Background: Autism spectrum disorder (ASD) is a pervasive neurodevelopmental disorder characterised by diminished social reciprocity and communication skills and the presence of stereotyped and restricted behaviours. Executive functioning deficits, such as working memory, are associated with core ASD symptoms. Working memory allows for temporary storage and manipulation of information and relies heavily on frontal-parietal networks of the brain. There are few reports on the neural correlates of working memory in youth with ASD. The current study identified the neural systems underlying verbal working memory capacity in youth with and without ASD using functional magnetic resonance imaging (fMRI). Methods: Fifty-seven youth, 27 with ASD and 30 sex- and age-matched typically developing (TD) controls (9– 16 years), completed a one-back letter matching task (LMT) with four levels of difficulty (i.e. cognitive load) while fMRI data were recorded. Linear trend analyses were conducted to examine brain regions that were recruited as a function of increasing cognitive load. Results: We found similar behavioural performance on the LMT in terms of reaction times, but in the two higher load conditions, the ASD youth had lower accuracy than the TD group. Neural patterns of activations differed significantly between TD and ASD groups. In TD youth, areas classically used for working memory, including the lateral and medial frontal, as well as superior parietal brain regions, increased in activation with increasing task difficulty, while areas related to the default mode network (DMN) showed decreasing activation (i.e., deactivation). The youth with ASD did not appear to use this opposing cognitive processing system; they showed little recruitment of frontal and parietal regions across the load but did show similar modulation of the DMN. Conclusions: In a working memory task, where the load was manipulated without changing executive demands, TD youth showed increasing recruitment with increasing load of the classic fronto-parietal brain areas and decreasing involvement in default mode regions. In contrast, although they modulated the default mode network, youth with ASD did not show the modulation of increasing brain activation with increasing load, suggesting that they may be unable to manage increasing verbal information. Impaired verbal working memory in ASD would interfere with the youths’ success academically and socially. Thus, determining the nature of atypical neural processing could help establish or monitor working memory interventions for ASD. Keywords: Autism spectrum disorder, Verbal working memory, Cognitive load, Executive functioning, fMRI * Correspondence: email@example.com V. M. Vogan and K. E. Francis contributed equally to this work. Diagnostic Imaging & Research Institute, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada Department of Psychology, University of Toronto, 100 St. George St., Toronto, Ontario M5S 3G3, Canada Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 2 of 12 Background temporal areas, whereas TD adults showed bilateral Autism spectrum disorder (ASD) is a neurodevelopmental dlPFC activation and less posterior activity. Following a disorder characterised by diminished social reciprocity and multi-pronged analysis approach, the authors concluded communication skills, as well as the presence of stereotyped that TD adults used verbal encoding strategies to and restricted behaviours . There is considerable evi- complete the task, whereas adults with ASD used non- dence that individuals with ASD also have impaired execu- verbal and visually oriented strategies with their WM tive and cognitive function [1–8]. The deficits in executive network shifted towards a right hemisphere dominance. processing may contribute to the autistic symptomology, as The ‘n-back’ protocol is commonly used to manipulate proposed by the ‘executive dysfunction theory’ of ASD [5, cognitive load while studying WM [9, 10, 27, 32, 38–45]. 