Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent component analysis

G Model

ARTICLE IN PRESS

NSL 30486 1–6

Neuroscience Letters xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet

Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis

1

2

3

4 5

Q1

Donna L. Murdaugh, Kavita D. Nadendla, Rajesh K. Kana ∗ Department of Psychology, University of Alabama at Birmingham, Birmingham, AL, USA

6

7 8 9 10 11 12 13

h i g h l i g h t s • • • • •

This fMRI study examined Theory of Mind (ToM) and brain in adults with autism. Stimuli required participants to make intentional and physical causal attribution. We used independent component analysis to examine brain responses. We found reduced brain response in autism in right temporoparietal junction. TPJ response to ToM was more robust than medial prefrontal cortex response.

14

15 28

a r t i c l e

i n f o

a b s t r a c t

16 17 18 19 20

Article history: Received 18 February 2014 Received in revised form 18 March 2014 Accepted 21 March 2014

21

27

Keywords: Autism Independent component analysis fMRI Functional connectivity Theory of Mind

29

1. Introduction

22 23 24 25 26

30 31 32 33 34 35

Neuroimaging studies have consistently identified a network of brain regions responsible for making inferences of others’ mental states. This network includes the medial prefrontal cortex (MPFC), posterior superior temporal sulcus (pSTS) at the temporoparietal junction (TPJ), and temporal poles. Although TPJ and MPFC are key nodes of the Theory of Mind (ToM) network, their relative functional roles are still debated. This study sought to examine the contribution of these regions in causal attribution and to explore the nature of the ToM network in people with autism spectrum disorders (ASD). Participants watched a series of comic strip vignettes in the MRI scanner, and identified the most logical ending to each vignette, which sometimes required intentional causal attribution. Independent component analysis was done to isolate temporally correlated brain networks. The functional networks for intentional causality included the TPJ and MPFC, with an increased contribution of TPJ. There was also a significant group difference in the TPJ, with reduced response in participants with ASD. These results suggest an increased role of TPJ in intentional causality. In addition, the reduced response in ASD in TPJ may reflect their difficulties in social cognition. © 2014 Published by Elsevier Ireland Ltd.

The mindblindness account attributes deficits in Theory of Mind (ToM) as key to the social and communication impairments in ASD [1]. Having a ToM involves recognizing others as intentional agents with beliefs or goals. Previous neuroimaging studies have consistently identified a network of brain regions responsible for inferring others’ mental states, including the medial prefrontal

∗ Corresponding author at: Department of Psychology, University of Alabama at Birmingham, CIRC 235G, 1719 6th Avenue South, Birmingham, AL 35294-0021, USA. Tel.: +1 205 934 3171; fax: +1 205 975 6330. E-mail addresses: [email protected], [email protected] (R.K. Kana).

cortex (MPFC), posterior superior temporal sulcus (pSTS) at the temporoparietal junction (TPJ), and temporal poles [2–4]. Although these regions have shown consistent activation in ToM tasks, their specific role in mentalizing is a topic of debate. One argument is that the MPFC is directly involved in inferring mental states, while the TPJ supports it by gathering important cues of intentionality and causality [5]. Another proposal suggests the TPJ has dual roles of attributing mental states to others as well as integrating those attributions to explain and predict behavior [3,6]. These findings imply the sensitivity of TPJ to intentionality. However, TPJ functional specialization does not reach full development until late childhood or early adolescence [7], which may explain why some individuals with ASD display delayed development of basic mentalizing ability [8]. Specific to ASD, previous studies have

http://dx.doi.org/10.1016/j.neulet.2014.03.051 0304-3940/© 2014 Published by Elsevier Ireland Ltd.

