik is consistent with threshold estimates in the extant literatures and the current study. This consistency of gap detection thresholds across studies, using a variety of assessment techniques suggests that early blindness does not significantly alter gap detection thresholds. It is also worth noting that while studies of auditory sound localization in EB have indicated greater acuity for sound localization, the advantage appears to be greatest under monaural listening condition (Lessard et al., 1998). Therefore, it appears that EB subjects do not have substantial differences from SC individuals in between-channel auditory abilities either. Psychophysical measures of Gap detection thresholds have been associated subcortical structures (Phillips et al., 1997; Phillips, 1999). Because electrophysiological recordings of peripheral auditory nerve fibers typically match psychophysical estimates of MGT, it has been suggested that within-channel gap detection reflects the refresh rate of the subcortical afferents that feed into auditory cortex (Zhang et al., 1990; Frisina, 2001). However, intact auditory cortex is also necessary to sustain normal gap detection thresholds (Kelly et al., 1996). Electrophysiological studies have routinely indicated that primary auditory cortex is sensitive to thresholds of within channel gap detection. Interestingly, in humans, dipole source modeling suggests both within- and between-channel gap detection have nearly identical sources located near primary auditory cortices suggesting some common timing mechanisms for intraand inter-aural temporal resolution (Heinrich et al., 2004). Clearly, the inter-play between cortical and sub-cortical systems in sustaining overt detection of gaps is necessary. 4.1. Age effects on gap detection The influence of age on gap thresholds in the present study suggested a small but significant increase in the average gap duration with increased age. SCy participants were, on average, more accurate than SCm individuals at detecting gaps at all gap lengths except the longest gap of 7 ms. Despite similar performance curves (curve estimates) between young and old sighted listeners, the SCmÕs average logistic function was shifted to the right (indicating poorer detection) of the SCy groupÕs curve. Some support for an age-related decline comes from He et al. (1999), who observed no differences between young (mean age of 32) and old listeners (mean age of 71) except when the gap
5
was placed within the first 5% or last 95% of the noise markers. Moore et al. (1992) found some minor differences between young and old individuals and their MGT estimates, but attributed the majority of the variance to a few of the oldest subjects who had markedly different scores (see also, Barsz et al., 2002). The majority of gap detection scores in that study for the older group fell within the same range as the younger subjects. Snell (1997) however, reported that older subjects (mean age of 70) had a larger variability in scores, although mean scores (MGTs) were significantly lower than that of younger observers (mean age of 26). The authors suggested that age has subtle effects on gap detection (a general ‘‘slowing’’) but that the increased variability within the group leads to a trend in poorer temporal resolution. Generalizations between gap detection abilities of older listeners reported in the literature and the current paradigm are confounded by differences within mean age scores, but nevertheless suggests that age effects are present while detecting amplitude decays within a steady state signal. Further work is needed to determine the degree to which gap detection abilities are altered in normal aging. The results of the current study also suggest that young listeners have a less stringent response criterion than older listeners regardless of visual status. This is indicated by a significantly increased number of false positives relative to both SCm and EB groups. When false positives were taken into consideration by adjusting curve estimates, detection differences between SCy and SCm subjects existed at all gap conditions except the longest gap duration. Speculatively, this suggests that early blindness may have protective consequences against the potential minor alterations that appear to occur as a result of aging. This suggestion is supported by the nearly identical MGT values between EB and young control participants and slightly higher but non-significant MGT values in older control listeners. 4.2. Conclusion The current finding that EB subjects performed comparably to age-matched sighted subjects suggest that gap detection is not substantially altered in our sample of EB subjects and suggests that alterations in auditory perception in the blind reflect changes involving more complex auditory mechanisms than those tapped by gap detection. In light of these results combined with the robust effects found with more complex auditory perceptual tasks (e.g., Stevens and Weaver, in press), the auditory perceptual enhancements observed in EB individuals appear to reflect changes in functions at later stages of auditory perception (Niemeyer and Starlinger, 1981; Bavelier and Neville, 2002; Stevens and Weaver, in press). Further studies will be necessary in order to better specify the systems that contribute critically to gap detection thresholds, and how these systems differ from those involved in complex auditory tasks in which the EB subjects do excel.
