Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study

Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Q1 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study Marjorie Dole a,b,n, Fanny Meunier c,d, Michel Hoen d,e,f a

Université Grenoble Alpes, LPNC, Grenoble, France CNRS, LPNC, Grenoble, France c L2C2, CNRS UMR 5304, Institut des Sciences Cognitives, Bron, France d Université de Lyon, F-69000 Lyon, France e INSERM U1028, Lyon Neuroscience Research Center, Brain Dynamics and Cognition Team, Lyon F-69000, France f CNRS UMR 5292, Lyon Neuroscience Research Center, Brain Dynamics and Cognition Team, Lyon F-69000, France b

art ic l e i nf o

a b s t r a c t

Article history: Received 24 September 2013 Received in revised form 23 May 2014 Accepted 24 May 2014

Dyslexia is a language-based neurodevelopmental disorder. It is characterized as a persistent deficit in reading and spelling. These difficulties have been shown to result from an underlying impairment of the phonological component of language, possibly also affecting speech perception. Although there is little evidence for such a deficit under optimal, quiet listening conditions, speech perception difficulties in adults with dyslexia are often reported under more challenging conditions, such as when speech is masked by noise. Previous studies have shown that these difficulties are more pronounced when the background noise is speech and when little spatial information is available to facilitate differentiation between target and background sound sources. In this study, we investigated the neuroimaging correlates of speech-in-speech perception in typical readers and participants with dyslexia, focusing on the effects of different listening configurations. Fourteen adults with dyslexia and 14 matched typical readers performed a subjective intelligibility rating test with single words presented against concurrent speech during functional magnetic resonance imaging (fMRI) scanning. Target words were always presented with a four-talker background in one of three listening configurations: Dichotic, Binaural or Monaural. The results showed that in the Monaural configuration, in which no spatial information was available and energetic masking was maximal, intelligibility was severely decreased in all participants, and this effect was particularly strong in participants with dyslexia. Functional imaging revealed that in this configuration, participants partially compensate for their poorer listening abilities by recruiting several areas in the cerebral networks engaged in speech perception. In the Binaural configuration, participants with dyslexia achieved the same performance level as typical readers, suggesting that they were able to use spatial information when available. This result was, however, associated with increased activation in the right superior temporal gyrus, suggesting the need to reallocate neural resources to overcome speech-in-speech difficulties. Taken together, these results provide further understanding of the speech-in-speech perception deficit observed in dyslexia. & 2014 Published by Elsevier Ltd.

Keywords: Dyslexia Cocktail party Speech-in-noise fMRI Informational masking

1. Introduction The current study evaluated the impacts of different ecologically inspired listening situations on the cortical networks implicated in speech-in-noise perception in adults with dyslexia and typical-reading controls.

n Corresponding author. Permanent address: Laboratoire de Psychologie et NeuroCognition, Université Pierre Mendès France, BP 47, 38040 Grenoble Cedex 9, France. Tel.: þ 33 4 76 82 78 31. E-mail addresses: [email protected] (M. Dole), [email protected] (F. Meunier), [email protected] (M. Hoen).

Dyslexia is a neurodevelopmental disorder that causes persistent difficulties in the acquisition of written language abilities such as reading, writing and spelling (Lyon, Shaywitz, & Shaywitz, 2003). Difficulties associated with dyslexia persist into adulthood, causing measurable language processing deficits in adults (Frauenheim, 1978; Ramus et al., 2003). These written-language deficits are thought to arise from difficulties in the development of spoken language, particularly in the processing of phonological information (Ramus et al., 2003; Snowling, 2000). In support of this hypothesis, longitudinal studies have revealed that phonological abilities predict reading acquisition in children at familial risk for dyslexia (Boets et al., 2010; Puolokanaho et al., 2007; Torppa, Lyytinen, Erskine, Eklund, & Lyytinen, 2010). The phonological

http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016 0028-3932/& 2014 Published by Elsevier Ltd.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

impairments frequently observed in dyslexia include difficulties with tasks evaluating phonological awareness (Bruck & Treiman, 1990; Sprenger-Charolles, Colé, Lacert, & Serniclaes, 2000) or short-term verbal memory (Nelson & Warrington, 1980; Trecy, Steve, & Martine, 2013). Recent observations further suggest that access to phonological representations, not phonological representations themselves, are hindered in dyslexia (Boets et al., 2013). With speech processing at the origin of the formation of phonetic/ phonological representations, it was recently suggested that these issues could be related to an underlying speech perception deficit, particularly under noisy conditions (Ziegler, Pech-Georgel, George, & Lorenzi, 2009). Under quiet listening conditions, there is little evidence for speech perception difficulties in dyslexia; such difficulties are, however, more consistently reported when listening situations are more challenging, such as in the presence of background noise (Boets, Ghesquière, van Wieringen, & Wouters, 2007; Boets, Wouters, van Wieringen, & Ghesquière, 2007; Boets et al., 2011). In an early experiment, Brady, Shankweiler, and Mann (1983) reported better performance by typical readers compared with poor readers on a speech perception task when speech was masked by noise, but such a difference was not found under quiet conditions. More recently, Hazan, Messaoud-Galusi, and Rosen (2013) refined this view by showing that children with dyslexia exhibited difficulties in speech-in-noise perception but only under more difficult conditions, such as when the speaker's intonation was variable or when the listeners were under greater memory and cognitive load, as in discrimination tasks. The existence of a speech-in-noise perception deficit in dyslexia is interesting because the ability to understand one speech stream while other talkers are speaking in the background i.e., the cocktail party phenomenon (as first described by Cherry, 1953) relies on multiple cognitive aspects that may be differentially affected in dyslexia. A first relevant dimension of the cocktail party phenomenon is the fact that two types of interference between target speech and background noise are commonly described: energetic and informational masking (Brungart, 2001; Brungart, Simpson, Ericson, & Scott, 2001). Energetic masking refers to the overlap of a target and a competing sound in time and frequency at the cochlear level, leading to the encoding of both sounds in ascending auditory pathways. Informational masking refers to listening situations in which the target and masker are clearly audible but the listener is unable to segregate them because of the similar nature of the information contained in the competing sounds, which interfere with one another. This notion of informational masking is of particular interest for speech-in-speech perception, when different levels of linguistic information (e.g., phonological or lexical information) can cause interference, as evidenced by multilinguistic babbles (Boulenger, Hoen, Ferragne, Pellegrino, & Meunier, 2010; Brouwer, Van Engen, Calandruccio, & Bradlow, 2012) or speech-derived noises (Hoen et al., 2007). In dyslexia, the cocktail party situation can thus be used to clarify what levels of interference hinder intelligibility. In this context, for example, Ziegler et al. (2009) showed that both the addition of external noise and the degradation of target speech signals (referred to as ‘internal noise’) can hinder intelligibility in children with dyslexia. This result suggests that the problem faced by participants with dyslexia is energetic, i.e., low level, in essence and not informational; at a minimum, it is not the processing of speech-in-noise per se that is impaired, but rather the processing of noisy speech. A second crucial aspect of the cocktail party phenomenon is the ability to perform auditory scene analysis. The perception of concurrent speech streams as different auditory objects improves intelligibility and is simpler if concurrent sounds originate from distinct positions in space (Bronkhorst, 2000; Divenyi, 2004). Characterizing spatial auditory processing during speech-inspeech perception in dyslexia may actually prove helpful for

understanding which processes related to speech recognition are impaired. Spatial abilities are mostly based on the processing of the differences in the sounds reaching the two ears, such as interaural level (ILD) and time differences (ITD) (Bronkhorst & Plomp, 1988; Drullman & Bronkhorst, 2000). Perceptual facilitation due to target-background separation by relying on these cues is commonly observed and is referred to as spatial unmasking. In participants with dyslexia, Hill, Bailey, Griffiths, and Snowling (1999) measured the ability to detect a pure tone presented in concurrent Gaussian noise and did not observe any significant difference between participants with dyslexia and matched controls, suggesting typical spatial segregation and binaural integration abilities in dyslexia, at least for the processing of non-speech targets. However, some studies using dichotic pitch stimuli suggest the contrary by showing impaired binaural facilitation in participants with dyslexia (Dougherty, Cynader, Bjornson, Edgell, & Giaschi, 1998; Edwards et al., 2004; but see Amitay, Ahissar, & Nelken, 2002; Chait et al., 2007). These discrepancies can be explained by former work showing that spatial unmasking differed depending on the type of masking (energetic or informational), with spatial unmasking being more effective for informational compared with energetic masking (Arbogast, Mason, & Kidd, 2002; Freyman, Balakrishnan, & Helfer, 2001; Freyman, Helfer, McCall, & Clifton, 1999; Hawley, Litovsky, & Culling, 2004). Thus, the question regarding the origins of the speech-in-noise perception deficits in dyslexia is still far from clear, and further studies should be dedicated to disentangling spatial effects from other masking effects. In this context, we conducted a behavioral evaluation of spatial processing during a speech-in-noise perception task in adults with dyslexia by comparing three listening configurations: Monaural, Dichotic and Spatialized (Dole, Hoen, & Meunier, 2012). The major observation of our behavioral study was the presence of a significant speech-in-noise perception deficit in participants with dyslexia when the background was speech, but only in the Monaural configuration, when the target and background were not spatially separable and energetic masking was maximal. Conversely, when spatial information was available, participants with dyslexia performed as well as typical readers did, suggesting typical spatial processing abilities and potential compensation for detrimental energetic masking with spatial unmasking. A central issue that remains largely open is the identification of neural correlates associated with speech-in-speech perception in different listening configurations to determine whether participants with dyslexia show particular patterns of cerebral activity. Previous neuroimaging studies investigating neural correlates of the cocktail party situation, which have to date only involved typical reading participants, showed that speech perception in noise engages a neural network including the bilateral superior temporal gyrus (STG), left inferior frontal gyrus (IFG), left thalamus, dorsolateral prefrontal cortex and right posterior parietal cortex (Salvi et al., 2002; Wong et al., 2009; Wong, Uppunda, Parrish, & Dhar, 2008; Zekveld, Heslenfeld, Festen, & Schoonhoven, 2006; see Scott & McGettigan, 2013, for a recent review). In an elegant series of studies using speech stimuli presented with different background noises, Scott and colleagues refined this view by determining speech-processing areas engaged in the informational and energetic components of masking. The energetic components appear to engage a network including phonological and motor areas, such as the IFG, supplementary motor area, premotor regions and prefrontal regions (Davis & Johnsrude, 2003; Obleser & Kotz, 2010; Obleser, Wise, Alex Dresner, & Scott, 2007; Scott, Rosen, Wickham, & Wise, 2004; Zekveld et al., 2006). For informational masking, their studies showed that the bilateral STG was the main site of informational interference (Scott et al., 2004). This first study was refined in a second experiment, which revealed

