Cortical basis for dichotic pitch perception in developmental dyslexia

Brain & Language 123 (2012) 104–112

Contents lists available at SciVerse ScienceDirect

Brain & Language journal homepage: www.elsevier.com/locate/b&l

Cortical basis for dichotic pitch perception in developmental dyslexia Marita Partanen a,⇑, Kevin Fitzpatrick a, Burkhard Mädler b, Dorothy Edgell c, Bruce Bjornson a, Deborah E. Giaschi a a

University of British Columbia, British Columbia Children’s Hospital, 4480 Oak St., Vancouver, BC, Canada V6H 3V4 University of Bonn, Department of Neurosurgery, Sigmund-Freud-Strasse 25, 53105 Bonn, Germany c University of Victoria, 3800 Finnerty Rd., Victoria, BC, Canada V8P 5C2 b

a r t i c l e

i n f o

Article history: Accepted 5 September 2012 Available online 6 October 2012 Keywords: Dyslexia Temporal processing Auditory perception Phonological processing Functional magnetic resonance imaging

a b s t r a c t The current study examined auditory processing deficits in dyslexia using a dichotic pitch stimulus and functional MRI. Cortical activation by the dichotic pitch task occurred in bilateral Heschl’s gyri, right planum temporale, and right superior temporal sulcus. Adolescents with dyslexia, relative to age-matched controls, illustrated greater activity in left Heschl’s gyrus for random noise, less activity in right Heschl’s gyrus for all auditory conditions, and less activity in right superior temporal sulcus for a dichotic melody. Subsequent analyses showed that these group differences were attributable to dyslexic readers who performed poorly on the psychophysical task. Furthermore, behavioral performance on phonological reading was correlated to activity from dichotic conditions in right Heschl’s gyrus and right superior temporal sulcus. It is postulated that these differences between reader groups is primarily due to a noise exclusion deficit shown previously in dyslexia. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction Developmental dyslexia is defined as difficulty learning to read, despite average intelligence, lack of sensory impairments, and adequate access to educational resources (Lyon, Shaywitz, & Shaywitz, 2003). This learning disability affects 5–12% of the population (Katusic, Colligan, Barbaresi, Schaid, & Jacobsen, 2001), and while the etiology is unknown, it appears to have genetic and neurobiological bases (Fisher & DeFries, 2002; Habib, 2000). It is well established that most persons with dyslexia experience difficulty with phonological encoding and decoding (Snowling, 1981; Stanovich & Siegel, 1994). Some children may also have problems with orthographic processing, which includes awareness of irregular words that do not have phonological representations, such as ‘‘yacht’’ (Castles & Coltheart, 1993; Edwards & Hogben, 1999). The existence of a phonological deficit in dyslexia is not disputed; however, it may be one aspect of a ‘‘multifactorial’’ disorder (Caylak, 2011). Another component to reading may be rapid auditory processing (reviewed in Caylak, 2011; Farmer & Klein, 1995; Hämäläinen, Salminen, & Leppänen, 2012). Tallal (1980) originally hypothesized that young children who have difficulty processing rapidly presented information, such as speech sounds, have subsequent difficulties with phonemic awareness and reading. Auditory processing deficits in dyslexia have been shown in many studies measuring ⇑ Corresponding author. Address: 4480 Oak St., Room A146, Vancouver, BC, Canada V6H 3V4. Fax: +1 604 875 2683. E-mail address: [email protected] (M. Partanen). 0093-934X/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.bandl.2012.09.002

temporal ordering of tones (e.g., Tallal, 1980), auditory gap detection (e.g., Boets, Wouters, van Wieringen, & Ghesquière, 2007), frequency-modulated tone discrimination and detection (e.g., Edwards et al., 2004; Talcott et al., 2003), and binaural pitch perception (e.g., Dougherty, Cynader, Bjornson, Edgell, & Giaschi, 1998; Edwards et al., 2004; McAnally & Stein, 1996). However, the relationship between reading and auditory temporal processing continues to be an area of debate (Landerl & Willburger, 2010; Ramus, 2003; Rosen & Manganari, 2001). In adults with normal hearing, rapid auditory processing occurs primarily in the left temporal cortex, including Heschl’s gyrus and superior temporal gyrus, whereas spectral or melody perception occurs in the right temporal cortex (Warrier et al., 2009; Zatorre & Belin, 2001; Zatorre, Belin, & Penhune, 2002). Furthermore, larger left Heschl’s gyri were correlated with greater activation in this region when adults listened to rapidly presented noise stimuli (Warrier et al., 2009), and thicker gray matter in left Heschl’s gyrus was correlated with better performance on a frequency-modulated tone detection task in children (Sutherland et al., 2012). In contrast, right Heschl’s gyrus showed a positive relationship between volume and spectral-related activity (Warrier et al., 2009). These anatomical and functioning neuroimaging studies show that although both hemispheres are involved in auditory processing, the left hemisphere is predominantly implicated for rapid auditory perception and the right hemisphere for melody perception. Only a few functional neuroimaging studies have examined auditory temporal processing deficits in dyslexia. Temple et al. (2000) showed that rapidly changing acoustic stimuli activated