6
K.E. Weaver, A.A. Stevens / Hearing Research 211 (2006) 1–6
Acknowledgements The authors thank the anonymous reviewers for their excellent comments and the OHSU Tinnitus Clinic for use of their sound-proof booth. This work was supported by a a grant from the National Eye Institute (EY-13682). References Amedi, A., Raz, N., Pianka, P., Malach, R., Zohary, E., 2003. Early ÔvisualÕ cortex activation correlates with superior verbal memory performance in the blind. Nat. Neurosci. 6 (7), 758–766. Barsz, K., Ison, J.R., Snell, K.B., Walton, J.P., 2002. Behavioral and neural measures of auditory temporal acuity in aging humans and mice. Neurobiol. Aging 23 (4), 565–578. Bavelier, D., Neville, H.J., 2002. Cross-modal plasticity: where and how? Nat. Rev. Neurosci. 3 (6), 443–452. Buus, S., Florentine, M., 1985. Gap detection in normal and impaired listeners: the effect of level and frequency. In: Michelsen, A. (Ed.), Time Resolution in Auditory Systems. Springer-Verlag, London, pp. 159–179. Buus, S., 2002. Psychophysical methods and other factors that affect the outcome of psychoacoustic measurements. In: Tranebjaerg, L., et al. (Eds.), Genetics and the Function of the Auditory System. Proceedings of the 19th Danavox Symposium, pp. 183–225. Fitzgibbons, P.J., 1983. Temporal gap detection in noise as a function of frequency, bandwidth, and level. J. Acoust. Soc. Am. 74 (1), 67–72. Florentine, M., Buus, S., Geng, W., 1999. Psychometric functions for gap detection in a yes–no procedure. J. Acoust. Soc. Am. 106 (6), 3512– 3520. Florentine, M., Marvit, P., Buss, B., 2001. Maximum-likelihood yes–no procedure for gap-detection: effect of track length. J. Am. Acad. Audiol. 12, 113–120. Frisina, R.D., 2001. Subcortical neural coding mechanisms for auditory temporal processing. Hear. Res. 158 (1–2), 1–27. Giraudi, D., Salvi, R., Henderson, D., Hamernik, R., 1980. Gap detection by the chinchilla. J. Acoust. Soc. Am. 68 (3), 802–806. Green, D.M., 1993. A maximum-likelihood method for estimating thresholds in a yes–no task. J. Acoust. Soc. Am. 93 (4 Pt 1), 2096–2105. Green, D.M., 1985. Temporal factors in psychoacoustics. In: Michelson, A. (Ed.), Time Resolution in Auditory Systems. Springer, London, pp. 122–140. Green, D.M., Forrest, T.G., 1989. Temporal gaps in noise and sinusoids. J. Acoust. Soc. Am. 86 (3), 961–970. Gou, Y., Burkard, R., 2002. Onset and offset from the inferior colliculus and auditory paired noiseburts: inner hair cell loss. Hear. Res. 171, 158–166. Gu, X., Green, D.M., 1994. Further studies of a maximum-likelihood yes– no procedure. J. Acoust. Soc. Am. 96, 93–101. He, N.J., Horwitz, A.R., Dubno, J.R., Mills, J.H., 1999. Psychometric functions for gap detection in noise measured from young and aged subjects. J. Acoust. Soc. Am. 106 (2), 966–978. Heinrich, A., Alain, C., Schneider, B.A., 2004. Within- and betweenchannel gap detection in the human auditory cortex. Neuroreport 15 (13), 2051–2056. Hugdahl, K., Ek, M., Takio, F., Rintee, T., Tuomainen, J., Haarala, C., Hamalainen, H., 2004. Blind individuals show enhanced perceptual and attentional sensitivity for identification of speech sounds. Brain Res. Cogn. Brain Res. 19 (1), 28–32.