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

36 37 38 39 40 41 42 43 44 45 46 47 48 49

G Model NSL 30486 1–6

D.L. Murdaugh et al. / Neuroscience Letters xxx (2014) xxx–xxx

2 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 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

ARTICLE IN PRESS

reported diminished activation in the MPFC and TPJ, as well as weaker connectivity between them [9–11]. Some of the discrepancy regarding the specific roles of the nodes of the ToM network may be due in part to the inconsistency of the type of tasks used, as well as, to the specific regions of interest (ROI) examined in different studies. With new neuroimaging multivariate methods, such as the application of independent components analysis (ICA), we now can identify functionally relevant neural networks elicited by cognitive tasks. The benefit of this approach is that ICA has the advantage of being data-driven without the need of a seed voxel or temporal filtering [12]. As such, ICA uses algorithmic constraints so that each voxel in a component that has the same time course can be considered a functionally connected network without being limited to a priori ROIs. Since ICA methods do not impose prior constraints on the shape of the HRF, it might detect responses that would not have been revealed by a General Linear Model (GLM) analysis. In addition, ICA can be correlated to the time-course of the fMRI task in order to determine which functional networks are being elicited by the specific task. Indeed, studies have used ICA successfully in identifying functional networks in healthy individuals during performance of a task [12–15]. However, to our knowledge, this is the first study to assess task related brain responses using ICA in ASD. Previous neuroimaging studies in ASD using ICA have thus far focused solely on the resting state network (e.g., [16,17]. While there is considerable overlap between the resting state network and the ToM network, further deduction of the differential roles of individual nodes of the ToM network can be conducted using ICA with ToM task-related data. In the present study, we used group ICA in order to differentiate key functional networks associated with ToM in a causal inference task. Specifically, we presented a pictorial causal inference task known to elicit robust activation of the ToM network [2,18,19]. The type of task chosen is critical, as it is unclear whether all previous ToM tasks used in neuroimaging studies were optimal in eliciting activity from all regions of the ToM network (see [11] for commentary on ToM task selection). Traditional ToM tasks, such as verbal stories or narratives are language-oriented and the performance of individuals with ASD might be confounded by linguistic constraints. The current study avoids this bias by using non-verbal social situations. This task depicted scenarios that required either a physical causal attribution or an intentional causal attribution to build a logical ending to the scene. We hypothesized that the ASD group will differ in their brain responses, from control participants only in intentional causal attribution. Based on the altered connectivity accounts of ASD [10,20], we expect individuals with ASD to show weaker connectivity between brain regions in the independent functional networks underlying intentional causal attribution, specific to the TPJ. We also predicted that the TPJ would play a decisive role in mentalizing, with the MPFC taking on the role of more auxiliary functions involved in ToM processes such as response selection and inhibitory control [21]. And lastly, based on the functional underconnectivity hypothesis of autism, we expect that the top component for ASD will show weaker connectivity between brain regions when compared with age-matched controls.

the 15 participants with ASD had received a diagnosis of Asperger’s Disorder. The ASD and control participants did not significantly differ in age (21.4 ± 3.9, ASD and 22.6 ± 4.2, Control: t(34) = −0.728, p = 0.473). The mean Wechsler Abbreviated Scale of Intelligence (WASI) full scale intelligence quotients for the two groups were not significantly different (105.2 ± 17.7, ASD and 113.3 ± 8.4, Control: t(34) = −1.336, p = 0.194). Participants were excluded on the basis of metal implanted in their bodies, history of kidney disease, seizure disorder, diabetes, hypertension, anemia, or sickle cell disease. The control participants were screened through a parent-report (for participants younger than 18 years) or self-report (for participants older than 18 years) history questionnaire that asked if the participant had ever been diagnosed with Autism, Asperger’s Disorder, PDD-NOS, Attention Deficit Hyperactivity Disorder, a Learning Disability, Mental Retardation, Cerebral Palsy, or Tourette’s/Tic Disorder. All participants or their legal guardians gave written informed consent, approved by the UAB Institutional Review Board, to participate in the study and were compensated for their participation. 2.2. Experimental stimuli The stimuli consisted of a series of black and white comic strip vignettes (adapted from [18]) depicting scenarios that demand either a physical causal attribution or an intentional causal attribution. The first part of the vignette was presented for 5 s and the participants’ task was to choose a logical ending to the story from the three choices in the second panel presented for 6 s. The whole vignette remains on the screen for a total of 11 s. Participants were to indicate the answer by a button press. Participants viewed a total of 11 physical cartoons, and 11 character (intentional) cartoons (Supplemental Fig. 1) presented in an event-related design. 2.3. Data acquisition Functional MRI data were collected on a Siemens 3.0 Tesla Allegra head-only scanner. A single-shot gradient-recalled echo-planar pulse sequence was used for rapid image acquisition (TR = 1000 ms, TE = 30 ms, flip angle = 60◦ ). Seventeen adjacent oblique-axial slices were acquired in an interleaved sequence with 5 mm slice thickness, 1 mm gap, a 24 cm × 24 cm field of view, and a 64 × 64 matrix, resulting in an in-plane resolution of 3.75 mm × 3.75 mm × 5 mm. 2.4. fMRI data analyses: preprocessing fMRI data were pre-processed and statistically analyzed using SPM8 (Wellcome Department of Cognitive Neurology, London, UK). Images were motion-corrected using INRIalign, an algorithm unbiased by local signal changes [24]. After motion correction, a mean functional image was computed for each separate study and then matched to the EPI template provided within SPM8. Data were then spatially normalized to standard Montreal Neurological Institute (MNI) brain space and spatially smoothed using a threedimensional Gaussian kernel of 8 mm full-width at half-maximum (FWHM).