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

that informational masking could create two main asymmetrically distributed effects, one related to linguistic processes engaging the left superior temporal sulcus (STS)/STG and the other involving the right STG and reflecting processes associated with the processing of competing streams, which occurs when several auditory objects are present at the same time (Scott, Rosen, Beaman, Davis, & Wise, 2009). This idea is consistent with previous observations suggesting the involvement of the right STS/STG in voice-signature-driven activities or auditory object representation (Zatorre & Belin, 2001; Zatorre, Bouffard, & Belin, 2004). However, spatial unmasking resulting from the processing of ILDs or ITDs at a lower level seems to engage bilateral primary auditory areas (Schadwinkel & Gutschalk, 2010). To date, only a few studies have been specifically designed to investigate neural correlates of the speech-in-noise processing deficit in dyslexia. Electrophysiological studies suggest atypical encoding of speech stimuli when presented with background noise among participants with dyslexia, both at the brainstem (Cunningham, Nicol, Zecker, Bradlow, & Kraus, 2001) and cortical (Warrier, Johnson, Hayes, Nicol, & Kraus, 2004; Wible, Nicol, & Kraus, 2002) levels. Moreover, neuroimaging studies conducted in participants with dyslexia performing various tasks showed functional and morphological anomalies in brain regions specifically involved in speech processing (see Habib, 2000; Heim & Keil, 2004 for reviews). However, the way in which these regions are differentially activated in dyslexia in a cocktail party situation involving background speech remains an open question. Furthermore, to date, no study has specifically investigated how neural networks involved in speech-in-speech perception are modulated by different listening configurations involving different types of auditory space analyses. In the current study, we used functional magnetic resonance imaging (fMRI) to identify the neural correlates of speech-in-speech processing under different listening configurations involving varying amounts of spatial separation between target speech and background. This information will allow us to more precisely determine the influence of spatial information on the energetic/informational masking balance and identify the neural networks involved in spatial release from masking in typical readers and participants with dyslexia. Neural activation in the speech-related regions should differ for the Monaural versus the two other configurations due to the reduced intelligibility and increased amount of energetic masking produced by the absence of spatial information. Furthermore, given the deficit observed in the Monaural configuration in participants with dyslexia, we should observe reduced activation in these areas in this configuration in participants with dyslexia compared with typical readers.

2. Participants and methods

3

67

Table 1 Summary of the characteristics of the two participants groups.

Q4 68

Variable

Adults with dyslexia Typical readers p

Age (years) Gender (male/female) Handedness Raven's (scores) Reading age (months)

23.57 9/5 84.29 48.69 136.77

Word decoding Phonological Score (/20) Time (s) Nonwords Score (/20) Time Orthographic Score (/20) Time (s) Spelling Words Phonological Nonwords Orthographic Sentences Orthography Grammar Phonological awareness Phoneme deletion Score (/10) Time (s) Acronyms Score (/10) Time (s) Visual assessment Letter sequence Score (/20) Time (s) Bell's test Score (/35)

(6.26) (10.89) (4.04) (17.23)

25.79 8/6 90.0 51.08 168.92

(6.34) (11.09) (5.15) (3.68)

19.57 (0.65) 14.29 (4.76)

20 (0) 9.58 (1.93)

17.57 (2.31) 25.07 (6.51)

19.23 (1.23) 12.92 (2.53)

19.36 (1.01) 13.14 (4.11)

19.85 (0.555) 10.08 (2.10)

0.307 0.107 0.202 o 0.001n

0.0248n 0.0039n 0.03n o 0.001n 0.135 0.0236n

8.64 (1.39) 9.14 (1.03) 9.64 (0.63)

9.61 (0.51) 9.85 (0.38) 9.85 (0.38)

0.027n 0.037n 0.279

9.43 (1.02) 8.07 (2.2)

9.85 (0.55) 9.77 (0.6)

0.172 0.008n

8.08 (2.10) 36.29 (9.13)

10 (0) 23.69 (3.77)

0.001n o 0.001n

8.07 (2.46) 70 (35.4)

9.46 (0.77) 43.15 (8.99)

0.104 0.003n

19.75 35.5 33.77 14.23

(0.45) (7.10) (1.17) (2.13)

0.273 0.002n 0.602 0.107

(1.12) (0.84)

6.61 (0.65) 5.69 (1.1)

0.045n 0.019n

(0.27)

16 (0)

0.365

(1.33)

19.75 (0.45)

0.087

19.46 45.69 34 15.93

Rapid automatized naming (s) Working memory Digit span—ascending 5.78 Digit span—descending 4.64 Word repetition Score (/16) 15.93 Nonword repetition Score (/20) 18.93

(0.78) (7.80) (1.109) (3.27)

Asterisks indicate significant differences at p o 0.05. Standard deviation is indicated in brackets.

informed consent and were paid for their participation. The current protocol was approved by a local ethics committee (CPP Sud-Est IV; ID RCB: 2008-A00708-47).

2.1. Participants

2.2. Psychometric evaluation and diagnostic

Fourteen adults with dyslexia (5 females, mean age: 23.57 years, standard deviation [SD]: 6.26) and 14 typical readers (6 females, mean age: 26.23 years, SD: 6.36), selected to match participants with dyslexia with respect to age, non-verbal IQ and handedness (independent t-test, all p 40.05; see Table 1), participated in this study. The entire sample of participants with dyslexia and all but two participants from the typical readers group were different from the individuals in our previous behavioral study (Dole et al., 2012). All participants were recruited from local universities and were engaged in higher education programs. All were right-handed (scores Z 70 on the Edinburgh Handedness Inventory proposed by Oldfield, 1971), and all had audiometric pure-tone thresholds r 25 dB Hearing Level (HL) at all frequencies in the 250–8000 Hz range. Participants reported no history of psychiatric or neurological disorders. Participants were included in the dyslexia group if they reported a childhood history of reading/spelling disorders and had a documented history of prior diagnosis of dyslexia from a psychologist or education specialist. In addition, only participants still exhibiting reading and/or phonological deficits were included in this group. All participants provided written

Nonverbal IQ was assessed using Raven's Standard Progressive Matrices (Raven, 1938). All participants obtained a score above the 40th percentile, and the scores did not significantly differ between the two groups (p4 0.05). Reading age was assessed using the French l’Alouette Reading Test (Lefavrais, 1967), and the neuropsychological battery ODEDYS (Jacquier-Roux, Valdois, Zorman, Lequette, & Pouget, 2005) was administered to all participants. This battery includes word decoding and spelling tests, phonological awareness tests (phoneme deletion and acronyms), visual attention tests (Letter sequence and Bell's test), Rapid Automatized Naming and working memory tests (digit span and word/nonword repetition). Table 1 summarizes the characteristics measured in each group. All but two participants with dyslexia included in this study performed below 2 SDs of the typical readers’ mean performance in the l’Alouette Reading Test; the remaining two participants performed below 2 SDs of the typical readers’ mean performance on the phonological awareness tests. As a group, adults with dyslexia enrolled in this study had a lower reading age according to the l’Alouette test (p o 0.001) and significant deficits compared with

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

controls on word/nonword decoding tests. They obtained lower scores on phonological word (p o 0.05) and nonword (p o 0.05) decoding and were slower in both word and nonword decoding (words—phonological decoding: p o 0.005; nonwords: p o 0.001; words—orthographic decoding: p o 0.05). Spelling was also affected (word spelling: phonological words: p o 0.05; nonwords: p o0.05; sentence spelling—grammar: p o 0.01), as well as phonological awareness, with participants with dyslexia being both weaker and slower on the phoneme deletion test (scores: po 0.005; time: p o 0.001) and slower on the acronym test (p o0.005). They were also slower in performing the letter sequence test (p o0.005). Performances on Bell's test and the Rapid Automatized Naming test did not differ between the two groups (both p4 0.05). Finally, participants with dyslexia showed impaired working memory on the digit span test (ascending and descending digit span, both p o 0.05), whereas no deficit was observed on the word and nonword repetition test (both p 40.05).