6]. Prior literature on the neural underpinnings of ASD, as The typical n-back task involves viewing a series of stim- well as the cognitive difficulties that follow, suggests that uli, then indicating whether the current stimulus is the working memory (WM) impairments are associated with same as the one presented ‘n’ (1, 2, 3, etc.) trials before. functional abnormalities in the frontal lobe, especially pre- The difficulty level is indexed by the total number of frontal cortical activity [3, 7, 9–12]. The protracted frontal interfering items between repeating stimuli. By increas- lobe maturation means that the functions relying on the ing load in this manner, different mental strategies frontal lobes are particularly vulnerable to developmental required to complete the task are also employed, includ- disturbances [13, 14]. ing executive functioning and procedural strategies. Working memory is the ability to temporarily store Manipulating both WM and other cognitive functions and manipulate information [15, 16]. WM is seen as an across load makes WM-specific changes difficult to essential element of cognitive control [16–19], critical quantify and link to specific brain regions. In the present for learning and academic achievement , as well as study, we used a one-back letter matching task (LMT) social competency . Previous literature suggests that [46–48] that avoids these confounds. LMT holds execu- individuals with ASD have greater difficulty with tive function constant across difficulty levels, while sys- visuo-spatial than verbal WM, which is more often com- tematically manipulating memory load, which better parable to typically developing (TD) individuals [22–24]. isolates the effects of cognitive load on verbal WM. A Prior work also reports, however, that WM in ASD is in- developmental investigation of LMT in typically devel- tact for simple memory tasks [22–26] including simple oping children and adults showed an opposing cognitive verbal WM , but impaired on more complex tasks processing system, with increasing cognitive load and in- [22, 23, 25, 26, 28] including verbal WM , compared creasing recruitment of brain areas related to WM, while to typically developing (TD) individuals, or broadly com- decreasing activation of areas in the default mode net- promised . A number of studies found that when work (DMN); adults showed larger load-dependent performing WM tasks of increasing complexity or cogni- changes than children in the bilateral superior parietal tive load, children with ASD were impaired compared to gyri, inferior/dorsolateral prefrontal and left middle TD children [8, 26, 29]. frontal gyri . The neuroimaging literature has identified a system of Limited neuroimaging studies exist to examine the im- lateral prefrontal, premotor and posterior parietal cortices pact of WM load on brain activity in ASD. In a recent underlying WM function [31, 32], with children showing investigation by Rahko et al. , adolescents with ASD more widespread activation patterns than adults . (ages 11–18 years) were observed to have reduced During a verbal WM two-back task, Nagel et al.  modulation of brain activity with increasing cognitive found that children (ages 10–16 years) recruited the left load in the insula, motor and auditory and somatosen- frontal and temporal lobes. Similarly, Thomason et al.  sory cortices compared to TD adolescents during a used a verbal WM block design task and observed that visuo-spatial n-back WM task. An earlier study by children (ages 7–12 years) showed activation in the left Vogan et al.  utilising a colour matching task (a frontal and parietal cortical regions, but activation in these visuo-spatial version of LMT) showed that children with regions was reduced compared to adults. ASD (ages 7–13 years) demonstrated reduced modula- Few studies have used neuroimaging to investigate ver- tion in the dlPFC, medial premotor cortex and precu- bal WM in ASD, with most studies using visual-spatial neus with increasing cognitive load. tasks (e.g., [10, 12, 36, 37]); this, our understanding of The current study used functional magnetic resonance the neural correlates underlying verbal WM deficits in imaging (fMRI) with a verbal WM task to explore neural ASD, particularly in children, remains modest. Koshino systems underlying WM, and the effects of cognitive load, et al.  used a letter matching task and found that, des- in children and young adolescents with and without ASD. pite comparable behavioural performance, adults with In this study, the cognitive load was manipulated by in- ASD showed right-lateralised activation in the dorsolat- creasing task difficulty level (see the “Methods” section for eral prefrontal cortex (dlPFC) and parietal and inferior full task description). We hypothesised that children with Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 3 of 12 ASD would perform with a lower accuracy than their Informed consent, MRI scanning, and the cognitive matched TD controls on the LMT with increasing cogni- and clinical testing involved in this study were carried tive load. Moreover, we expected that children with ASD out at the Hospital for Sick Children in Toronto. All the would