104

2. Materials and methods

2.5. fMRI data analyses: independent component analysis

105

2.1. Participants

After the data were preprocessed in SPM8, all 36 participants were included in a group independent component analysis (ICA) [13] using the fMRI Group ICA Toolbox (GIFT; http://icatb.sourceforge.net/, version 1.3e). A total of 34 independent components were estimated using dimensionality estimation performed using the minimum description length criteria, modified to account for spatial correlation [25]. The GIFT toolbox organizes the data into batch scripts, with the first script compressing the

106 107 108 109 110 111

Fifteen high-functioning young adults with ASD (all male, one left-handed) and twenty-one typical control participants (all male, right-handed) were included in this study. The participants with ASD had received a diagnosis based on the Autism Diagnostic Interview-Revised (ADI-R) [22] symptoms, Autism Diagnostic Observation Schedule (ADOS) [23], and clinical impressions. Six of

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

131

132 133 134 135 136 137 138 139 140 141

142

143 144 145 146 147 148 149

150

151 152 153 154 155 156 157 158 159 160

161

162 163 164 165 166 167 168 169

G Model NSL 30486 1–6

ARTICLE IN PRESS D.L. Murdaugh et al. / Neuroscience Letters xxx (2014) xxx–xxx

3

Table 1 Brain regions identified in the top three components from the intentional causality condition. The table depicts anatomical location, MNI coordinates for peak activation Q2 voxel in each brain region, and t scores from random effects analyses across all participants (n = 36; p < 0.001). Brain region Intentional Component A Superior temporal gyrus Middle/superior temporal gyrus Medial prefrontal cortex Middle frontal gyrus Middle occipital Anterior cingulate cortex Inferior frontal cortex Component B Posterior cingulate cortex/fusiform gyrus Middle temporal gyrus Postcentral gyrus Component C Superior temporal gyrus/insula Superior temporal gyrus Postcentral gyrus Middle cingulate gyrus Superior parietal lobule Posterior cingulate gyrus Precuneus Culmen Inferior parietal lobule Cuneus

170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199

200

201 202 203 204

BAa

Hemb

MNI

t

39 19 10

R L R R L R R L

52, −62, 18 −50, −64, 12 4, 58, 20 46, 6, 44 −34, 0, 54 8, −34, −4 4, 6, 50 −46, 22, 24

19.53 16.91 10.14 9.61 5.34 9.23 7.86 5.12

L R R

−4, −54, 6 48, −68, 24 52, −18, 60

36.34 9.59 7.48

L R L R R R L L R L R

−50, −26, 12 54, −20, 14 −32, −38, 70 0, 0, 46 34, −70, 48 4, −34, 30 −12, −66, 12 −24, −74, 50 12, −36, −10 −52, −36, 58 2, −96, 2