2.3. Stimuli Stimuli were created by inserting a target word into a randomly selected 5-s duration of babble noise. Target words were inserted 2.5 s after the onset of the noise to allow participants to familiarize themselves with the noise before presentation of a target word (see Dole et al., 2012).

2.3.1. Target words Targets were 180 disyllabic words selected within the middle range of lexical frequency (ranging from 0.55 to 97.77 occurrences per million, mean [M] ¼13.44, SD ¼ 16.57) according to the French database Lexique 2 (New, Pallier, Brysbaert, & Ferrand, 2004). They were pronounced by a 27-year-old French woman, recorded in a soundproof room and sampled and saved with a resolution of 44 kHz.

2.3.2. Background noise As concurrent noise, we used background babble composed of 4 voices (2 males, 2 females). Every talker was originally recorded in a soundproof room, reading extracts from a French book. Individual recordings were modified according to the following protocol: (i) removal of silences and pauses of more than 1 s; (ii) deletion of sentences containing pronunciation errors, exaggerated prosody or proper nouns; (iii) removal of low-amplitude background noise using noisereduction techniques optimized for speech signals (CoolEdit Pro& 1.1, Dynamic Range Processing, preset Vocal limiter); (iv) intensity calibration in dB-A and normalization of each source at 70 dB-A; (v) final mixing of the 4 sources in final soundtracks; and (vi) cutting each soundtrack into 150 segments of 5 s each.

2.3.3. Listening configurations: Monaural, Binaural and Dichotic Finally, each target word was inserted into a randomly selected section of noise according to three listening configurations: Monaural, Binaural and Dichotic (Fig. 1). In the Monaural configuration, both the target word and noise were presented to the right ear at a signal-to-noise ratio of 0 dB, with no sound being presented in the left ear. This configuration was the most difficult configuration, with both signals being encoded in the same cochlea, thus maximizing energetic masking. In the Binaural configuration, both the target word and noise were presented in both ears, but with an ILD of 10 dB in favor of the right ear for the noise, which created the illusion that the target speech and babble in the background were originating from two different spatial locations. Finally, in the Dichotic configuration, target words were presented to the right ear, whereas noise was presented to the left ear at equal intensity. This configuration was designed to represent a maximal separation between the target and background, thus leaving only informational masking. The 180 words were evenly split across conditions, resulting in 60 words per condition. Between the three conditions, word frequency and phoneme number were counterbalanced.

2.4. Procedure 2.4.1. Task and procedure Inside the MRI scanner, participants were asked to carefully listen to the stimuli and judge the intelligibility of target words on a subjective scale ranging from 1 (unintelligible) to 4 (very intelligible). A response screen consisting of an arrow with four scales appeared immediately after the presentation of auditory stimuli, and responses were provided by button press via a 4-button box placed under the right hand of participants. The 180 stimuli were divided into two sessions of 90 items each, separated by a short pause; word frequency and phoneme number were counterbalanced between sessions. Stimuli were presented with variable ISI (mean duration: 4.5 s, range: 1–11 s). Auditory stimuli were presented with an audio system compatible with high magnetic fields (MR-Confon). Practice items were presented to participants before entering the scanner to ensure good comprehension of the instructions and to familiarize them with the presentation mode and the target voice.

Fig. 1. Conditions used in the experiment. S indicates the target speech, and N indicates background babble noise. 2.4.2. fMRI data acquisition MR imaging was performed at the Center for Neuroimaging at the hospital La Timone (Marseilles, France). A 3.0 T Brucker Medspec 30/80 AVANCE scanner was used. A T2-weighted gradient-echo-planar imaging (EPI) sequence sensitive to blood oxygen level-dependent contrast was used for functional scans (repetition time, 2400 ms; echo time, 30 ms; flip angle, 811; field of view, 192  192 mm; matrix size, 64  64). In total, 36 axial slices covering the entire brain were acquired in an interleaved mode with a voxel size of 3  3  3 mm3. A total of 801 volumes were acquired. To correct images for geometric distortions, a B0 fieldmap was obtained from two gradient-echo data sets acquired with a FLASH sequence. The fieldmap was subsequently used during data pre-processing. A T1-weighted highresolution three-dimensional anatomical volume was also acquired using a MPRAGE sequence: repetition time, 9.4 ms; echo time, 4.42 ms; flip angle, 301; field of view, 256  256  180 mm3; voxel size, 1  1  1 mm3.

2.4.3. Behavioral post-test To ensure that subjective ratings of intelligibility constituted a fair evaluation of the real intelligibility of the stimuli, participants completed a speech-in-speech perception test following the MR image acquisition, this time evaluating objective intelligibility. Configurations were the same as those in the scanner, and participants had to repeat the words they heard. Half of the target words had already been presented during MR acquisition, and half were new words. The number of trials was limited to 60, resulting in 20 words per condition, and the number of phonemes and word frequency were counterbalanced between the conditions. The stimuli duration, word onset and voices were the same as those in the MRI experiment. Briefly, analyses confirmed that the behavioral results followed the same patterns inside and outside the scanner (see Supplementary material S1 for further details).

2.5. Statistical analyses 2.5.1. Subjective intelligibility scores Mean subjective intelligibility scores were computed for each configuration and each subject, and a two-way analysis of variance (ANOVA) was performed, with Group (Typical Readers vs. Adults with Dyslexia) as the between-subjects factor and Configuration (Dichotic vs. Binaural vs. Monaural) as the within-subjects factor. Statistical tests were performed with a threshold of p o 0.05, and post hoc t-tests were performed for statistically significant effects with a correction for multiple comparisons (false discovery rate, FDR; Benjamini & Hochberg, 1995).

2.5.2. Functional data analysis fMRI data analyses were performed using the SPM8 statistical parametric mapping software (Wellcome Department of Cognitive Neurology, UK, www.fil.ion. ucl.ac.uk/spm/software/spm8/), running under Matlab 7.9 (Mathworks Inc., Natick, USA) in combination with the SPM extension Anatomical Automatic Labeling (Tzourio-Mazoyer et al., 2002) to localize the effects. The first 5 volumes of each scanning session, during which the MR signal reaches a steady state, were discarded. Data pre-processing included slice-time correction, realignment of images on the first volume of each session, unwarping, co-registration and normalization using the DARTEL tool from SPM8. To this end structural T1-weighted scans of the 28 participants were segmented into different tissue types. Intensity averages of the gray and white matter images were generated for use as an initial template for DARTEL registration (6 iterations). This template was aligned with the MNI template using affine transform, and each functional scan was then aligned with this template. Finally, images were smoothed by a 10-mm full-width halfmaximum (FWHM) Gaussian kernel and high-pass filtered (cut-off of 128 s). Data analysis was performed using the general linear model (GLM; Friston, Frith, Frackowiak, & Turner, 1995) as implemented in SPM8. Global differences in scan intensities were removed by scaling each fMRI value in proportion to the global intensity of the scan. To model the design, onsets of each target word were taken, with an event duration set to 1 s to approximately match the duration of the words. The four conditions (Dichotic, Binaural, Monaural and motor response) and a silent baseline condition (inter-stimulus intervals) were modeled as five regressors convolved with a canonical hemodynamic response function (HRF). The GLM was then used to generate parameter estimates of activity in each voxel and for each condition. Statistical parametric maps were generated from linear contrasts between the HRF parameter estimates for the different conditions and for

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

between-condition contrasts and then entered into the second-level randomeffects group analysis. Second-level analyses were performed using a random effects model (Friston, Holmes, Price, Büchel, & Worsley, 1999). The goals of this study were two-fold: first, evaluate the neural activity associated with speech-in-speech perception according to different listening configurations in a control (typical reading) population; second, test differences between participants with dyslexia and typical reading controls. To this end, regions of activation were first identified in typical readers only. A flexible factorial model was used, with Subject and Configuration as main factors. In this first analysis, we were interested in the specific effects of each configuration compared with those of the two other configurations; therefore, we contrasted each configuration versus another (for example, for the configuration A versus configurations B and C comparison, the contrasts were A 4B and A 4C). These two contrasts were thresholded at p o 0.0005 uncorrected, with a cluster extent defined at p o 0.005 determined using a Monte Carlo simulation procedure (Slotnick, Moo, Segal, & Hart, 2003). The two resulting maps were then combined as a single map representing both contrasts, A 4B and A 4C. This first analysis enabled the definition of functional regions of interest (ROIs) using MarsBar Toolbox (http://marsbar.sourceforge.net/; Brett, Anton, Valabregue, & Poline, 2002) based on the five main local maxima obtained from the entire group analysis, which served as a basis for further between-group analyses. These five ROIs were located in the left Heschl gyrus/STG (MNI coordinates: [  39  24 9]), right STG ([63  15 6], supplementary motor area (SMA: [0 18 57]), left medial frontal gyrus (MFG)/IFG ([  45 15 36]) and right precentral gyrus/IFG ([42 6 30]). To achieve the second goal, between-group effects were explored with a second flexible factorial model that added Group as a factor. To compare the effects of introducing different degrees of spatial separation in the two groups, ROI analyses were performed on the Group x Configuration interaction; we focused on the Binaural4Monaural, Dichotic4Binaural and Dichotic4 Monaural contrasts. The statistical threshold was set at p o 0.05 FDR corrected. Finally, to better understand the specific changes in activity elicited by each configuration in each group, percent signal change values were also extracted for each participant, each ROI and each configuration using an in-house toolbox. For each ROI, a two-way ANOVA was performed with Group as the between-subjects factor and Configuration as the within-subjects factor. Differences were considered significant when po 0.05. Post hoc t-tests were performed on statistically significant differences, with a FDR correction for multiple comparisons.