under-recruit frontal and parietal cortical regions experimental procedures used were approved by the related to verbal WM, relative to TD children, and that hospital’s Research Ethics Board. All participants gave the difference would increase with greater cognitive de- informed verbal assent, and a parent or legal guardian of mand. We predicted that cortical activity would be linearly all participants gave informed written consent. modulated (increasing in WM areas, decreasing in DMN ASD clinical diagnosis was confirmed through expert areas) by task difficulty; however, we anticipated that the clinical judgement and the Autism Diagnostic Observa- youth with ASD would have a less pronounced pattern of tion Schedule (ADOS)  for all participants with linear activation/deactivation. ASD. The ADOS was conducted by a trained individual with established inter-rater research reliability. Methods Letter matching task Participants The LMT is a verbal WM task. LMT is presented visu- Ninety one participants (47 ASD, 44 TD) were recruited ally to participants and has linguistic/phonological fea- through community support centres, parent support tures. Participants attended to letters embedded in a groups, email listservs, hospital ads and schools for this global “A” figure. Participants were taught to focus only study. Six TD participants and 20 ASD participants were on the eight relevant letters (A, B, E, H, K, M, N, T) pre- excluded from analyses due to inadequate task perform- sented in uppercase and to ignore irrelevant letters “O” ance; see below (lines 197–202) for our threshold for and “P” (Fig. 1). The task was designed with both rele- task performance (ASD = 12, TD = 2), protocol comple- vant and irrelevant letters, as well as the irrelevant global tion (ASD = 5, TD = 2) and excessive movement (ASD letter, since tasks containing misleading or irrelevant fea- = 3, TD = 0), and two TDs were excluded for tures evoke interference and elicit cognitive control, age-matching. The age- and sex-matched sample was which has been shown to provide more reliable mea- composed of 27 children with ASD (5 girls and 22 boys) sures of WM capacity [47, 51]. The number ‘n’ of rele- and 30 TD children (8 girls and 22 boys) aged 9 to vant letters in the figure, referred to as capacity, 16 years old. Although groups differed slightly on IQ as increased by one item for each increasing difficulty level. determined by the Wechsler Abbreviated Scale of Difficulty level was assigned n + 2 to account for these Intelligence , t = 2.16, p = 0.04, both groups had cognitive control and executive functions. LMT is a (39) IQs within the average range, see Table 1 for additional one-back task in which participants were instructed to participant characteristics. identify relevant letter(s) and remember if the letter(s) in Participants were not included in the study with any the current stimulus figure matched those from the pre- significant psychiatric comorbidities , medical ill- vious figure, disregarding letter repetition and location. nesses, neurological disorders, prematurity, colour blind- Repetition of both irrelevant and relevant letters within ness, uncorrected vision, IQ < 80 or any standard MRI a stimulus was usual (see Fig. 1), and although the num- contraindicators, such as ferromagnetic implants. TD bers and placement of the letters changed, the partici- participants were also not included if they had a history pants always ignored the same two letters, O and P. of learning disability, developmental delay, a sibling with Stimuli were presented one at a time for 3 s, during ASD or attention deficit hyperactivity disorder (ADHD). which time children indicated their response using a These factors were not the current primary diagnosis for dual-key MRI compatible keypad in their right hand; any of the ASD subjects. one button for the same relevant letters embedded in Table 1 Demographic and neuropsychological test characteristics of the sample Variables ASD TD Significant test % Mean (SD) % Mean (SD) Demographic data Sex (% male) 81 73 χ = 0.54, p = 0.46 (1) Age 12.56 (1.46) 12.96 (1.89) t = 0.91, p = 0.37 (54) Full-scale IQ 105.52 (14.41) 112.27 (7.91) t = 2.16, p = 0.04* (39) ADOS total 11.89 (4.30) N/A *p < 0.05 ADOS scores range from 3 to 20 with greater symptom severity reflected by higher scores Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 4 of 12 Fig. 1 Protocol description of the letter matching task (LMT). a The task consisted of six difficulty levels where thenumberofrelevantletters (A, B,E,H,K, M, N,T) increased with each difficulty