M. Partanen et al. / Brain & Language 123 (2012) 104–112

the left superior and middle frontal gyri for average reading adults, but not for adults with dyslexia. Similar results were shown for children with dyslexia; after completing a reading remediation program, children showed increased activity in several regions used for rapid auditory processing, including the frontal and parietal lobes as well as the thalamus (Gaab, Gabrieli, Deutsch, Tallal, & Temple, 2007). Two recent studies with German speakers have examined auditory processing in adults with and without dyslexia. Steinbrink, Ackermann, Lachmann, and Riecker (2009) presented click stimuli or syllables at rates between 1 and 9 Hz. In average readers, with increased rate presentation, increased activation was shown in left insular cortex, right cerebellum, and thalamus and decreased activation was shown in right insular cortex. In contrast, the dyslexic group did not show this pattern of results in bilateral insula when syllables were presented. The authors concluded that dyslexic readers have speech-specific auditory temporal processing deficits; however, since their click and syllable stimuli differed in signal complexity, they did not rule out that stimulus characteristics also contributed to their results. In another study, Steinbrink, Groth, Lachmann, and Riecker (2012) used long and short German vowels to measure auditory temporal processing and phonological processing. In average readers, activation from the temporal task was shown in bilateral superior temporal gyri, bilateral anterior insula, and left inferior frontal gyrus. Although the dyslexic readers had lower discrimination accuracy than controls on their temporal task, there were no significant differences between groups in cortical activity. Subsequent analyses showed that low performers on the discrimination task (<60% accuracy; all dyslexic readers) had decreased activity in bilateral insular cortex and left inferior frontal gyrus in comparison to high performers (>90% accuracy; five average readers, one dyslexic reader). Given that low performers were all dyslexic readers, this suggests that some adults with dyslexia have impairments in insular cortex and left inferior frontal gyrus in response to rapid auditory stimuli. In summary, the functional neuroimaging studies have shown that dyslexic readers have impaired activation in response to rapid auditory processing tasks and these deficits are distributed throughout the brain, including frontal (Gaab et al., 2007; Steinbrink et al., 2012; Temple et al., 2000), parietal (Gaab et al., 2007), and insular (Steinbrink et al., 2009, 2012) cortices, as well as the thalamus (Gaab et al., 2007). Further evidence, however, is needed to determine the neurobiological basis of auditory processing deficits in dyslexia, particularly in children and adolescents. We have previously shown that children with dyslexia have difficulty with dichotic pitch perception (Dougherty et al., 1998; Edwards et al., 2004). Dichotic pitch is an auditory stimulus created using two white noise sources that are presented to both ears. The fusing of binaural cues is needed to perceive pitch, whereas monaural cues on their own cannot be used. Our stimulus uses an ongoing interaural time difference (ITD) of 0.6 ms and thus, it requires sensitivity to the rapid temporal structure of the acoustic signals. It also probes the auditory system’s ability to extract signals from noise and to compute sound location. Therefore, the dichotic pitch task is similar to previous dyslexia research as it examines rapid auditory processing (i.e., the ongoing ITD) and binaural pitch perception (e.g., McAnally & Stein, 1996). Neuroimaging results with adults illustrate that dichotic pitch stimuli activate bilateral Heschl’s gyri, planum temporale, and planum polare (Garcia, Hall, & Plack, 2010; Puschmann, Uppenkamp, Kollmeier, & Thiel, 2010), as well as right superior temporal sulcus (Giaschi, Bjornson, Dougherty, & Au Young, 2000). In the current study, we measured dichotic pitch processing in adolescents with dyslexia and adolescents with average reading ability using functional magnetic resonance imaging (fMRI) to assist in elucidating the neural basis of dyslexia. Our first goal was to establish that a similar dichotic pitch perception system is in-

105

volved in adolescents because previous studies on auditory development reported more regions of activity in adults than in children (e.g., Koelsch, Fritz, Schulze, Alsop, & Schlaug, 2005). Our second goal was to examine deficits in dyslexia and we hypothesized that adolescents with dyslexia would have altered activity in regions such as Heschl’s gyrus, planum temporale, planum polare, and superior temporal sulcus, relative to adolescents with average reading ability (Garcia et al., 2010; Giaschi et al., 2000; Puschmann et al., 2010). 2. Methods 2.1. Participants Participants included 10 adolescents with average reading ability (6 male, 4 female) and 9 adolescents with dyslexia (6 male, 3 female) between 12 and 16 years old. All adolescents were righthanded and had a mean IQ score greater than or equal to 1 SD below the standardized norm on the Wechsler Intelligence Scale for Children-IV (WISC-IV) (M = 10, SD = 3; i.e., all adolescents had scaled scores greater than 7; Wechsler, 2003). The mean IQ score was derived from the Block Design and Vocabulary subtests, which measured perceptual reasoning and verbal comprehension, respectively. Adolescents were grouped into average or dyslexic reading groups based on their performance on the Gray Oral Reading Test, 4th edition (GORT-4) (Wiederholt & Bryant, 2001). Reading fluency was established by the rate and accuracy of reading aloud short paragraphs. Average readers were characterized by a fluency score greater than or equal to 1 SD below the standardized norm (M = 10, SD = 3; i.e., scaled score of 7 or greater) and dyslexic readers were characterized by a fluency score less than or equal to 1.5 SD below the standardized norm (i.e., scaled score of 5 or less). Groups were similar on age and IQ measures, while there were significant differences between groups on the GORT-4 reading measures (see Table 1). Additionally, all participants had normal hearing with a threshold of 25 dB HL or less in both ears at 500 and 1000 Hz, which was assessed using a Maico MA-39 audiometer. 2.2. Behavioral tasks 2.2.1. Reading measures A modified version of the Coltheart and Leahy (1996) reading lists was used to assess phonological and orthographic reading ability. Participants were asked to read aloud single words and non-words, which were printed on cards and placed in random order in a book. There were 30 words in each of three lists: regular words (e.g., plant), pronounceable non-words (e.g., norf) to measure phonological reading ability, and irregular words (e.g., yacht) to measure orthographic reading ability. Each participant read the words in the same order at their own pace and the number of errors was recorded. 2.2.2. Dichotic pitch Auditory temporal processing was assessed with a dichotic pitch task (Dougherty et al., 1998), which has been shown to differentiate between average and dyslexic readers (Dougherty et al., 1998; Edwards et al., 2004). Dichotic pitch stimuli were generated by filtering two independent flat-amplitude noise sources. One source was band-pass filtered to create the signal tones, while the other source was notch filtered to create the background noise. The signal tones had an ongoing ITD of ±0.6 ms and the background noise had an ITD of 0 ms, which created the perception of tones that appeared to come from the left or right and noise from the middle of the head. The complementary signal and noise filters were modified to adjust the signal-to-noise ratio (SNR) from 0

106

M. Partanen et al. / Brain & Language 123 (2012) 104–112

Table 1 Participant characteristics for age, IQ, reading, and dichotic pitch. Measure

Average reader group

Dyslexic reader group

M

SD

M

SD

Age (years)a

14.92

1.07

14.43

1.72

IQ (WISC-IV)b Verbal comprehensiona Perceptual reasoning Average IQa

11.00 10.80 11.00

2.26 2.53 1.97

11.33 10.78 11.06

3.64 4.12 3.43

Paragraph reading (GORT-4)b Ratea,* Accuracy* Fluency*

11.60 11.00 11.70

2.07 1.76 2.67

5.00 5.11 3.22

1.00 2.03 1.72

.12 .37 .34 .02

.97 .93 .43 .99

2.41 1.91 1.93 1.14

1.21 1.30 1.30 2.61

Single word readingc Regular words* Exception words* Non-wordsa,* Dichotic pitcha,c a

Equal variances not assumed. Reported as scaled scores (M = 10, SD = 3). c Reported as z-scores. Significant difference between groups, independent samples t-test, p < .001.