3. Results 3.1. Behavioral data The results of the ANOVA performed on subjective intelligibility scores revealed a significant main effect of Configuration (F(2, 52) ¼100.25, p o0.001); post hoc comparisons showed that all configurations differed significantly from each other (all p o0.05), with subjective intelligibility increasing when spatial separation between the target and background increased (Monaural: mean ¼2.70; SD ¼0.58; Binaural: mean ¼3.50; SD ¼ 0.30; Dichotic: mean ¼3.69; SD ¼0.32). We also obtained a significant Group  Configuration interaction (F(2, 52) ¼ 7.66, p o0.005), demonstrating that the effects of Configuration differed between the two

5

groups of participants. In particular, participants with dyslexia experienced more difficulty in the Monaural configuration (M ¼2.50; SD ¼0.57) compared with typical readers (M ¼2.90; SD¼ 0.53; p o0.05), but this was not the case for the Binaural and Dichotic configurations (both p 40.05; Fig. 2). Taken together, these results demonstrated reduced performance in participants with dyslexia during speech-in-speech perception when the target word and the background were co-localized, but not when spatial information between the two sources was available.

3.2. Functional data To explore cortical activation elicited by speech-in-speech perception in different listening configurations, whole-brain activation was first analyzed in the typical readers group only (n¼14). To identify functional networks specifically engaged in each listening configuration, we first contrasted each individual configuration against the two other configurations: [Dichotic4Monaural] and [Dichotic4Binaural]; [Binaural4Dichotic] and [Binaural4Monaural]; [Monaural4Dichotic] and [Monaural4Binaural]. As shown in Fig. 3 and Table 2, in typical readers, the Dichotic configuration contrasted against the Monaural configuration resulted in increased activation in the right STG (BA 22), whereas the Dichotic configuration contrasted against the Binaural configuration revealed an increased activation in the right BA 42. The Binaural configuration contrasted against the Dichotic configuration revealed increased activation in the left Heschl gyrus, extending into the STG and left pars orbitalis (BA 47). Contrasted against the Monaural configuration, the Binaural configuration revealed increased activation in the right BA 41. The Monaural configuration contrasted against the two other configurations resulted in a network of frontal activations, including the left supplementary motor area (SMA), left MFG extending into the left IFG and right precentral gyrus extending into the right IFG. In summary, the two configurations in which spatial information was introduced mainly produced activation in the STG, whereas the Monaural configuration elicited activation in a more extensive neural network, including inferior and medial frontal regions, as well as motor regions. To answer our second research question and examine group differences in this activation, we used the five main clusters derived from these first results to define five functional ROIs: left and right STG, left MFG/IFG (hereafter labeled ‘left IFG’), right precentral gyrus/IFG (hereafter labeled ‘right IFG’) and SMA. The effect of introducing different degrees of spatial separation was

Fig. 2. Subjective intelligibility scores obtained for typical readers (black line) and participants with dyslexia (gray line). In the Dichotic configuration, target speech was presented in the right ear, whereas background noise was presented in the left ear. In the Monaural configuration, both target speech and background noise were presented in the right ear. In the Binaural configuration, target speech and background noise were presented in both ears, but with an interaural level difference of 10 dB for the noise. The asterisk indicates a statistically significant difference between typical readers and adults with dyslexia. Error bars represent standard error.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 3. fMRI activation in the typical readers group for each configuration contrasted against the two other configurations, displayed on the cortical surface of the brain. Top: [Dichotic4Binaural] and [Dichotic4Monaural] contrasts. Middle: [Binaural4Dichotic] and [Binaural4Monaural] contrasts. Bottom: [Monaural4 Dichotic] and [Monaural4Binaural] contrasts. The displayed results are significant at po0.0005 uncorrected, with a cluster extent of po0.005.

investigated by exploring group differences in three contrasts, namely, Binaural4Monaural, Dichotic4Binaural and Dichotic 4 Monaural. For the Binaural4Monaural contrast, this analysis revealed larger activation in the right STG (p o0.001 FDR; Fig. 4, left) as well as lower activation in the SMA (po 0.05 FDR) in participants with dyslexia compared with typical readers. No significant differences were observed for the Dichotic4Binaural and Dichotic4 Monaural contrasts (all p 40.05). Finally, to better understand the specific effects of each individual configuration, we extracted the percent signal change in each ROI for each subject and each configuration. For each ROI, a twoway ANOVA was performed, with Group (typical readers, adults with dyslexia) as the between-subjects factor and Configuration (Dichotic, Binaural, Monaural) as the within-subjects factor. For bilateral IFG and left STG, we obtained a significant effect of Configuration only (left IFG: F(2, 52)¼ 47.7, po0.001; right IFG: F(2, 52)¼62.260, po0.001; left STG: F(2, 52)¼23.17, po0.001; see Fig. 5A–C). Further exploration revealed a gradual increase in activity in the left and right IFG with decreasing spatial separation between the target and noise, with the Monaural configuration eliciting greater activation than the Binaural configuration (both po0.001 FDR-corrected), which produced greater activation than the Dichotic

configuration (left IFG: po0.01 FDR; right IFG: po0.001 FDR). In the left STG, we obtained a lower activation in the Dichotic configuration than in the two other configurations (both po0.001 FDR), and there was a tendency toward a greater activation in the Binaural configuration compared with the Monaural configuration (p¼ 0.080 FDR). The effect of Group tended to be significant for the left IFG only (F(1, 26)¼3.31; p¼0.080), indicating that participants with dyslexia tended to show weaker activation in the left IFG compared with typical readers, regardless of the configuration. To better characterize the activation pattern in the left and right IFG, the percent signal change was also tested against 0 in each group using independent t-tests. This analysis confirmed the progressive increase in activation in the bilateral IFG when decreasing spatial separation among typical readers, for whom activation tended toward significance in the Dichotic configuration (left IFG: p¼ 0.078; right IFG: p ¼0.082) and was significant in the two other configurations (Binaural: left IFG: p o0.005; right IFG: po 0.05; Monaural: left IFG: p o0.001; right IFG: po 0.001). In participants with dyslexia, the left and right IFG were not significantly activated in the Dichotic and Binaural configurations (all p 40.05), but they were significantly activated in the Monaural configuration (left IFG: p o0.05; right IFG: p o0.01). For the SMA and the right STG, ANOVAs yielded a different pattern of results, with a significant Group x Configuration interaction for the right STG (F(2, 52) ¼ 3.73, p o0.05) that only tended toward significance for the SMA (F(2, 52) ¼2.94; p¼ 0.061; see Fig. 6A and B). For the SMA, this interaction revealed that typical readers showed a gradual decrease in activation with increasing spatial separation (Dichotic compared with Binaural, p o0.01; Binaural over Monaural, po 0.005). Conversely, for participants with dyslexia, we observed greater activation in the SMA for the Monaural configuration than for the two other configurations (both p o0.001), and activation did not differ between the Binaural and Dichotic configurations (p ¼0.419). One-sample t-tests performed against 0 confirmed that whereas activation in typical readers was non-significantly different from 0 in the Dichotic configuration (p ¼0.600), tended toward significance in the Binaural configuration (p ¼0.071) and was significant in the Monaural configuration (p ¼0.014), this was not the case for participants with dyslexia, for whom activation was never significant regardless of configuration (all p4 0.05). Finally, the right STG in typical readers showed a reversed pattern of activation compared with the other regions, with a gradual increase in activation with increasing spatial separation (Monaural: 1.01%, Binaural: 1.592%, Dichotic: 2.11%, Binaural vs. Monaural: po 0.05, Dichotic vs. Binaural: p o0.05). This situation was not the case in participants with dyslexia, for whom activation was at a similar level in the Binaural (2.18%) and Dichotic configurations (2.22%; Dichotic vs. Binaural: p ¼0.92; Binaural vs. Monaural: p o0.001). Taken together, these results suggest that, in typical readers, decreasing spatial separation is accompanied by a gradual increase in activation in the bilateral IFG and SMA, whereas the reverse pattern was observed in the right STG. Participants with dyslexia did not show any significant activation in the SMA regardless of configuration, but they did show increased activation in the right STG in a spatial unmasking situation. Finally, to evaluate the brain/behavior relationship, in each ROI, we subtracted the percent signal change values obtained for the Monaural configuration from those obtained for the Binaural configuration and correlated these differences with the intelligibility score changes associated with the Binaural–Monaural configuration (amount of spatial unmasking). The latter analysis resulted in a positive correlation between cerebral activation and behavior in the right STG only (F(1, 26)¼ 11.92, p o0.005; r ¼0.561), indicating that greater activation in the right STG was associated with an increase in the amount of unmasking of speech

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

7

Table 2 Activations obtained for the contrasts: [Dichotic>Binaural] and [Dichotic>Monaural], [Monaural>Dichotic] and [Monaural>Binaural], and [Binaural>Dichotic] and [Binaural>Monaural], in the typical readers group. AAL label

BA

Cluster size (voxels)

p Cluster (FWE)

p Voxel (FWE)

p Voxel (uncorr.)