23.78 21.08 5.03 11.10 8.48 5.78 5.58 5.33 5.22 5.16 4.91

32

30 3 41 1

23 7 40 18

fMRI time series data from all participants in a principal component analysis (PCA). There were two PCA data reduction stages, each with a dimensionality of 25, which helped to reduce the impact of noise as well as to make the estimation computationally tractable [13,15]. The data reduction was followed by a group spatial ICA, performed on the participants’ aggregate data (ICs) [13]. Each component was visually inspected for artifacts, which finally resulted in 18 components. Visual inspection of discarded components suggested that they represented eye movements, head motion, or cardiac-induced pulsatile artifact at the base of the brain. This resulted in a final number of 18 components that were selected for further analysis to identify the components that best correlated with our task. Next, a temporal correlation analysis was performed between the 18 surviving components and the task timecourse for the physical trials and intentional trials separately. The most relevant components were then selected based on both the signal timecourse closely matching the task paradigm and the probability rank of the component being statistically significant. As such, the top 3 components from both the physical trials and intentional trials were selected; with all final components having correlation coefficients greater than R2 = 0.06. For each participant, the top three correlated components for both the physical and intentional causality timecourse were converted to Fisher’s z values. Individual maps of all subjects were entered into random effect one sample t-tests in SPM8 with a threshold of p < 0.001, corrected at the cluster level for false discovery rate (FDR) at p < 0.01, to create a sample-specific component map. These maps were used as a mask for group analyses within the corresponding component. Thus, the results represent both groups of participants, and are not biased by component maps defined from control participants only.

2.6. fMRI data analyses: statistical analyses The GIFT maps of individual components were entered into SPM8 for group analyses. The Fisher’s z values of these individual correlation maps represent the fit of a specific voxel BOLD timecourse to the group averaged component’s timecourse. Thus, group

analyses test the connectivity strength (i.e., signal synchronization) of each voxel to the whole spatial component. Random effects two-sample t-tests of the top three components for each condition examined group differences between ASD and control subjects. The resulting statistical maps were then masked with the general map of the relevant component for all 36 participants to explore results within each network separately. The between group results were reported at a statistical threshold p < 0.01, cluster corrected at 170 contiguous voxels, with all coordinates in MNI space. Due to the specific focus of our hypotheses on the intentional causality components, the results for the physical causality components are presented in the supplementary material.

205 206 207 208 209 210 211 212 213 214 215 216

3. Results

217

3.1. Independent neural components correlated with fMRI task

218

The three components with the highest probability rank for physical causality entailed a number of regions involved in different aspects necessary for interpreting the task, including areas involved in visual and visual association, motor and motor imagery [26], somatosensory, and memory and learning, such as, left middle occipital gyrus, left superior frontal gyrus, right middle frontal gyrus, left superior parietal lobule, right fusiform gyrus, hippocampus, supplemental motor area, and rolandic operculum (Supplemental Table 1). For intentional causality, the three components with the highest probability rank revealed activation of: (A) ToM network, including bilateral pSTS at the TPJ, right MPFC, and right ACC; (B) regions shown to be involved in self-other evaluation and emotion perception [4,27], including bilateral posterior cingulate cortex (PCC) and right middle temporal gyrus; and (C) regions encompassing functions related to social perception and mental state attribution, specifically, bilateral superior temporal gyrus (including pSTS), right PCC, and left precuneus (Fig. 1 and Table 1). For behavioral results and more detailed description of brain regions found in

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237

G Model NSL 30486 1–6 4

ARTICLE IN PRESS D.L. Murdaugh et al. / Neuroscience Letters xxx (2014) xxx–xxx

Fig. 1. Independent components representing the three components with the highest temporal correlation with the intentional causality task condition. Components A, B, and C correspond to the regions listed in Table 1 respectively. Maps are thresholded at p < 0.05, FDR corrected.

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

G Model NSL 30486 1–6

ARTICLE IN PRESS D.L. Murdaugh et al. / Neuroscience Letters xxx (2014) xxx–xxx

Fig. 2. Scatterplot showing average cluster z-scores for each subject for the TPJ and MPFC in Component A. A paired t-test revealed significantly stronger withinnetwork connectivity in the TPJ compared to MPFC during intentional causal attribution (t(35); p < 0.0001).