b

*

(no signal present; only background noise) to 1 (equal signal and noise amplitudes) and greater than 1 (signal amplitude higher than noise). At SNR levels less than 1, the tones can only be perceived by combining the input from each ear using binaural cues. At SNR levels greater than 1, there are peaks in the amplitude spectrum that may provide monaural cues to pitch in addition to the binaural location cues. Four tones of 200 ms each (400, 575, 750, 900 Hz), with an ISI of 100 ms of noise only, were presented in either ascending or descending temporal order on a Macintosh G4 computer through Sennheiser HD-265 headphones. An adaptive staircase adjusted the SNR to determine the minimum SNR required for correct pitch change direction. The staircase began with an SNR of 10 so that the tones were initially audible monaurally for all participants. Participants were asked to determine the direction of pitch change of these tones, by indicating ‘up’ or ‘down’ on a response pad. The SNR decreased (increased difficulty) with two successive correct answers and increased with one incorrect answer. The initial step size of 2 SNR was halved after each response reversal. The staircase continued until 50 trials were completed or 10 reversals had occurred. Thresholds were determined by fitting a Weibull function to the data using a maximum-likelihood minimization procedure (Watson, 1979). Threshold was defined as the point of maximum slope on the fitted curve, which occurs at 82% correct in a 2-alternative forced-choice procedure (Strasburger, 2001). 2.3. MRI session 2.3.1. Data acquisition A Philips Gyroscan Intera whole body 3.0 Tesla MRI scanner was used with an 8-channel phased array head coil and parallel imaging (SENSE). A high-resolution T1-weighted 3-D anatomic scan (3DT1TFE/MPRAGE) was acquired first and lasted 6 min 34 s. Images were acquired in 170 axial slices (field-of-view, FOV: 256 mm; 256  256 mm matrix; TR: 8.0 ms; TE: 3.7 ms; flip angle: 8 deg; 1 mm slice thickness; 1 mm3 voxel size). T2-weighted scans were acquired using a single-shot echo-planar imaging (EPI) sequence. A sparse sampling method was used with an effective TR of 7000 ms (actual time of data acquisition was 2020 ms followed by 4980 ms rest period; TE: 30 ms). All images were acquired in 36 interleaved axial slices (FOV: 240 mm; 80  80 mm matrix; flip angle: 90°; 3 mm slice thickness; 1 mm inter-slice

gap). Images were reconstructed with an 80  80 mm matrix, which produced 3 mm3 isotropic voxels. The dichotic pitch task acquired 61 volumes per participant (427 s). Participant responses were obtained using a fiber optic response system (Lumitouch), and each participant wore headphones (Avotec) and earplugs. 2.3.2. Dichotic pitch fMRI task Given that the dichotic pitch stimulus is difficult to hear in an MRI scanner, the stimuli were presented using an event-related sparse sampling design (Belin, Zatorre, Hoge, Evans, & Pike, 1999; Friston, Zarahn, Josephs, Henson, & Dale, 1999; Van den Noort, Specht, Rimol, Ersland, & Hugdahl, 2008). Four conditions consisting of SNR levels of 10 (easy), 1 (difficult), and 0 (random), as well as a silent period (baseline) were presented 15 times in random order, for a total of 60 trials. For each trial, there was a 5 s period during which the stimulus could be played (i.e., scanner is quiet), and was followed by a 2 s TR during which images were collected. The beginning of the sound stimuli was jittered at 1 s, 1.5 s, or 2 s after the silent period began. The stimuli were presented for 1.8 s, which consisted of four tones in ascending or descending order as described above. Images were obtained 3–6 s after the beginning of the stimulus. The three jitter times were presented five times per condition. For SNR levels of 10 and 1, a question mark was presented at fixation, which queued the participant to respond during the trial. The task was to determine the direction of pitch by indicating ‘up’ or ‘down’ on the response pad. During SNR levels of 0 and the silent period, a central fixation cross was presented indicating to the participant that no response was needed during the trial, and they were asked to listen to the tones. Accuracy was recorded for SNR levels of 10, 1, and 0. No response for SNR 0 (accuracy of 0%) was expected for this task. 2.3.3. Pre-processing BrainVoyager QX 2.2.8 (Brain Innovation, Maastricht, The Netherlands) was used for MRI analyses (Goebel, Esposito, & Formisano, 2006). For the T2-weighted scans, slice scan time correction was performed using a cubic spline interpolation. 3-D motion correction using a six parameter rigid-body tri-linear interpolation was performed to correct for small head movements. All images were aligned to the first image in the time course. Translation (mm) and rotations (deg) were estimated for each time course and

107

M. Partanen et al. / Brain & Language 123 (2012) 104–112

images were discarded if these parameters exceeded 3 mm or 3°. Linear trends and non-linear low-frequency drifts of 2 cycles or less were removed with high-pass temporal filtering. For the T1weighted scan, data were interpolated to 1 mm3 isotropic voxels and then transformed into standardized stereotaxic atlas space (Talairach & Tournoux, 1988). The T2-weighted data were aligned to each participant’s standardized anatomical image. FMRI data for one average reader and two dyslexic readers were excluded due to excessive motion or incomplete data. Subsequent analyses were completed on groups with 9 average readers and 7 dyslexic readers.

2.3.4. FMRI analyses A random-effects general linear model was used to determine whole-brain voxel-wise activity for all participants (n = 16). Three predictors from the dichotic pitch task were included in the model: SNR 10 (easy), SNR 1 (difficult), and SNR 0 (random) conditions. The silent condition was utilized as the baseline in this model. Due to the nature of the sparse sampling design (i.e., one volume acquired per trial), the typical hemodynamic response function could not be fit to the data. Rather, an event-related response was defined by the onset of the auditory stimulus in each trial. Activation maps of the t-statistic were created with a Bonferroni-corrected level of p < .05 to adjust for multiple comparisons. Using the contrast of all auditory conditions (SNR 10, 1, 0) versus the silent baseline condition, regions of interest (ROIs) were established. Increased activity within bilateral Heschl’s gyri, planum temporale, planum polare, and right superior temporal sulcus was expected for this task (Garcia et al., 2010; Giaschi et al., 2000; Puschmann et al., 2010).

3.2.2. Regions of interest With all participants grouped together, whole-brain activation for the dichotic pitch task was found in bilateral Heschl’s gyri, right planum temporale, and right superior temporal sulcus. Given the anatomical variation between participants, activation clusters were projected onto a standardized image (Colin27 brain; Collins et al., 1998; see Fig. 1). Four ROIs were established from these areas and were used for further analyses (see Table 2). Within each ROI, a mean standardized beta coefficient was extracted for each SNR condition and participant; these values were used as a measure of cortical activity. The beta coefficients were analyzed with a between-within analysis of variance (ANOVA) to determine any differences between regions, SNR condition, or group: ROI (4 regions)  SNR (10, 1, 0)  Group (average, dyslexic). Results are reported with adjusted degrees of freedom using the Greenhouse–Geisser method when sphericity was violated for within-subjects factors and Brown-Forsythe method (1974) when homogeneity of variance was violated for between-subjects factors. Effect sizes (Cohen’s f) are reported for significant interactions or main effects (Cohen, 1992). The ANOVA results demonstrated a significant 3-way interaction between ROI, SNR, and Group, F (3.73, 52.28) = 5.44, p < .001, f = .62 (large effect); thus, simple effect analyses were completed for each ROI separately. In left Heschl’s gyrus, an interaction between SNR condition and Group was found, F (1.47, 20.56) = 4.97, p < .05, f = .60 (large effect). Follow-up pairwise comparisons demonstrated that in the SNR 0 condition, dyslexic readers had higher cortical activity than average readers (p < .05). Other contrasts between groups were not significant (p > .05). Additionally in the

3. Results 3.1. Behavioral results Age-related normative performance was derived from previous studies for the reading tasks (Coltheart & Leahy, 1996; Edwards & Hogben, 1999) and from previous data collected in our lab for the dichotic pitch task (Edwards, Giaschi, Low, & Edgell, 2005). Zscores based on normative performance were established for each participant and set to a maximum value of ±4. Outliers were found on the orthographic reading task (1 – dyslexic group) and on the dichotic pitch task (4 – dyslexic group). Further inspection of the dichotic pitch thresholds indicated that two readers in the dyslexic group did not perceive tones in the dichotic range (i.e., SNR < 1), whereas all average readers were able to perceive tones in this range. Adolescents with dyslexia demonstrated significantly lower scores than the average readers on all reading tasks, but their poorer performance on the dichotic pitch task did not reach significance (see Table 1).