Tmax

MNI coordinates x

y

z

[Dichotic>Binaural] and [Dichotic>Monaural] Dichotic>Binaural Right STG Dichotic>Monaural Right STG

42

98

0.001

0.102

o0.001

63

 15

9

5.66

22

347

o 0.001

0.001

o0.001

63

 15

6

7.72

[Monaural>Dichotic] and [Monaural>Binaural] Monaural>Dichotic Right precentral (extending in pars opercularis) Left MFG (extending in pars opercularis) Left STG Left SMA (extending in SFG)

9 9 42 8

440 180 43 291

o 0.001 o 0.001 0.0367 o 0.001

0.0096 0.0214 0.0583 0.0625

o0.001 o0.001 o0.001 o0.001

42  45  57 0

6 15  36 18

30 36 12 57

6.74 6.42 5.94 5.9

8 9

87 39

0.0019 0.0499

0.0299 0.5539

o0.001 o0.001

45  39

6 15

45 33

6.26 4.69

[Binaural>Dichotic] and [Binaural>Monaural] Binaural>Dichotic Left Heschl gyrus (extending in STG) Left pars orbitalis (extending in pars opercularis)

13 47

174 30

o 0.001 0.1024

0.0917 0.6041

o0.001 o0.001

 39 42

 24 27

9 3

5.71 4.62

Binaural>Monaural Right STG

41

60

0.0108

0.3425

o0.001

54

 18

6

5.01

Monaural>Binaural Right Precentral Left IFG (pars opercularis)

Statistical threshold set at p o 0.0005, with a cluster extent of p o 0.005.

Fig. 4. Left: between-group effects (adults with dyslexia4typical readers) in the Binaural4Monaural contrast displayed on the cortical surface of the brain. For representation purposes, the displayed results are thresholded at po 0.001 uncorrected. Right: percent signal change in the right STG ROI in the Monaural configuration subtracted from the Binaural configuration, displayed as a function of behavioral unmasking (Binaural–Monaural intelligibility scores). Black dots: typical readers. Gray dots: adults with dyslexia.

Fig. 5. Percent signal change values for the Dichotic, Binaural and Monaural configurations in typical readers and adults with dyslexia in the left IFG (A), right IFG (B) and left STG (C) functional ROIs.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

Fig. 6. Percent signal change values for the Dichotic, Binaural and Monaural configurations in typical readers and adults with dyslexia in the SMA (A) and right STG (B) functional ROIs.

in the Binaural configuration compared with the Monaural configuration (Fig. 4, right).

4. Discussion Our study was designed to evaluate cortical mechanisms involved in speech-in-speech perception in participants with dyslexia compared with typical reading controls. More specifically, we evaluated modulations of activation in functional networks in Monaural, Binaural and Dichotic listening configurations. The main finding confirms the presence of a significant speech-in-speech perception deficit in adults with dyslexia. Importantly, this finding highlights the dependency of this impairment on the listening configuration by showing that participants with dyslexia demonstrate speech-in-speech perception impairments only when target and background sound sources are co-localized. Conversely, when spatial information was provided, the scores obtained by participants with dyslexia did not differ from those of typical readers. This result is in agreement with our previous findings (Dole et al., 2012) and suggests the presence of typical spatial processing abilities in adults with dyslexia. The second important finding was the presence of increased activation in the right STG in a spatial unmasking situation in participants with dyslexia compared with typical readers, which suggests that the typical behavioral spatial abilities of adults with dyslexia are associated with atypical activation patterns involving the right STG in particular. In the following section, we will discuss these findings and their implications in the context of dyslexia. Generally, functional networks identified in the typical readers group confirmed the involvement of several language-related regions in speech-in-speech perception, mainly bilateral STG, bilateral IFG, right precentral gyrus, left MFG and SMA (Salvi et al., 2002; Scott et al., 2004, 2009; Scott, Rosen, Lang, & Wise, 2006; Wong et al., 2008, 2009; Zekveld et al., 2006). These first results also highlighted the differential involvement of these areas according to listening configuration; specifically, whereas the Dichotic and Binaural configurations specifically involved portions of the STG, the Monaural configuration, which was the most difficult configuration because of the absence of spatial cues, seemed to also involve a large prefrontal network including the bilateral IFG and MFG, extending into the SMA. This result highlights the effect of spatial information on the energetic/informational masking balance. Indeed, in the Monaural configuration, both signals were encoded in the same cochlea, which

consequently increased the amount of energetic masking. It is thus likely that the observed activation was related to increased difficulties in signal extraction associated with this type of masking. In agreement with this hypothesis, Scott et al. (2004), in an experiment designed to specifically test energetic masking, found more activation in the frontal pole and dorsolateral frontal and parietal regions for speech-in-noise compared with speech-inspeech perception. In the same context, Adank, Davis, and Hagoort (2012) observed increased activation in the bilateral IFG, frontal operculum and right cingulate gyrus for speech perceived in background noise compared with speech-in-silence. Furthermore, the presence of bilateral IFG involvement in situations of energetic masking or reduced intelligibility due to distortions of the speech signal is well established (Davis & Johnsrude, 2003; Zekveld et al., 2006). In this framework, Giraud et al. (2004) used natural speech as well as simple and complex speech envelope noises to differentiate the mechanisms involved in speech perception from the attentional mechanisms specifically directed at detecting phonetic indices in noisy stimuli. Before training, participants were unable to understand sentences presented concurrently with either simple or complex speech envelope noise, whereas after training, the sentences produced with the complex speech envelope noises became understandable. By comparing brain activation before and after training, they found that Broca's area (BA 44) was activated during effortful searching for phonetic cues. This activation would therefore be specifically involved in an enhanced processing of phonological information during difficult listening situations. In the same study, the authors also showed that the interaction of auditory attention and successful comprehension occurred in the anterior cingulate, bilateral anterior insula and right MFG (BA 9). Interestingly, in our study, this MFG activation extended to the SMA and could be related to the motor theory of speech perception, which argues for an involvement of the motor system in speech perception, particularly in challenging listening situations (Bishop & Miller, 2009; D’Ausilio, Bufalari, Salmas, & Fadiga, 2012; Pulvermüller et al., 2006; Pulvermüller & Fadiga, 2010). According to this theory, during speech perception, motor primitives are activated as a result of an acoustically evoked motor resonance. This theory is supported by the observations that passive listening to syllables involves motor and premotor areas (Fadiga, Craighero, Buccino, & Rizzolatti, 2002; Iacoboni, 2008; Pulvermüller et al., 2006) and that the presupplementary motor area is involved in the perception of degraded speech (Adank & Devlin, 2010; Shahin, Bishop, & Miller, 2009). In summary, we can thus conclude that the frontal activations observed in the Monaural configuration were related to the increased

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

complexity of speech perception due to increased energetic masking. As a consequence, to successfully achieve the task, participants needed to rely more on phonological decoding and alternative networks, such as motor and attentional networks. Surprisingly, between-group analysis did not reveal any significant activation differences between typical readers and participants with dyslexia in the Monaural configuration; this result was unexpected, as we observed clear and significant behavioral deficits in the Dyslexia group, both in subjective and objective intelligibility (Dole et al., 2012). Given the consistency of results showing, on one hand, activation differences in the superior temporal regions and inferior frontal regions in individuals with dyslexia in various tasks (Cao, Bitan, Chou, Burman, & Booth, 2006; Hoeft et al., 2006; Kronbichler et al., 2006; McCrory, Frith, Brunswick, & Price, 2000; Paulesu et al., 2001) and, on the other hand, the involvement of these regions in speech-in-noise processing (Scott et al., 2004,, 2009), we would have expected a significant deficit in at least one of these regions. However, careful examination of the percent signal change for each configuration allowed us to highlight some subtle differences in the activation patterns between the two groups of participants. Indeed, whereas the activation in the frontal and motor regions became progressively significant in typical readers when reducing the amount of spatial separation between speech and background noise, this was not the case for participants with dyslexia, especially in the SMA, which was never significantly activated. This result suggests an atypical pattern of functionality in the SMA in participants with dyslexia, which could be at the origin of their speech-in-speech difficulties. Indeed, the SMA has been implicated in the perception of degraded speech (Adank & Devlin, 2010; Shahin et al., 2009), and its involvement has been proposed to be related to the mobilization of articulatory representations, which would be of particular interest in the context of hearing speech in background noise. Another possible role that has been proposed for the SMA during speech perception is in the processing of temporal information and speech regularities during speech perception (Geiser, Zaehle, Jancke, & Meyer, 2008). According to a model developed by Kotz and Schwartze (2010), the SMA is involved in the elaboration of a representation of the temporal structure of speech. It is thus possible that the additional amount of energetic masking created by the absence of spatial information in the Monaural configuration results in the need for the listener to rely more on the temporal structure of speech. The fact that this region seems to be less activated in adults with dyslexia would thus be in accordance with their difficulties with temporal processing that have been observed (Goswami, 2011; Lorenzi, Dumont, & Füllgrabe, 2000). Although still speculative, these considerations clearly highlight the need for further investigation of both the cerebral bases of the speech-in-speech deficit in dyslexia and the precise role of the SMA in the cocktail party phenomenon. In the two other configurations, spatial separation between target and background speech helped participants achieve the task. As a consequence, intelligibility increased in these configurations, leading to a significant reduction of energetic masking and, in the Dichotic condition, leaving only informational masking. Accordingly, in both groups, we observed reduced activation in the frontal and motor regions for these two configurations. Indeed, the task became simpler, and participants needed to rely less on phonological decoding and motor/attentional networks. However, there was clear involvement of the right STG, revealed by the increase in activation in this region for the Binaural4 Monaural and Dichotic4Monaural contrasts. The percent signal changes extracted for each configuration also confirmed this observation and more specifically showed a gradual increase in activation in the right STG along with increasing spatial separation between the target and background, at least among typical readers. This right