238 239

240 241

the three components for intentional causality see supplemental material. 3.2. Between group differences of independent component networks underlying intentional causality

251

Component A showed significantly reduced connectivity in the autism group compared to the control group in the pSTS at the TPJ (x = 58, y = −52, z = 6; p < 0.01, cluster corrected at 170 voxels). In addition, component B showed reduced connectivity in the autism group at the bilateral PCC (x = −6, y = −50, z = 4; x = 6, y = −42, z = 6; p < 0.01), and the precuneus (x = 12, y = −76, z = 24; p < 0.01). There were no group differences in the component C. See supplemental material for intentional component A figure of group differences (Supplemental Fig. 2) and group differences in the physical causality condition.

252

3.3. Differential functional connectivity of the TPJ and MPFC

242 243 244 245 246 247 248 249 250

263

Since the component with the highest probability rank for the intentional causality condition (component A) encompassed both the TPJ and the MPFC, we further assessed the relationship between each region and its correlation with the overall component, to determine differences in the magnitude of functional connectivity between either the TPJ (x = 52, y = −62, z = 18) or MPFC (x = 4, y = 58, z = 20) and the rest of the intentional component A network. As such, a paired t-test comparing average cluster z-scores of each participant (n = 36) revealed significantly stronger within-network connectivity in the TPJ compared to MPFC during intentional causal attribution (t(35); p < 0.0001; Fig. 2).

264

4. Discussion

253 254 255 256 257 258 259 260 261 262

265 266 267 268 269 270 271 272

In this study, brain regions that exhibit synchronous BOLD timecourses were classified into independent components by group ICA [15]. The three functional networks correlated with intentional causality trials elicited activation of three networks involved in different aspects of social cognition. Component A was the network that represented the highest correlation with our task time-course, and revealed a robust network consistent with previously identified ToM regions, including both the TPJ and the

5

MPFC [6,10,11,28,29]. Further investigation into the individual contribution of both the TPJ and MPFC within the ToM network revealed significantly greater functional connectivity of the TPJ than the MPFC. In addition, when we compared the ToM network between the individuals with autism and our control participants, we found decreased connectivity within the network in the individuals with autism specific to the TPJ. Taken together, these results demonstrate a distinct difference in the TPJ, but not the MPFC, in individuals with ASD. These differences likely reflect distinct roles of both the TPJ and the MPFC in mentalizing. As such, the TPJ may be considered critical in interpreting causal inference, and the dysfunction of it may underlie ToM-related social difficulties seen in ASD. Our study used a novel approach to differentiate between different networks involved in ToM, allowing for an unbiased selection of synchronous brain regions involved in causal inference. Our results are consistent with other neuroimaging studies that have identified the TPJ, MPFC, STS, and cingulate cortex as part of a distinct network involved in ToM [2–4]. Interestingly, the present results indicate that in healthy individuals, both the TPJ and MPFC are part of the same functional network. This result is consistent with previous studies which found robust MPFC and TPJ activation in response to similar verbal and non-verbal ToM tasks [2,18,30]. The current results were also consistent with some studies in terms of laterality [2,18], in that the TPJ activation in the present study was bilateral, whereas MPFC activation was rightlateralized. Gweon et al. [31] assessed ToM processes in both children and adults, and found that the right TPJ was involved in making inferences about others’ mental states regardless of age. However, the left TPJ activation was more robust in adults than in children, suggesting a potential sophistication of the development of ToM as a person ages. The left TPJ has been identified as being involved in metarepresentation of social information [32]. Unlike prior findings of decreased activation of the MPFC in individuals with autism [9,10,33], our results found no differences between autism and control groups in the MPFC. Instead, these results were most consistent with Lombardo et al. [11] findings, which also reported no differences between groups in the MPFC. Lombardo et al. [11] suggests that as an individual develops, the TPJ increases in activation, becoming more specialized in mentalizing processes, while the MPFC decreases in specialization of mentalizing, becoming more involved in general metarepresentation of social and non-social information. Our previous findings [19] found the TPJ and IFG to be primarily involved in intentional causality, but not the MPFC. The results of the current study suggest some benefit for the ICA method for detecting additional regions not seen in GLM based studies. Further investigation into the differential roles of the TPJ and MPFC in the ToM network revealed that the TPJ had significantly greater synchronization within the functional network than the MPFC. This suggests that the TPJ may be more specialized for causal inference, while the MPFC may be involved in more general aspects involved in ToM, such as attention, detection of incongruences, response selection, and inhibition [7,11,21]. While these executive processes are important for ToM, they are not domain specific, and this may be the reason the TPJ is more robustly synchronized with our ToM network than the MPFC. In summary, our results suggest that the deficits individuals with autism experience in social interactions mediated by ToM, such as causal inference, are likely related to altered functioning of the TPJ. While the MPFC is important in ToM processes it is not domain specific and is involved in social and nonsocial metarepresentations. Our results indicate a need for further investigation of the specific roles of the MPFC and the TPJ in mentalizing.