Fig. 1. Regions of interests (ROIs) established from the dichotic pitch fMRI task for all participants together. Increased activation for all auditory conditions was shown in left Heschl’s gyrus (green), right Heschl’s gyrus (pink), right planum temporale (purple), and right superior temporal sulcus (blue) (p < .05, corrected).

3.2. FMRI results 3.2.1. In-scanner accuracy In-scanner responses were obtained for all SNR conditions. There were high accuracies for SNR conditions of 10 and 1 (between 86% and 94%), and no group differences were found, t (14) = .62, p > .25 for SNR 10, and t (14) = 1.23, p > .10 for SNR 1. Additionally, low accuracies were shown on SNR 0 (between 3% and 5%), and there were no group differences, t (14) = .45, p > .25. However, low accuracies were expected for this condition, as the instructions were to not respond during the trial.

Table 2 Voxel size and coordinates for ROIs. ROI

Number of voxels

Talairach coordinates (x, y, z)

Left Heschl’s gyrus Right Heschl’s gyrus Right planum temporale Right superior temporal sulcus

191 420 615 202

45, 19, 6 47, 18, 7 56, 18, 7 58, 31, 6

108

M. Partanen et al. / Brain & Language 123 (2012) 104–112

Fig. 2. Mean standardized beta coefficients were extracted for each ROI, SNR condition, and group (bars indicate ±1 standard error). Left and right Heschl’s gyrus (HG), right planum temporale (PT), and right superior temporal sulcus (STS) are shown separately. Significant differences between reader groups are shown (p < .05).

average reader group, cortical activity increased as SNR increased; this was illustrated by significant differences between all pairs of SNR conditions (p < .05). However in the dyslexic reader group, this linear trend was not shown. Activation from SNR 0 was similar to SNR 1 and 10 (p > .05), whereas SNR 10 was higher than SNR 1 (p < .05), but not SNR 0 (p > .05). In Fig. 2, it appears that SNR 10 is higher than both SNR 1 and 0; however, the lack of a significant effect is likely to due the variability in the dyslexic reader group. A main effect of SNR condition, F (2, 28) = 29.05, p < .001, f = 1.44 (large effect) and a main effect of group, F (1, 14) = 4.69, p < .05, f = .58 (large effect), was illustrated in right Heschl’s gyrus. Pairwise contrasts showed that all SNR conditions were significantly different from each other (p < .05); as SNR increased, cortical activity in right Heschl’s gyrus also increased. Additionally, average readers had greater activity in this region than dyslexic readers (p < .05). In right planum temporale, there was a main effect of SNR condition, F (2, 28) = 18.61, p < .001, f = 1.15 (large effect). Similarly to the other regions, pairwise contrasts indicated that cortical activity increased with SNR. The SNR 10 condition had significantly greater activity than SNR 0 and SNR 1 conditions (p < .001); however, there was no difference between SNR 0 and 1 (p = .07). In contrast to the left and right Heschl’s gyri, there were no group differences found in this region. Finally, the right superior temporal sulcus showed an interaction between SNR and Group, F (2, 28) = 4.05, p < .05, f = .54 (large effect). Simple effect analyses were completed for each SNR condition separately. For SNR 0 and 10, no group differences were found (p > .05). However for SNR 1 (dichotic condition), the average readers had significantly higher activity than the dyslexic reader group (p < .05). Additional contrasts showed non-linear effects of SNR condition. For the average readers, SNR 1 showed higher activity than SNR 0 (p < .05), whereas SNR 10 was similar to SNR 0 and 1 (p > .05). The dyslexic reader group illustrated a different trend: SNR 10 was higher than SNR 0 and 1 (p < .05), while SNR 0 and 1 showed similar activity (p > .05). 3.2.3. Subgroup analyses We also sought to determine the differences between dyslexic readers who performed poorly on the dichotic pitch psychophysical task (i.e., below 1 z-score) from those who performed within an average range (i.e., above 1 z-score). With these criteria, there were three dyslexic readers with poor dichotic pitch and four dyslexic readers with average dichotic pitch thresholds. Analyses were conducted on SNR conditions only (i.e., beta coefficients were averaged across ROIs) to decrease the number of comparisons and focus on the effect of SNR. We expected that group differences

would be shown on the dichotic condition of SNR 1, as this condition requires the binaural integration of cues and is more difficult than the SNR 0 and SNR 10 conditions. A 3 Group (average reader, dyslexic reader with poor dichotic pitch, dyslexic reader with average dichotic pitch)  3 SNR (10, 1, 0) between-within ANOVA was conducted. Results showed a significant Group  Condition interaction, F (4, 26) = 3.50, p < .05, f = .74 (large effect). Follow-up pairwise comparisons were completed for each SNR condition. In SNR 0, dyslexic readers with poor dichotic pitch had significantly higher cortical activity than average readers and dyslexic readers with average dichotic pitch (both p < .01), while the latter two groups were similar to each other (p > .50). In SNR 1, dyslexic readers with poor dichotic pitch had less cortical activity than the other two groups as expected (both p < .01), and the average readers and dyslexic readers with average dichotic pitch were similar (p > .50). Finally in SNR 10, none of the comparisons were significant (p > .25). Fig. 3 illustrates the subgroup results and shows that group differences were primarily due to the dyslexic readers who had poor dichotic pitch thresholds. 3.2.4. Correlations between cortical activity and reading Finally, correlations between cortical activity and the behavioral reading tasks were conducted. Orthographic and phonological reading ability were correlated to cortical activity in each ROI and SNR condition. Results were reported using a Bonferroni method to correct for multiple comparisons (p < .05). In left Heschl’s gyrus and right planum temporale, there were no significant correlations

Fig. 3. Mean standardized beta coefficients averaged across ROIs (bars indicate ±1 standard error). Activity for each SNR condition is shown for average readers, dyslexic readers with average dichotic pitch (DP) thresholds, and dyslexic readers with poor DP thresholds. Significant differences between groups are shown (p < .05).