9

STG activation could reflect the fact that decreasing the energetic masking in the Binaural and Dichotic configurations increases the proportion of informational masking, leading to activation in the regions associated with informational masking for the Binaural4 Monaural and Dichotic 4Monaural contrasts. Indeed, Scott et al. (2004, 2009) demonstrated an involvement of the STG in informational masking, specifically attributing the right STS/STG activation to processes of selection between competing voices on the basis of acoustic cues and pitch variation (Scott et al., 2009). Following this hypothesis, we can interpret our findings as evidence of facilitated access to these cues allowed by increased spatial separation, which enhanced their processing in the right STG, thus improving intelligibility. In support of this explanation, we also found a positive correlation between the right STG activation and the increase in intelligibility in the Binaural configuration compared with the Monaural configuration. This argument is also reinforced by the fact that the posterior STG has been shown to be involved in spatial localization processes, specifically in the integration between localization mechanisms and segregation between competing auditory flows on the basis of spectro-temporal cues (Middlebrooks, 2002; Zatorre, Bouffard, Ahad, & Belin, 2002). The hypothesis that common regions are involved in the integration of localization and spectro-temporal cues is further strengthened by a recent study that found an interdependent neural encoding of localization, pitch and timbre in the auditory cortex (Bizley, Walker, Silverman, King, & Schnupp, 2009). In addition, spatial localization is known to engage the right hemisphere more than the left (Kaiser, Lutzenberger, Preissl, Ackermann, & Birbaumer, 2000; Itoh, Yumoto, Uno, Kurauchi, & Kaga, 2000; Palomäki, Alku, Mäkinen, May, & Tiitinen, 2000; Zatorre et al., 2002), particularly for the processing of ILDs (Johnson & Hautus, 2010; Palomäki, Tiitinen, Mäkinen, May, & Alku, 2005; Spierer, Bellmann-Thiran, Maeder, Murray, & Clarke, 2009; Tardif, Murray, Meylan, Spierer, & Clarke, 2006). These results therefore suggest that the STG activation obtained in the two configurations in which spatial information was available was related to processes of integration of spatial localization and segregation between competing voices based on the processing of pitch or timbre variations. In this context, participants with dyslexia achieved the same scores as those of typical readers, which suggests that they were able to use the spatial information contained in our stimuli. However, a different pattern of activation between the two groups was obtained for the Binaural4Monaural contrast, with participants with dyslexia exhibiting increased activation in the right STG compared with typical readers. This result confirms the finding that they were able to use spatial information to overcome their difficulties encountered in the Monaural configuration, as well as the idea that this effect could be due to a reallocation of neural resources, with decreased involvement of frontal networks being compensated for by increased activity in the superior temporal regions. 4.1. Speech in noise and dyslexia Taken together, our results provide further evidence for a significant speech-in-speech perception deficit in dyslexia. A previous study (Ziegler et al., 2009), which demonstrated the presence of a significant deficit for both internal (distorted speech) and external (noise added to speech) noise, suggested that the difficulties experienced by participants with dyslexia are related to the processing of noisy phonological information. The present results, together with those of our previous study (Dole et al., 2012), confirm this finding by highlighting the specificity of this deficit in complex energetic masking situations. The fact that adults with dyslexia performed as well as typical readers in the

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

two configurations in which spatial information was available suggests that it was not the concurrent linguistic information that created difficulties, but rather the additional amount of energetic masking. These results are in line with studies highlighting the absence of a general deficit of speech-in-noise perception in dyslexia, suggesting that these difficulties are not a matter of impoverished phonological representations per se, but more likely represent a problem associated with gaining access to these phonological representations (Boets et al., 2013; Hazan et al., 2013; Ramus & Szenkovitz, 2008). fMRI results extend these observations by showing that speech-in-noise processing involves a dynamic frontotemporal network, modulated by the amount of spatial information available. Participants with dyslexia seem to exhibit functional deficits in frontal-motor networks in speech-in-speech situations. However, they also showed increased activation in the right superior temporal regions, which clearly suggests that they develop different processing strategies to overcome their speech-in-speech difficulties, notably involving increased metabolic activity in the right superior temporal regions. It is possible that this increased right STG activation reflects the greater allocation of neural resources to the integration of spatial information and vocal information to improve the processing of speech when masked by surrounding noise in adults with dyslexia. Right hemisphere over-activation has indeed been consistently observed in dyslexia and is generally interpreted as a compensation effect that develops over a lifetime (Démonet, Taylor, & Chaix, 2004; Pugh et al., 2000; Richlan, Kronbichler, & Wimmer, 2009; Shaywitz et al., 2002; Shaywitz & Shaywitz, 2005). Sarkari et al. (2002), using magnetoencephalography, found increased engagement of the right temporoparietal areas in children with dyslexia, and Rumsey et al. (1999) found that temporoparietal activation in the right hemisphere was correlated with reading performance in the dyslexia group only, which suggests an involvement of the right hemisphere in compensatory functions. In light of these results, and given the fact that speech-in-noise is a situation encountered daily, it would not be surprising that compensatory strategies are developed for speech perception, and our study suggests that this compensation involves the right superior temporal regions. The fact that the participants with dyslexia in our study were, for the most part, students who had achieved high-level degrees is also of interest. Indeed, they achieved sufficiently good literacy abilities to pursue high-level studies, which suggests that they are at least partially compensated, thus maximizing the chance to observe compensation. It may be interesting to test the same effects in non-compensated adults with dyslexia to determine whether the same results are found. One possible limitation of this study could be the relatively small sample of participants (14 vs. 14). It is possible that with an increased number of participants, the observed effects, particularly in the frontal/motor regions, would have been more significant and the reported results should be taken into account considering this potential limit, in particular when generalizing them to the entire dyslexia population. However, the present study is the first study to investigate the neural correlates of speech-in-noise perception difficulties in participants with dyslexia with fMRI, and it is the first study to suggest increased activation in the right STG together with decreased activation in regions involved in motor aspects of speech perception.

5. Conclusion First, our results aid in the elucidation of the cortical mechanisms underlying speech-in-noise perception in a complex energetic masking situation with no spatial information between the

target and background by showing a strong dependence on the neural networks linked to phonological decoding and motor regions. In addition to a clear behavioral impairment, adults with dyslexia appear to suffer from functional deficits in cortical regions that have been associated with the processing of noisy speech, suggesting that their behavioral responses may result from impairment at this level. In the Binaural listening situation, a significant release from energetic masking occurred for all participants. In this configuration, participants with dyslexia were able to optimally use spatial information, showing no speech-in-speech deficits. This finding appears to be a result of the reallocation of neural resources in participants with dyslexia, with increased activation in the right superior temporal regions. This result suggests a functional plasticity in adults with dyslexia, which could help them manage their speech-in-speech difficulties. These results provide us with a better understanding of speech-in-noise processing in dyslexia, and future research should provide further characterization of right-hemisphere involvement.

Funding This work was supported by a European Research Council grant [SpiN project 209234 to F.M.] and a Ph.D. grant from the RhôneAlpes Region, France (‘Cluster régional de recherche: Handicap, Vieillissement et Neurosciences’), to M.D.

Uncited reference Studebaker (1985).