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337

G Model NSL 30486 1–6

D.L. Murdaugh et al. / Neuroscience Letters xxx (2014) xxx–xxx

6 338

339 340 341

342

343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384

ARTICLE IN PRESS

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.neulet.2014.03.051. References [1] U. Frith, Mind blindness and the brain in autism, Neuron 32 (2001) 969–979. [2] H.L. Gallagher, F. Happé, N. Brunswick, P.C. Fletcher, U. Frith, C.D. Frith, Reading the mind in cartoons and stories: an fMRI study of ‘theory of mind’ in verbal and nonverbal tasks, Neuropsychologia 38 (2000) 11–21. [3] R. Saxe, A. Wexler, Making sense of another mind: the role of the right temporoparietal junction, Neuropsychologia 43 (2005) 1391–1399. [4] A. Di Martino, K. Ross, L.Q. Uddin, A.B. Sklar, F.X. Castellanos, M.P. Milham, Functional brain correlates of social and nonsocial processes in autism spectrum disorders: an activation likelihood estimation meta-analysis, Biol. Psychiatry 65 (2009) 63–74. [5] H.L. Gallagher, C.D. Frith, Functional imaging of ‘theory of mind’, Trends Cogn. Sci. 7 (2003) 77–83. [6] R. Saxe, N. Kanwisher, People thinking about thinking people – the role of the temporo-parietal junction in theory of mind, NeuroImage 19 (2003) 1835–1842. [7] R. Saxe, S. Whitfield-Gabrieli, J. Scholz, K.A. Pelphrey, Brain regions for perceiving and reasoning about other people in school-aged children, Child Dev. 80 (2009) 1197–1209. [8] F.G.E. Happé, The role of age and verbal ability in the theory of mind task performance of subjects with autism, Child Dev. 66 (1995) 843–855. [9] F. Castelli, C. Frith, F. Happé, U. Frith, Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes, Brain 125 (2002) 1839–1849. [10] R.K. Kana, T.A. Keller, V.L. Cherkassky, N.J. Minshew, M.A. Just, Atypical frontalposterior synchronization of theory of mind regions in autism during mental state attribution, Soc. Neurosci. 4 (2009) 135–152. [11] M.V. Lombardo, B. Chakrabarti, E.T. Bullmore, MRC AIMS Consortium, S. BaronCohen, Specialization of right temporo-parietal junction for mentalizing and its relation to social impairments in autism, NeuroImage 56 (2011) 1832–1838. [12] M.J. McKeown, S. Makeig, G.G. Brown, T.P. Jung, S.S. Kindermann, A.J. Bell, T.J. Sejnowski, Analysis of fMRI data by blind separation into independent spatial components, Hum. Brain Mapp. 6 (1998) 160–188. [13] V.D. Calhoun, T. Adali, G.D. Pearlson, J.J. Pekar, Spatial and temporal independent component analysis of functional MRI data containing a pair of task-related waveforms, Hum. Brain Mapp. 13 (2001) 43–53. [14] V.D. Calhoun, J.J. Pekar, V.B. McGinty, T. Adali, T.D. Watson, G.D. Pearlson, Different activation dynamics in multiple neural systems during simulated driving, Hum. Brain Mapp. 16 (2002) 158–167. [15] V.D. Calhoun, K.A. Kiehl, G.D. Pearlson, Modulation of temporally coherent brain networks estimated using ICA at rest and during cognitive tasks, Hum. Brain Mapp. 29 (2008) 828–838. [16] M. Assaf, K. Jagannathan, V.D. Calhoun, L. Miller, M.C. Stevens, R. Sahl, J.G. O’Boyle, R.T. Schultz, G.D. Pearlson, Abnormal functional connectivity of default