M. Partanen et al. / Brain & Language 123 (2012) 104–112

between reading and cortical activity. This result corresponded to the overall ANOVA results, as only one significant group difference was illustrated (i.e., SNR 0 in left Heschl’s gyrus). Conversely, there were significant correlations between SNR 1 (dichotic) and phonological reading ability in right Heschl’s gyrus, r (16) = .64, p < .05, and right superior temporal sulcus, r (16) = .66, p < .05. These correlations were similar to the overall ANOVA results, as both of these regions illustrated group differences. Therefore, these results suggest that phonological reading ability is correlated to cortical activity in right Heschl’s gyrus and right superior temporal sulcus, but only when participants were listening to a dichotic melody (SNR 1). 4. Discussion 4.1. Behavioral performance on reading and dichotic pitch Adolescents with dyslexia obtained lower scores than average readers on measures of phonological and orthographic reading. This result is in accordance to previous findings, which have suggested that there are subgroups of dyslexic readers who have deficits in both phonological and orthographic processing (Castles & Coltheart, 1993; Edwards & Hogben, 1999). Children and adults may have difficulty in establishing a mental lexicon of words (orthographic processing) in addition to associating sounds to letters (phonological processing). Our findings support the hypothesis that both component reading processes are affected in dyslexia. In the dichotic pitch task, however, the dyslexic reader group performed similarly to the average reader group. Several studies have shown that only a proportion of persons with dyslexia have deficits in rapid auditory processing (reviewed in Hämäläinen et al., 2012). Inspection of z-scores in the current study illustrates that four adolescents with dyslexia had poor dichotic pitch (z < 1), whereas the other five adolescents had scores in the average range (z > 1). Previous studies using similar binaural pitch tasks have shown no deficits in dyslexic groups (Chait et al., 2007; Hill, Bailey, Griffiths, & Snowling, 1999; Santurette et al., 2010), however, Santurette et al. (2010) showed that a proportion of their dyslexic group had lower accuracies than average reader adults. Therefore, the current results suggest that there is a subsample of adolescents with dyslexia who have auditory temporal processing problems, which is further supported by our fMRI results, as discussed below. 4.2. Cortical activation for dichotic pitch Increased activation from the fMRI dichotic pitch task was shown in bilateral regions of Heschl’s gyrus, right planum temporale, and right superior temporal sulcus. It was hypothesized that these regions would be activated based on previous neuroimaging studies using similar dichotic pitch tasks (Garcia et al., 2010; Giaschi et al., 2000; Puschmann et al., 2010). Bilateral Heschl’s gyri and right planum temporale showed linear effects of pitch salience, which was illustrated by increased cortical activity with increases in SNR. In contrast, the right superior temporal sulcus did not show this linear effect in the average reader group. These results imply that Heschl’s gyrus and right planum temporale are sensitive to changes in pitch salience, whereas the right superior temporal sulcus may be involved with other components of the dichotic pitch task. Our results are supported by previous findings, which implicate the lateral Heschl’s gyrus in the processing of pitch (reviewed in Bizley & Walker, 2010) and changes in SNR (Ernst, Verhey, & Uppenkamp, 2008). Other researchers, however, suggest that the planum temporale is also involved in pitch perception depending on the type of auditory stimulus (Hall & Plack, 2009). Additionally,

109

there may be a hierarchy of pitch processing; the lateral Heschl’s gyrus is engaged for fixed pitch tasks, whereas surrounding regions such as planum temporale are activated for sequences with changing pitch (Patterson, Uppenkamp, Johnsrude, & Griffiths, 2002). Therefore, our results provide further evidence that pitch perception is localized to Heschl’s gyrus and right planum temporale, whereas the right superior temporal sulcus may be involved with higher-level components of pitch processing. Additionally, the current results showed greater activation in right than left temporal regions. As discussed in the introduction, melody perception is localized to right temporal cortex, whereas rapid auditory processing is localized to left temporal cortex (Warrier et al., 2009; Zatorre & Belin, 2001; Zatorre et al., 2002). Given that our stimulus included components of both temporal processing and melody extraction, we would expect bilateral activation in temporal cortex. However, the task was to indicate the direction of pitch of ascending and descending tones, and increased activity in the right hemisphere may reflect this aspect of pitch or melody extraction. Additionally, the current results showed activation in the right superior temporal sulcus, which has been shown previously for melodic sequences (Zatorre & Belin, 2001). The right superior temporal sulcus may be sensitive to high-level analysis of complex auditory information (Belin, Zatorre, Lafaille, Ahad, & Pike, 2000), which may be needed for pitch direction discrimination. Results from the current study provide evidence that pitch or melody extraction is processed asymmetrically in right temporal regions. In contrast to previous studies that used similar dichotic pitch stimuli (Garcia et al., 2010; Puschmann et al., 2010), our results showed no significant activation in planum polare. These differences in results may be due to developmental differences between adolescents and adults. For example, Koelsch et al. (2005) showed that adults had greater activation than children in left prefrontal, supramarginal gyrus, and temporal regions during a musical chord discrimination task. However, most studies of the auditory system have not directly compared adults and children and thus, developmental differences can only be hypothesized. We suggest that the lack of activity in planum polare may be due to our younger sample, but further research is needed to examine this claim. 4.3. Cortical activation in dyslexia Group differences from the fMRI dichotic pitch task were shown in bilateral Heschl’s gyri and right superior temporal sulcus. Similarly to Stoodley, Hill, Stein, and Bishop (2006), although we did not find group differences on the psychophysical dichotic pitch task, there were differences in cortical activity. These results suggest that neuroimaging measures may be more sensitive than behavioral measures at detecting auditory processing deficits in dyslexia. In left Heschl’s gyrus, adolescents with dyslexia showed greater activation than adolescents with average reading ability, but only in the random noise condition (SNR 0). A linear effect of SNR was shown in the average reader group, whereas activity from SNR 0 was similar to SNR 1 and 10 in the dyslexic group, which implies a quadratic effect of SNR. These results suggest that adolescents with dyslexia may have difficulty discriminating between pitch (SNR 1 and 10) and non-pitch (SNR 0) stimuli, as there were no significant differences between these conditions. Additionally in right Heschl’s gyrus, the average reader group illustrated greater activation than the dyslexic reader group. This difference was shown across three SNR conditions, which implies that the deficit is not related to pitch salience. Additionally, since group differences were shown in bilateral regions of Heschl’s gyrus, this suggests that auditory deficits in dyslexia are not specific to the left hemisphere. Similarly to our results, Steinbrink and colleagues showed that