Acknowledgments We are extremely grateful to all of the adults with dyslexia and participants in the control group who participated in this study. We especially thank Jean-Luc Anton, Muriel Roth and Bruno Nazarian, from the Brain Imaging Centre La Timone (Marseilles), for their invaluable help with fMRI data collection and analyses, as well as Dr. Evelyne Veuillet and Aurore Gautreau for additional help with data collection and stimuli generation. We are also grateful to Emilie Cousin and Cédric Pichat for their help with data analysis.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.neuropsychologia. 2014.05.016. References Adank, P., & Devlin, J. T. (2010). On-line plasticity in spoken sentence comprehension: adapting to time-compressed speech. NeuroImage, 49, 1124–1132. Adank, P., Davis, M. H., & Hagoort, P. (2012). Neural dissociation in processing noise and accent in spoken language comprehension. Neuropsychologia, 50, 77–84. Amitay, S., Ahissar, M., & Nelken, I. (2002). Auditory processing deficits in reading disabled adults. Journal of the Association for Research in Otolaryngology, 3, 302–320. Arbogast, T. L., Mason, C. R., & Kidd, G., Jr. (2002). The effect of spatial separation on informational and energetic masking of speech. Journal of the Acoustical Society of America, 112, 2086–2098. Bishop, C. W., & Miller, L. M. (2009). A multisensory cortical network for understanding speech in noise. Journal of Cognitive Neuroscience, 21, 1790–1805. Bizley, J. K., Walker, K. M., Silverman, B. W., King, A. J., & Schnupp, J. W. (2009). Interdependent encoding of pitch, timbre, and spatial location in auditory cortex. Journal of Neuroscience, 29, 2064–2075.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Q3 91 92 93 94 Q2 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B: Statistical Methodology, 57, 289–300. Boets, B., De Smedt, B., Cleuren, L., Vandewalle, E., Wouters, J., & Ghesquière, P. (2010). Towards a further characterization of phonological and literacy problems in Dutch-speaking children with dyslexia. British Journal of Developmental Psychology, 28, 5–31. Boets, B., Ghesquière, P., van Wieringen, A., & Wouters, J. (2007). Speech perception in preschoolers at family risk for dyslexia: Relations with low-level auditory processing and phonological ability. Brain and Language, 101, 19–30. Boets, B., Op de Beeck, H. P., Vandermosten, M., Scott, S. K., Gillebert, C. R., Mantini, D., et al. (2013). Intact but less accessible phonetic representations in adults with dyslexia. Science, 342, 1251–1254. Boets, B., Vandermosten, M., Poelmans, H., Luts, H., Wouters, J., & Ghesquière, P. (2011). Preschool impairments in auditory processing and speech perception uniquely predict future reading problems. Research in Developmental Disabilities, 32, 560–570. 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. Boulenger, V., Hoen, M., Ferragne, E., Pellegrino, F., & Meunier, F. (2010). Real-time lexical competitions during speech-in-speech comprehension. Speech Communication, 52, 246–253. Brady, S., Shankweiler, D., & Mann, V. (1983). Speech perception and memory coding in relation to reading ability. Journal of Experimental Child Psychology, 35, 345–367. Brett, M., Anton, J. L., Valabregue, R., & Poline, J. B. (2002). Region of interest analysis using an SPM Toolbox. NeuroImage, 16. Bronkhorst, A. W. (2000). The cocktail party phenomenon: A review of research on speech intelligibility in multiple-talker conditions. Acustica, 86, 117–128. Bronkhorst, A. W., & Plomp, R. (1988). The effect of head-induced interaural time and level differences on speech intelligibility in noise. Journal of the Acoustical Society of America, 83, 1508–1516. Brouwer, S., Van Engen, K. J., Calandruccio, L., & Bradlow, A. R. (2012). Linguistic contributions to speech-on-speech masking for native and non-native listeners: Language familiarity and semantic content. Journal of the Acoustical Society of America, 131, 1449–1464. Bruck, M., & Treiman, R. (1990). Phonological awareness and spelling in normal children and dyslexics: The case of initial consonant clusters. Journal of Experimental Child Psychology, 50, 156–178. Brungart, D. S. (2001). Informational and energetic masking effects in the perception of two simultaneous talkers. Journal of the Acoustical Society of America, 109, 1101–1109. Brungart, D. S., Simpson, B. D., Ericson, M. A., & Scott, K. R. (2001). Informational and energetic masking effects in the perception of multiple simultaneous talkers. Journal of the Acoustical Society of America, 110, 2527–2538. Cao, F., Bitan, T., Chou, T. L., Burman, D. D., & Booth, J. R. (2006). Deficient orthographic and phonological representations in children with dyslexia revealed by brain activation patterns. Journal of Child Psychology and Psychiatry, 47, 1041–1050. 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. Cherry, E. C. (1953). Some experiments on the recognition of speech, with one and two ears. Journal of the Acoustical Society of America, 25, 975–979. Cunningham, J., Nicol, T., Zecker, S. G., Bradlow, A., & Kraus, N. (2001). Neurobiologic responses to speech in noise in children with learning problems: Deficits and strategies for improvement. Clinical Neurophysiology, 112, 758–767. D’Ausilio, A., Bufalari, I., Salmas, P., & Fadiga, L. (2012). The role of the motor system in discriminating normal and degraded speech sounds. Cortex, 48, 882–887. Davis, M. H., & Johnsrude, I. S. (2003). Hierarchical processing in spoken language comprehension. Journal of Neuroscience, 23, 3423–3431. Démonet, J. F., Taylor, M. J., & Chaix, Y. (2004). Developmental dyslexia. Lancet, 363, 1451–1460. Divenyi, P. L. (2004). Speech segregation by humans and machines. Dordrecht, The Netherlands: Kluwer Academic Publisher. 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. 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. Drullman, R., & Bronkhorst, A. W. (2000). Multichannel speech intelligibility and talker recognition using monaural, binaural, and three-dimensional auditory presentation. Journal of the Acoustical Society of America, 107, 2224–2235. 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. Fadiga, L., Craighero, L., Buccino, G., & Rizzolatti, G. (2002). Speech listening specifically modulates the excitability of tongue muscles: A TMS study. European Journal of Neuroscience, 15, 399–402. Frauenheim, J. G. (1978). Academic achievement characteristics of adult males who were diagnosed as dyslexic in childhood. Journal of Learning Disabilities, 11, 476–483.

11

Freyman, R. L., Balakrishnan, U., & Helfer, K. S. (2001). Spatial release from informational masking in speech recognition. Journal of the Acoustical Society of America, 109, 2112–2122. Freyman, R. L., Helfer, K. S., McCall, D. D., & Clifton, R. K. (1999). The role of perceived spatial separation in the unmasking of speech. Journal of the Acoustical Society of America, 106, 3578–3588. Friston, K. J., Frith, C. D., Frackowiak, R. S., & Turner, R. (1995). Characterizing dynamic brain responses with fMRI: A multivariate approach. NeuroImage, 2, 166–172. Friston, K. J., Holmes, A. P., Price, C. J., Büchel, C., & Worsley, K. J. (1999). Multisubject fMRI studies and conjunction analyses. NeuroImage, 10, 385–396. Geiser, E., Zaehle, T., Jancke, L., & Meyer, M. (2008). The neural correlates of speech rhythm as evidenced by metrical speech processing. Journal of Cognitive Neuroscience, 20, 541–552. Giraud, A. L., Kell, C., Thierfelder, C., Sterzer, P., Russ, M. O., Preibisch, C., et al. (2004). Contributions of sensory input, auditory search and verbal comprehension to cortical activity during speech processing. Cerebral Cortex, 14, 247–255. Goswami, U. (2011). A temporal sampling framework for developmental dyslexia. Trends in Cognitive Science, 15, 3–10. Habib, M. (2000). The neurological basis of developmental dyslexia: An overview and working hypothesis. Brain, 123, 2373–2399. Hazan, V., Messaoud-Galusi, S., & Rosen, S. (2013). The effect of talker and intonation variability on speech perception in noise in children with dyslexia. Journal of Speech, Language, and Hearing Research, 56, 44–62. Hawley, M. L., Litovsky, R. Y., & Culling, J. F. (2004). The benefit of binaural hearing in a cocktail party: Effect of location and type of interferer. Journal of the Acoustical Society of America, 115, 833–843. Heim, S., & Keil, A. (2004). Large-scale neural correlates of developmental dyslexia. European Child & Adolescent Psychiatry, 13, 125–140. 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, L53–58. Hoeft, F., Hernandez, A., McMillon, G., Taylor-Hill, H., Martindale, J. L., Meyler, A., et al. (2006). Neural basis of dyslexia: A comparison between dyslexic and nondyslexic children equated for reading ability. Journal of Neuroscience, 26, 10700–10708. Hoen, M., Meunier, F., Grataloup, C., Pellegrino, F., Grimault, N., Perrin, F., et al. (2007). Phonetic and lexical interferences in informational masking during speech-in-speech comprehension. Speech Communication, 49, 905–916. Iacoboni, M. (2008). The role of premotor cortex in speech perception: Evidence from fMRI and rTMS. Journal of Physiology (Paris), 102, 31–34. Itoh, K., Yumoto, M., Uno, A., Kurauchi, T., & Kaga, K. (2000). Temporal stream of cortical representation for auditory spatial localization in human hemispheres. Neuroscience Letters, 292, 215–219. Jacquier-Roux, M., Valdois, S., Zorman, M., Lequette, C., & Pouget, G. (2005). ODEDYS: Un outil de dépistage des dyslexies version 2. Grenoble: Laboratoire Cogni-Sciences, IUFM de Grenoble. Johnson, B. W., & Hautus, M. J. (2010). Processing of binaural spatial information in human auditory cortex: Neuromagnetic responses to interaural timing and level differences. Neuropsychologia, 48, 2610–2619. Kaiser, J., Lutzenberger, W., Preissl, H., Ackermann, H., & Birbaumer, N. (2000). Right-hemisphere dominance for the processing of sound-source lateralization. Journal of Neuroscience, 20, 6631–6639. Kotz, S. A., & Schwartze, M. (2010). Cortical speech processing unplugged: A timely subcortico-cortical framework. Trends in Cognitive Science, 14, 392–399. Kronbichler, M., Hutzler, F., Staffen, W., Mair, A., Ladurner, G., & Wimmer, H. (2006). Evidence for a dysfunction of left posterior reading areas in German dyslexic readers. Neuropsychologia, 44, 1822–1832. Lefavrais, P. (1967). Le test de l’Alouette. Paris, France. Edition du Centre de Psychologie Appliquée. Lorenzi, C., Dumont, A., & Füllgrabe, C. (2000). Use of temporal envelope cues by children with developmental dyslexia. Journal of Speech, Language and Hearing Research, 43, 1367–1379. Lyon, G. R., Shaywitz, S. E., & Shaywitz, B. A. (2003). A definition of dyslexia. Annals of Dyslexia, 53, 1–14. McCrory, E., Frith, U., Brunswick, N., & Price, C. (2000). Abnormal functional activation during a simple word repetition task: A PET study of adult dyslexics. Journal of Cognitive Neuroscience, 12, 753–762. Middlebrooks, J. C. (2002). Auditory space processing: Here, there or everywhere? Nature Neuroscience, 5, 824–826. Nelson, H. E., & Warrington, E. K. (1980). An investigation of memory functions in dyslexic children. British Journal of Psychology, 71, 487–503. New, B., Pallier, C., Brysbaert, M., & Ferrand, L. (2004). Lexique 2: A new French lexical database. Behavior Research Methods, Instruments & Computers, 36, 516–524. Obleser, J., & Kotz, S. A. (2010). Expectancy constraints in degraded speech modulate the language comprehension network. Cerebral Cortex, 20, 633–640. Obleser, J., Wise, R. J., Alex Dresner, M., & Scott, S. K. (2007). Functional integration across brain regions improves speech perception under adverse listening conditions. Journal of Neuroscience, 27, 2283–2289. Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9, 97–113. Palomäki, K., Alku, P., Mäkinen, V., May, P., & Tiitinen, H. (2000). Sound localization in the human brain: Neuromagnetic observations. NeuroReport, 11, 1535–1538.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