[17]

[18] [19]

[20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

mode sub-networks in autism spectrum disorder patients, NeuroImage 53 (2010) 247–256. E. Hagan, R. Stoyanova, S. Baron-Cohen, A. Calder, Reduced functional connectivity within and between ‘social’ resting state networks in autism spectrum conditions, SCAN 8 (2012) 694–701. E. Brunet, Y. Sarfati, M.C. Hardy-Baylé, J. Decety, A PET investigation of the attribution of intentions with a nonverbal task, NeuroImage 11 (2000) 157–166. R.K. Kana, L.E. Libero, C.P. Hu, H.D. Deshpande, J.S. Colburn, Functional brain networks and white matter underlying theory-of-mind in autism, SCAN (2012) (Epub ahead of print). M.A. Just, V.L. Cherkassky, T.A. Keller, N.J. Minshew, Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity, Brain 127 (2004) 1811–1821. R. Saxe, L.E. Schulz, Y.V. Jiang, Reading minds versus following rules: dissociating theory of mind and executive control in the brain, Soc. Neurosci. 1 (2006) 284–298. C. Lord, M. Rutter, A. Le Couteur, Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders, J. Autism Dev. Disord. 24 (1994) 659–685. C. Lord, S. Risi, L. Lambrecht, E.H. Cook, B.L. Leventhal, P.C. DiLavore, A. Pickles, M. Rutter, The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism, J. Autism Dev. Disord. 30 (2000) 205–223. L. Freire, J.F. Mangin, Motion correction algorithms may create spurious brain activations in the absence of subject motion, NeuroImage 14 (2001) 709–722. Y.O. Li, T. Adali, V.D. Calhoun, Estimating the number of independent components for functional magnetic resonance imaging data, Hum. Brain Mapp. 28 (2007) 1251–1266. A.J. Szameitat, A. McNamera, S. Shen, A. Sterr, Neural activation and functional connectivity during motor imagery of bimanual everyday actions, PLOS ONE 7 (2012) e38506. D. Sabatinelli, E.E. Fortune, Q. Li, A. Siddiqui, C. Krafft, W.T. Oliver, S. Beck, J. Jeffries, Emotional perception: meta-analyses of face and natural scene processing, NeuroImage 54 (2011) 2524–2533. A. Ciaramidaro, M. Adenzato, I. Enrici, S. Erk, L. Pia, B.G. Bara, H. Walter, The intentional network: how the brain reads varieties of intentions, Neuropsychologia 45 (2007) 3105–3113. L. Young, D. Dodell-Feder, R. Saxe, What gets the attention of the temporoparietal junction? An fMRI investigation of attention and theory of mind, Neuropsychologia 48 (2010) 2658–2664. P.C. Fletcher, F. Happe, U. Frith, S.C. Baker, R.J. Dolan, R.S.J. Frackowiak, C.D. Frith, Other minds in the brain: a functional imaging study of ‘theory of mind’ in story comprehension, Cognition 57 (1995) 109–128. H. Gweon, D. Dodell-Feder, M. Bedny, R. Saxe, Theory of mind performance in children correlates with functional specialization of a brain region for thinking about thoughts, Child Dev. 83 (2012) 1853–1868. I. Apperly, D. Samson, C. Chiavarino, W. Bickerton, G. Humphreys, Testing the domain-specificity of a theory of mind deficit in brain-injured patients: evidence for consistent performance on non-verbal, Cognition 103 (2007) 300–321. F. Happé, S. Ehlers, P. Fletcher, U. Frith, M. Johansson, C. Gillberg, R. Dolan, R. Frackowiak, C. Frith, ‘Theory of mind’ in the brain. Evidence from a PET scan study of Asperger syndrome, NeuroReport 8 (1996) 197–201.

Please cite this article in press as: D.L. Murdaugh, et al., Differential role of temporoparietal junction and medial prefrontal cortex in causal inference in autism: An independent components analysis, Neurosci. Lett. (2014), http://dx.doi.org/10.1016/j.neulet.2014.03.051

385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437