110

M. Partanen et al. / Brain & Language 123 (2012) 104–112

adults with dyslexia did not activate bilateral insular cortex when syllables were presented (Steinbrink et al., 2009) and dyslexic readers with poor discrimination accuracy had decreased activation in bilateral insula and left inferior frontal gyrus when listening to rapid auditory stimuli (Steinbrink et al., 2012). In contrast, other studies only showed group differences in left superior and middle frontal gyri for rapidly changing acoustic stimuli (Gaab et al., 2007; Temple et al., 2000). However, given that these authors contrasted their two acoustic conditions (slow versus rapid sequences), it is unknown whether there were group differences within auditory cortex. Overall, we suggest that adolescents with dyslexia have deficits in bilateral Heschl’s gyrus during pitch extraction, and this may be one component of the reading process. This was illustrated by the activation differences between average and dyslexic reader groups in bilateral Heschl’s gyrus. An additional group difference was shown in right superior temporal sulcus for the dichotic condition (SNR 1); adolescents with average reading ability showed greater activation than adolescents with dyslexia. As discussed previously, the right superior temporal sulcus may be involved with higher-level pitch extraction (Belin et al., 2000). In the dichotic condition, the participant must combine the rapid temporal cues from both ears in order to perceive pitch. We suggest that the right superior temporal sulcus may be one cortical region that is involved with dichotic pitch perception, and dyslexic readers may have a deficit within this region. Subsequent analyses showed that dyslexic readers with poor dichotic pitch thresholds were significantly different than average readers and dyslexic readers with average dichotic pitch thresholds. As illustrated in Fig. 3, those with poor dichotic pitch had greater activity during a noise condition (SNR 0) and less activity during a dichotic condition (SNR 1), whereas there were no group differences at a higher signal-to-noise ratio condition (SNR 10). These neuroimaging results provide more substantial evidence than behavioral results that there is a sub-sample of dyslexic readers who have auditory processing difficulties. Group differences in the dichotic condition were expected, as binaural cues are needed in order to perceive pitch. This result suggests that a sub-sample of adolescents with dyslexia have difficulty with binaural pitch perception, which has been shown previously in children (Dougherty et al., 1998; Edwards et al., 2004). Higher activity in the noise condition was not expected, however, it can be argued that some dyslexic readers have difficulty with pitch extraction in low signal-to-noise conditions. Furthermore, there were no group differences at the high signal-to-noise condition. These results are in accordance to studies that have shown speech-in-noise perception deficits in children (Boets et al., 2007; Ziegler, Pech-Georgel, George, & Lorenzi, 2009) and adults (Dole, Hoen, & Meunier, 2012) with dyslexia. Specifically, speech perception deficits are prominent when stimuli are presented within noise, but not in silent conditions. We suggest that our dichotic pitch task includes aspects of noise exclusion in addition to rapid auditory processing and that some dyslexic readers have difficulty with these components of the task. Furthermore, we provide novel neurobiological evidence that there is a sub-sample of dyslexic readers who have auditory processing deficits. 4.4. Relationship between reading and auditory processing The relationship between dichotic pitch processing and reading was further examined with correlation analysis. We found that cortical activation by the dichotic condition was correlated to phonological reading ability in right Heschl’s gyrus and right superior temporal sulcus. The specific relationship between dichotic stimuli and phonological reading, and not with orthographic reading, suggests there are similar skills involved with dichotic pitch and phonological processing. At the most basic level, phonological awareness is

the recognition that words are composed of individual phonemes; persons with dyslexia may have difficulty with this task because of the rapid formant transitions that occur in speech (e.g., Tallal, 1980). In the dichotic pitch task, the participant must fuse signals that are presented binaurally and differ by only 0.6 ms. Therefore it is plausible that this relationship is due to the rapid processing component of the tasks. However, the significant correlations in this study were shown in right temporal regions, which would be in contrast to studies suggesting that left temporal cortex is responsible for rapid processing of acoustic stimuli (Warrier et al., 2009; Zatorre & Belin, 2001; Zatorre et al., 2002). Another relationship between phonological processing and dichotic pitch is the ability to extract a signal from a distracting noise background. When a child is learning to read, they must associate letters to sounds or make a grapheme–phoneme correspondence (Ehri, 1998). Similar phonemes such as /p/ and /b/ need to be differentiated as different sounds and thus, the child detects the signal /p/ from its distractors /b/. In dichotic pitch, the participant needs to detect signal tones from background noise and the task becomes more difficult as the SNR decreases. As discussed above, persons with dyslexia may have noise exclusion deficits (Boets et al., 2007; Dole et al., 2012; Ziegler et al., 2009). We provide further evidence that adolescents with dyslexia have difficulty extracting the signal (tones) from distracting backgrounds, however, our results are unique as these deficits were shown in brain activity. 4.5. Considerations It should be noted that the fMRI dichotic pitch task is likely more difficult than the task presented in the lab. In the scanner, participants wore earplugs and headphones, and there is some scanner noise even though stimuli were presented with a sparse sampling design. However, all participants had high accuracies for the fMRI task and thus, the task was not too difficult in the scanner. 5. Conclusions Cortical activation from the dichotic pitch fMRI task was shown in bilateral Heschl’s gyri, right planum temporale, and right superior temporal sulcus. Effects of pitch salience were illustrated in bilateral Heschl’s gyri and right planum temporale; these regions were sensitive to changes in SNR, whereas right superior temporal sulcus may be involved with higher-level processing of pitch. Additionally, there was greater activation in right than left temporal regions, which may be explained by the melodic component of the dichotic pitch task. Group differences were shown in bilateral Heschl’s gyri and right superior temporal sulcus. Adolescents with dyslexia, in comparison to adolescents with average reading ability, illustrated greater activity in left Heschl’s gyrus for random noise, less activity in right Heschl’s gyrus for all auditory conditions, and less activity in right superior temporal sulcus for a dichotic melody. When all regions of interest were analyzed together, it was revealed that dyslexic readers with poor dichotic pitch thresholds had greater activity for random noise and less activity for a dichotic melody, in comparison to average readers and dyslexic readers with average dichotic pitch thresholds. These results provide further evidence that there is a subgroup of dyslexic readers who have auditory processing difficulties. Finally, behavioral performance on phonological reading was correlated to cortical activity from dichotic conditions in right Heschl’s gyrus and right superior temporal sulcus. Although we cannot exclude the rapid auditory components of the dichotic pitch task, there is greater evidence that our results are