Irwin, R.J., Purdy, S.C., 1982. The minimum detectable duration of auditory signals for normal and hearing-impaired listeners. J. Acoust. Soc. Am. 71 (4), 967–974. Kelly, J.B., Rooney, B.J., Phillips, D.P., 1996. Effects of bilateral auditory cortical lesions on gap-detection thresholds in the ferret (Mustela putorius). Behav. Neurosci. 110 (3), 542–550. Kujala, T., Lehtokos, A., Alho, K., Kekoni, J., Naatanen, R., 1997. Faster reaction times in the blind than sighted during bimodal divided attention. Acta Psychol. 96, 75–82. Lessard, N., Pare, M., Lepore, F., Lassonde, M., 1998. Early-blind human subjects localize sound sources better than sighted subjects. Nature 395 (6699), 278–280. Moore, B.C., Peters, R.W., Glasberg, B.R., 1992. Detection of temporal gaps in sinusoids by elderly subjects with and without hearing loss. J. Acoust. Soc. Am. 92 (4 Pt 1), 1923–1932. Muchnik, C., Efrati, M., Nemeth, E., Malin, M., Hildesheimer, M., 1991. Central auditory skills in blind and sighted subjects. Scand. Audiol. 20 (1), 19–23. Muchnik, C., Hildesheimer, M., Rubinstein, M., Sadeh, M., Shegter, Y., Shibolet, B., 1985. Minimal time interval in auditory temporal resolution. J. Audit. Res. 25 (4), 239–246. Neville, H.J., Bavelier, D., 2001. Variability of developmental plasticity. In: McClelland, J.L., Seigler, R.S. (Eds.), Mechanisms of Cognitive Development: Behavioral and Ellon Symposia on Cognition, pp. 271–301. Niemeyer, W., Starlinger, I., 1981. Do the blind hear better. Investigations on auditory processing in congenital or early acquired blindness. II. Central functions. Audiology 20 (6), 510–515. Phillips, D.P., 1999. Auditory gap detection, perceptual channels, and temporal resolution in speech perception. J. Am. Acad. Audiol. 10 (6), 343–354. Phillips, D.P., Taylor, T.L., Hall, S.E., Carr, M.M., Mossop, J.E., 1997. Detection of silent intervals between noises activating different perceptual channels: some properties of ‘‘central’’ auditory gap detection. J. Acoust. Soc. Am. 101 (6), 3694–3705. Plomp, R., 1964. Rate of decay of auditory sensation. J. Acoust. Soc. Am. 36, 277–282. Rauschecker, J.P., 1995. Compensatory plasticity and sensory substitution in the cerebral cortex. Trends Neurosci. 18 (1), 36–43. Roder, B., Rosler, F., 2003. Memory for environmental sounds in sighted, congenitally blind and late blind adults: evidence for cross-modal compensation. Int. J. Psychophysiol. 50 (1–2), 27–39. Roder, B., Rosler, F., Neville, H.J., 1999. Effects of interstimulus interval on auditory event-related potentials in congenitally blind and normally sighted humans. Neurosci. Lett. 264 (1–3), 53–56. Roberts, R.A., Lister, J.J., 2004. Effects of age and hearing loss on gap detection and the precedence effect: broadband stimuli. J. Speech Lang. Hear. Res. 47, 965–978. Snell, K.B., 1997. Age-related changes in temporal gap detection. J. Acoust. Soc. Am. 101 (4), 2214–2220. Starlinger, I., Niemeyer, W., 1981. Do the blind hear better. Investigations on auditory processing in congenital or early acquired blindness. I. Peripheral functions. Audiology 20 (6), 503–509. Stevens A.A., Weaver, K., in press. Auditory perceptual consolidation in early onset-blindness. Neuropsychologia. Werner, L.A., Folsom, R.C., Mancl, L.R., Syapin, C.L., 2001. Human auditory brainstem response to temporal gaps in noise. J. Speech Lang. Hear. Res. 44 (4), 737–750. Yates, J.T., Johnson, R.M., Starz, W.J., 1972. Loudness perception of the blind. Audiology 11 (5), 368–376. Zhang, W., Salvi, R.J., Saunders, S.S., 1990. Neural correlates of gap detection in auditory nerve fibers of the chinchilla. Hear. Res. 46 (3), 181–200.