M. Dole et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Palomäki, K. J., Tiitinen, H., Mäkinen, V., May, P. J., & Alku, P. (2005). Spatial processing in human auditory cortex: The effects of 3D, ITD, and ILD stimulation techniques. Brain Research. Cognitive Brain Research, 24, 364–379. Paulesu, E., Démonet, J. F., Fazio, F., McCrory, E., Chanoine, V., Brunswick, N., et al. (2001). Dyslexia: Cultural diversity and biological unity. Science, 291, 2165–2167. Pugh, K. R., Mencl, W. E., Jenner, A. R., Katz, L., Frost, S. J., Lee, J. R., et al. (2000). Functional neuroimaging studies of reading and reading disability (developmental dyslexia). Mental Retardation and Developmental Disabilities Research Reviews, 6, 207–213. Pulvermüller, F., & Fadiga, L. (2010). Active perception: Sensorimotor circuits as a cortical basis for language. Nature Reviews Neuroscience, 11, 351–360. Pulvermüller, F., Huss, M., Kherif, F., Martin, Moscoso del Prado, F., Hauk, O., et al. (2006). Motor cortex maps articulatory features of speech sounds. Proceedings of the National Academy of Science of the United States of America, 103, 7865–7870. Puolokanaho, A., Timo, A., Aro, M., Eklund, K., Leppänen, P. H. T., Poikkeus, A. M., et al. (2007). Very early phonological and language skills: Estimating individual risk of reading disability. Journal of Child Psychology and Psychiatry, 48, 23–931. Ramus, F., Rosen, S., Dakin, S. C., Day, B. L., Castellote, J. M., White, S., et al. (2003). Theories of developmental dyslexia: insights from a multiple case study of dyslexic adults. Brain, 126, 841–865. Ramus, F., & Szenkovitz, G. (2008). What phonological deficit? Quarterly Journal of Experimental Psychology, 61, 129–141. Raven, J. C. (1938). Progressive matrices: A perceptual test of intelligence. London: UK. Lewis (Pub). Richlan, F., Kronbichler, M., & Wimmer, H. (2009). Functional abnormalities in the dyslexic brain: A quantitative meta-analysis of neuroimaging studies. Human Brain Mapping, 30, 3299–3308. Rumsey, J. M., Horwitz, B., Donohue, B. C., Nace, K. L., Maisog, J. M., & Andreason, P. (1999). A functional lesion in developmental dyslexia: Left angular gyral blood flow predicts severity. Brain and Language, 70, 187–204. Salvi, R. J., Lockwood, A. H., Frisina, R. D., Coad, M. L., Wack, D. S., & Frisina, D. R. (2002). PET imaging of the normal human auditory system: Responses to speech in quiet and in background noise. Hearing Research, 170, 96–106. Sarkari, S., Simos, P. G., Fletcher, J. M., Castillo, E. M., Breier, J. I., & Papanicolaou, A. C. (2002). Contributions of magnetic source imaging to the understanding of dyslexia. Seminars in Pediatric Neurology, 9, 229–238. Schadwinkel, S., & Gutschalk, A. (2010). Activity associated with stream segregation in human auditory cortex is similar for spatial and pitch cues. Cerebral Cortex, 20, 2863–2873. Scott, S. K., & McGettigan, C. (2013). The neural processing of masked speech. Hearing Research, 303, 58–66. Scott, S. K., Rosen, S., Lang, H., & Wise, R. J. (2006). Neural correlates of intelligibility in speech investigated with noise vocoded speech—A positron emission tomography study. Journal of the Acoustical Society of America, 120, 1075–1083. Scott, S. K., Rosen, S., Wickham, L., & Wise, R. J. (2004). A positron emission tomography study of the neural basis of informational and energetic masking effects in speech perception. Journal of the Acoustical Society of America, 115, 813–821. Scott, S. K., Rosen, S., Beaman, C. P., Davis, J. P., & Wise, R. J. (2009). The neural processing of masked speech: Evidence for different mechanisms in the left and right temporal lobes. Journal of the Acoustical Society of America, 125, 1737–1743. Shahin, A. J., Bishop, C. W., & Miller, L. M. (2009). Neural mechanisms for illusory filling-in of degraded speech. NeuroImage, 44, 1133–1143.

Shaywitz, B. A., Shaywitz, S. E., Pugh, K. R., Mencl, W. E., Fulbright, R. K., Skudlarski, P., et al. (2002). Disruption of posterior brain systems for reading in children with developmental dyslexia. Biological Psychiatry, 52, 101–110. Shaywitz, S. E., & Shaywitz, B. A. (2005). Dyslexia (specific reading disability). Biological Psychiatry, 57, 1301–1309. Slotnick, S. D., Moo, L. R., Segal, J. B., & Hart, J, Jr (2003). Distinct prefrontal cortex activity associated with item memory and source memory for visual shapes. Brain Research. Cognitive Brain Research, 17, 75–82. Snowling, M. J. (2000). Dyslexia. Oxford: Blackwell Publishers. Spierer, L., Bellmann-Thiran, A., Maeder, P., Murray, M. M., & Clarke, S. (2009). Hemispheric competence for auditory spatial representation. Brain, 132, 1953–1966. Studebaker, G. A. (1985). A “rationalized” arcsine transform. Journal of Speech, Language and Hearing Research, 28, 455–462. Sprenger-Charolles, L., Colé, P., Lacert, P., & Serniclaes, W. (2000). On subtypes of developmental dyslexia: Evidence from processing time and accuracy scores. Canadian Journal of Experimental Psychology, 54, 87–104. Tardif, E., Murray, M. M., Meylan, R., Spierer, L., & Clarke, S. (2006). The spatiotemporal brain dynamics of processing and integrating sound localization cues in humans. Brain Research, 1092, 161–176. Torppa, M., Lyytinen, P., Erskine, J., Eklund, K., & Lyytinen, H. (2010). Language development, literacy skills, and predictive connections to reading in finish children with and without familial risk for dyslexia. Journal of Learning Disabilities, 43, 308–321. Trecy, M. P., Steve, M., & Martine, P. (2013). Impaired short-term memory for order in adults with dyslexia. Research in Developmental Disabilities, 34, 2211–2223. Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N., et al. (2002). Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage, 15, 273–289. Warrier, C. M., Johnson, K. L., Hayes, E. A., Nicol, T., & Kraus, N. (2004). Learning impaired children exhibit timing deficits and training-related improvements in auditory cortical responses to speech in noise. Experimental Brain Research, 157, 431–441. Wible, B., Nicol, T., & Kraus, N. (2002). Abnormal neural encoding of repeated speech stimuli in noise in children with learning problems. Clinical Neurophysiology, 113, 485–494. Wong, P. C., Jin, J. X., Gunasekera, G. M., Abel, R., Lee, E. R., & Dhar, S. (2009). Aging and cortical mechanisms of speech perception in noise. Neuropsychologia, 47, 693–703. Wong, P. C., Uppunda, A. K., Parrish, T. B., & Dhar, S. (2008). Cortical mechanisms of speech perception in noise. Journal of Speech, Language and Hearing Research, 51, 1026–1041. Zatorre, R. J., & Belin, P. (2001). Spectral and temporal processing in human auditory cortex. Cerebral Cortex, 11, 946–953. Zatorre, R. J., Bouffard, M., Ahad, P., & Belin, P. (2002). Where is ‘where’ in the human auditory cortex? Nature Neuroscience, 5, 905–909. Zatorre, R. J., Bouffard, M., & Belin, P. (2004). Sensitivity to auditory object features in human temporal neocortex. Journal of Neuroscience, 24, 3637–3642. Zekveld, A. A., Heslenfeld, D. J., Festen, J. M., & Schoonhoven, R. (2006). Top-down and bottom-up processes in speech comprehension. NeuroImage, 32, 1826–1836. Ziegler, J. C., Pech-Georgel, C., George, F., & Lorenzi, C. (2009). Speech-perceptionin-noise deficits in dyslexia. Developmental Science, 12, 732–745.

Please cite this article as: Dole, M., et al. Functional correlates of the speech-in-noise perception impairment in dyslexia: An MRI study. Neuropsychologia (2014), http://dx.doi.org/10.1016/j.neuropsychologia.2014.05.016i

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87