M. Partanen et al. / Brain & Language 123 (2012) 104–112

due to noise exclusion deficits. This claim is supported by: (1) group differences and correlations between reading and dichotic pitch activity were shown predominantly in the right hemisphere, which is implicated for spectral processing and not rapid auditory processing, and (2) subgroup differences were shown on low signal-to-noise conditions, which would suggest that some dyslexic readers have difficulty excluding noise from their environment. Therefore, children who have difficulty learning to read may have a general problem extracting signals from background noise, which is expressed as deficits in dichotic pitch processing and reading. Acknowledgments The authors would like to thank the UBC MRI Research Centre and Philips Healthcare, the staff and students who assisted with data collection including Emily Harrison, Grace Truong, and Rawn Stokoe, as well as Daniel Kim for his assistance with neuroimaging techniques. References Belin, P., Zatorre, R. J., Hoge, R., Evans, A. C., & Pike, B. (1999). Event-related fMRI of the auditory cortex. NeuroImage, 10, 417–429. http://dx.doi.org/10.1006/ nimg.1999.0480. Belin, P., Zatorre, R. J., Lafaille, P., Ahad, P., & Pike, B. (2000). Voice-selective areas in human auditory cortex. Nature, 403, 309–312. http://dx.doi.org/10.1038/ 35002078. Bizley, J. K., & Walker, K. M. M. (2010). Sensitivity and selectivity of neurons in auditory cortex to the pitch, timbre, and location of sounds. Neuroscientist, 16, 453–469. http://dx.doi.org/10.1177/1073858410371009. Boets, B., Wouters, J., van Wieringen, A., & Ghesquière, P. (2007). Auditory processing, speech perception and phonological ability in pre-school children at high-risk for dyslexia: A longitudinal study of the auditory temporal processing theory. Neuropsychologia, 45, 1608–1620. http://dx.doi.org/ 10.1016/j.neuropsychologia.2007.01.009. Brown, M. B., & Forsythe, A. B. (1974). 372: The ANOVA and multiple comparisons for data with heterogeneous variances. Biometrics, 30, 719–724. . Castles, A., & Coltheart, M. (1993). Varieties of developmental dyslexia. Cognition, 47, 149–180. http://dx.doi.org/10.1016/0010-0277(93)90003-E. Caylak, E. (2011). The auditory temporal processing deficit theory in children with developmental dyslexia. Journal of Pediatric Neurology, 9, 151–168. http:// dx.doi.org/10.3233/JPN-2011-0455. Chait, M., Eden, G., Poeppel, D., Simon, J. Z., Hill, D. F., & Flowers, D. L. (2007). Delayed detection of tonal targets in background noise in dyslexia. Brain and Language, 102, 80–90. http://dx.doi.org/10.1016/j.bandl.2006.07.001. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. http:// dx.doi.org/10.1037/0033-2909.112.1.155. Collins, D. L., Zijdenbos, A. P., Kollokian, V., Sled, J. G., Kabani, N. J., Holmes, C. J., et al. (1998). Design and construction of a realistic digital brain phantom. IEEE Transactions on Medical Imaging, 17, 463–468. http://dx.doi.org/10.1109/ 42.712135. Coltheart, M., & Leahy, J. (1996). Assessment of lexical and nonlexical reading abilities in children: Some normative data. Australian Journal of Psychology, 48, 136–140. http://dx.doi.org/10.1080/00049539608259520. Dole, M., Hoen, M., & Meunier, F. (2012). Speech-in-noise perception deficit in adults with dyslexia: Effects of background type and listening configuration. Neuropsychologia, 50, 1543–1552. http://dx.doi.org/10.1016/j. neuropsychologia.2012.03.007. Dougherty, R. F., Cynader, M. S., Bjornson, B. H., Edgell, D., & Giaschi, D. E. (1998). Dichotic pitch: A new stimulus distinguishes normal and dyslexic auditory function. Neuroreport, 9, 3001–3005. . Edwards, V. T., Giaschi, D. E., Dougherty, R. F., Edgell, D., Bjornson, B. H., Lyons, C., et al. (2004). Psychophysical indexes of temporal processing abnormalities in children with developmental dyslexia. Developmental Neuropsychology, 25, 321–354. http://dx.doi.org/10.1207/s15326942dn2503_5. Edwards, V. T., Giaschi, D. E., Low, P., & Edgell, D. (2005). Sensory and nonsensory influences on children’s performance of dichotic pitch perception tasks. Journal of the Acoustical Society of America, 117, 3157–3164. http://dx.doi.org/10.1121/ .1.1861599. Edwards, V. T., & Hogben, J. H. (1999). New norms for comparing children’s lexical and nonlexical reading: A further look at subtyping dyslexia. Australian Journal of Psychology, 51, 37–49. http://dx.doi.org/10.1080/00049539908255333. Ehri, L. C. (1998). Grapheme–phoneme knowledge is essential for learning to read words in English. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 3–40). Mahwah, NJ: Erlbaum. Ernst, S. M. A., Verhey, J. L., & Uppenkamp, S. (2008). Spatial dissociation of changes of level and signal-to-noise ratio in auditory cortex for tones in noise. NeuroImage, 43, 321–328. http://dx.doi.org/10.1016/j.neuroimage.2008.07.046.

111

Farmer, M. E., & Klein, R. M. (1995). The evidence for a temporal processing deficit linked to dyslexia: A review. Psychonomic Bulletin & Review, 2, 460–493. http:// dx.doi.org/10.3758/BF03210983. Fisher, S. E., & DeFries, J. C. (2002). Developmental dyslexia: Genetic dissection of a complex cognitive trait. Nature Reviews Neuroscience, 3, 767–780. http:// dx.doi.org/10.1038/nrn936. Friston, K. J., Zarahn, E., Josephs, O., Henson, R. N. A., & Dale, A. M. (1999). Stochastic designs in event-related fMRI. NeuroImage, 10, 607–619. http://dx.doi.org/ 10.1006/nimg.1999.0498. Gaab, N., Gabrieli, J. D. E., Deutsch, G. K., Tallal, P., & Temple, E. (2007). Neural correlates of rapid auditory processing are disrupted in children with developmental dyslexia and ameliorated with training: An fMRI study. Restorative Neurology and Neuroscience, 25, 295–310. . Garcia, D., Hall, D. A., & Plack, C. J. (2010). The effect of stimulus context on pitch representations in the human auditory cortex. NeuroImage, 51, 808–816. http:// dx.doi.org/10.1016/j.neuroimage.2010.02.079. Giaschi, D., Bjornson, B., Dougherty, R. F., & Au Young, S. (2000). Cortical mechanisms of central pitch processing revealed by dichotic pitch stimuli. NeuroImage, 11, S703. Goebel, R., Esposito, F., & Formisano, E. (2006). Analysis of functional image analysis contest (FIAC) data with BrainVoyager QX: From single-subject to cortically aligned group general linear model analysis and self-organizing group independent component analysis. Human Brain Mapping, 27, 392–401. http:// dx.doi.org/10.1002/hbm.20249. Habib, M. (2000). The neurological basis of developmental dyslexia: An overview and working hypothesis. Brain, 123, 2373–2399. http://dx.doi.org/10.1093/ brain/123.12.2373. Hall, D. A., & Plack, C. J. (2009). Pitch processing sites in the human auditory brain. Cerebral Cortex, 19, 576–585. http://dx.doi.org/10.1093/cercor/bhn108. Hämäläinen, J. A., Salminen, H. K., & Leppänen, P. H. T. (2012). Basic auditory processing deficits in dyslexia: Systematic review of the behavioral and eventrelated potential/field evidence. Journal of Learning Disabilities. http:// dx.doi.org/10.1177/0022219411436213. Hill, N. I., Bailey, P. J., Griffiths, Y. M., & Snowling, M. J. (1999). Frequency acuity and binaural masking release in dyslexic listeners. Journal of the Acoustical Society of America, 106, L53–L58. Katusic, S. K., Colligan, R. C., Barbaresi, W. J., Schaid, D. J., & Jacobsen, S. J. (2001). Incidence of reading disability in a population-based birth cohort, 1976–1982, Rochester, Minn. Mayo Clinic Proceedings, 76, 1081–1092. http://dx.doi.org/ 10.4065/76.11.1081. Koelsch, S., Fritz, T., Schulze, K., Alsop, D., & Schlaug, G. (2005). Adults and children processing music: An fMRI study. NeuroImage, 25, 1068–1076. http://dx.doi.org/ 10.1016/j.neuroimage.2004.12.050. Landerl, K., & Willburger, E. (2010). Temporal processing, attention, and learning disorders. Learning and Individual Differences, 20, 393–401. http://dx.doi.org/ 10.1016/j.lindif.2010.03.008. Lyon, G. R., Shaywitz, S. E., & Shaywitz, B. A. (2003). A definition of dyslexia. Annals of Dyslexia, 53, 1–14. http://dx.doi.org/10.1007/s11881-003-0001-9. McAnally, K. I., & Stein, J. F. (1996). Auditory temporal coding in dyslexia. Proceedings of the Royal Society of London, 263, 961–965. . Patterson, R. D., Uppenkamp, S., Johnsrude, I. S., & Griffiths, T. D. (2002). The processing of temporal pitch and melody information in auditory cortex. Neuron, 36, 767–776. http://dx.doi.org/10.1016/S0896-6273(02)01060-7. Puschmann, S., Uppenkamp, S., Kollmeier, B., & Thiel, C. M. (2010). Dichotic pitch activates pitch processing centre in Heschl’s gyrus. NeuroImage, 49, 1641–1649. http://dx.doi.org/10.1016/j.neuroimage.2009.09.045. Ramus, F. (2003). Developmental dyslexia: Specific phonological deficit or general sensorimotor dysfunction? Current Opinion in Neurobiology, 13, 212–218. http:// dx.doi.org/10.1016/S0959-4388(03)00035-7. Rosen, S., & Manganari, E. (2001). Is there a relationship between speech and nonspeech auditory processing in children with dyslexia? Journal of Speech, Language, and Hearing Research, 44, 720–736. doi: http://dx.doi.org/10.1044/ 1092-4388(2001/057). Santurette, S., Poelmans, H., Luts, H., Ghesquiére, P., Wouters, J., & Dau, T. (2010). Detection and identification of monaural and binaural pitch contours in dyslexic listeners. Journal of the Association for Research in Otolaryngology, 11, 512–524. http://dx.doi.org/10.1007/s10162-010-0216-5. Snowling, M. J. (1981). Phonemic deficits in developmental dyslexia. Psychological Research, 43, 219–234. http://dx.doi.org/10.1007/BF00309831. Stanovich, K. E., & Siegel, L. S. (1994). Phenotypic performance profile of children with reading disabilities: A regression-based test of the phonological-core variable-difference model. Journal of Educational Psychology, 86, 24–53. http:// dx.doi.org/10.1037/0022-0663.86.1.24. Steinbrink, C., Ackermann, H., Lachmann, T., & Riecker, A. (2009). Contribution of the anterior insula to temporal auditory processing deficits in developmental dyslexia. Human Brain Mapping, 30, 2401–2411. http://dx.doi.org/10.1002/ hbm.20674. Steinbrink, C., Groth, K., Lachmann, T., & Riecker, A. (2012). Neural correlates of temporal auditory processing in developmental dyslexia during German vowel length discrimination: An fMRI study. Brain and Language, 121, 1–11. http:// dx.doi.org/10.1016/j.bandl.2011.12.003. Stoodley, C. J., Hill, P. R., Stein, J. F., & Bishop, D. V. M. (2006). Auditory event-related potentials differ in dyslexics even when auditory psychophysical performance is normal. Brain Research, 1121, 190–199. http://dx.doi.org/10.1016/ j.brainres.2006.08.095.

112

M. Partanen et al. / Brain & Language 123 (2012) 104–112

Strasburger, H. (2001). Converting between measures of slope of the psychometric function. Perception and Psychophysics, 63, 1348–1355. http://dx.doi.org/ 10.3758/BF03194547. Sutherland, M. E., Zatorre, R. J., Watkins, K. E., Hervé, P.-V., Leonard, G., Pike, B. G., et al. (2012). Anatomical correlates of dynamic auditory processing: Relationship to literacy during early adolescence. NeuroImage, 60, 1287–1295. http://dx.doi.org/10.1016/j.neuroimage.2012.01.051. Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain. New York, NY: Thieme. Talcott, J. B., Gram, A., van Ingelghem, M., Witton, C., Stein, J. F., & Toennessen, F. E. (2003). Impaired sensitivity to dynamic stimuli in poor readers of a regular orthography. Brain and Language, 87, 259–266. http://dx.doi.org/10.1016/ S0093-934X(03)00105-6. Tallal, P. (1980). Auditory temporal perception, phonics, and reading disabilities in children. Brain and Language, 9, 182–198. http://dx.doi.org/10.1016/0093934X(80)90139-X. Temple, E., Poldrack, R. A., Protopapas, A., Nagarajan, S., Salz, T., Tallal, P., et al. (2000). Disruption of the neural response to rapid acoustic stimuli in dyslexia: Evidence from functional MRI. Proceedings of the National Academy of Sciences of the United States of America, 97, 13907–13912. http://dx.doi.org/10.1073/ pnas.240461697. Van den Noort, M., Specht, K., Rimol, L. M., Ersland, L., & Hugdahl, K. (2008). A new verbal reports fMRI dichotic listening paradigm for studies of hemispheric

asymmetry. NeuroImage, 40, 902–911. http://dx.doi.org/10.1016/ j.neuroimage.2007.11.051. Warrier, C., Wong, P., Penhune, V., Zatorre, R., Parrish, T., Abrams, D., et al. (2009). Relating structure to function: Heschl’s gyrus and acoustic processing. The Journal of Neuroscience, 29, 61–69. http://dx.doi.org/10.1523/jneurosci.348908.2009. Watson, A. B. (1979). Probability summation over time. Vision Research, 19, 515–522. http://dx.doi.org/10.1016/0042-6989(79)90136-6. Wechsler, D. (2003). Wechsler intelligence scale for children (4th ed.). San Antonio, TX: The Psychological Corporation. Wiederholt, J. L., & Bryant, B. R. (2001). Gray oral reading test (4th ed.). Austin, TX: Pro-Ed. Zatorre, R. J., & Belin, P. (2001). Spectral and temporal processing in human auditory cortex. Cerebral Cortex, 11, 946–953. http://dx.doi.org/10.1093/cercor/ 11.10.946. Zatorre, R. J., Belin, P., & Penhune, V. B. (2002). Structure and function of auditory cortex: Music and speech. Trends in Cognitive Sciences, 6, 37–46. http:// dx.doi.org/10.1016/S1364-6613(00)01816-7. Ziegler, J. C., Pech-Georgel, C., George, F., & Lorenzi, C. (2009). Speech-perception-innoise deficits in dyslexia. Developmental Science, 12, 732–745. http://dx.doi.org/ 10.1111/j.1467-7687.2009.00817.x.