level. Difficulty level = the number of relevant numbers + 2. Participants were instructed to ignore the global ‘A’ figure, letter location, letter repetition and irrelevant letters (‘O’ and ‘P’). b The task used a block design with each run consisting of 32-s task blocks for each difficulty level, followed by 20 s baseline blocks with figures containing only ‘O’ and ‘P’ (irrelevant letters). The task blocks were shown in pseudo-random order within each run. c An example of part of a baseline block sequence during which participants were instructed not to respond. d Example of part of a task block sequence; participants indicated if the current figure ‘A’ contained the same or different letters as the previous figure. In the first exemplar (in 1B), the target letters are M and N; in the subsequent exemplar (in 1D), the target letters are M and N (thus the same), then N and K (thus different) and then N and K again (thus, the correct response is ‘same’), as each stimulus is judged by whether the stimuli are the same or different as the preceding one. Stimuli were presented for 3 s followed by a 1-s inter-stimulus fixation cross the stimulus as the previous stimulus and one button for responded correctly within 3 s of stimulus onset. The different. A 1-s inter-stimulus interval during which a fMRI session took approximately 22 min, during which fixation cross was presented followed the task stimuli. reaction time and accuracy were recorded as behavioural The baseline trials included presentations in the same data. configuration as the task stimuli, except that the stimuli Participants were excluded from the analyses if they only included the irrelevant letters (O and P) in varying did not complete at least three runs of the task, with an configurations. They were presented with the same tim- accuracy of at least 70% (averaged across their runs) on ing as the task, but for 20 s; thus, only five stimuli per the two easiest levels (D3 and D4). Participants were also block (see Fig. 1c). All children were trained and com- required to have at least two runs where at least 50% of pleted practice trials successfully with an accuracy of at the blocks were 70% accurate, to ensure that participants least 80% prior to performing the task in the scanner. were performing better than chance (50%). Motion was Twenty-four task and 24 baseline blocks (168 total considered acceptable if participants moved less than task trials) were displayed over four runs. Each run in- 1.5 mm from their average head position in a minimum cluded a 32-s block for each of the six difficulty levels; of 60% of the volume within a task block. each task block consisted of eight stimuli of the same difficulty level. The levels were randomised for each run, Image acquisition with the same order of runs presented to all participants. All images were acquired on a 3T Siemens Trio MRI The task blocks alternated with the 20-s baseline blocks, system with a 12-channel head coil. Foam padding was where participants were taught to look but not respond used to provide head motion restriction and stabilisa- to the figures. Items were only correct if subjects tion. fMRI scans were a single-shot echo planar imaging Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 5 of 12 sequence (axial; FOV = 192 × 192 × 150 mm; 3 × 3 × 5 mm were chosen as this was the approach used in prior stud- voxels; TR/TE/FA = 2000/30/70). The visual stimuli for the ies with the same type of working memory protocols task (LMT) were shown using MR-compatible goggles. [48, 55, 56]. Individual subjects’ results were averaged Stimuli were displayed, and performance was documented across runs, then between-group comparisons were con- using presentation software (Neurobehavioral Systems Inc., ducted using FMRIB’s Local Analysis of Mixed Effects-1 Berkeley, CA, USA). Structural scans were used as anatom- (FLAME-1) . Using FLAME-1 allowed us to acquire ical references, collected as a high-resolution T1-weighted between-subject variance estimation, thus increasing our 3D MP-RAGE image (sagittal; FOV = 2000/30/70 mm; capacity to identify real activation . Cluster-based 1 mm iso voxels; TR/TE/TI/FA = 2300/2.96/900/9). During thresholding was determined by Z > |2.3| as well as a the structural scan, participants used MR-compatible gog- corrected cluster significance threshold of p < 0.05 to corr gles and earphones to watch a movie of their choice. identify significant activations. Regions of interest (ROIs) were identified by examining local maxima of regions Behavioural data analyses showing significant variation between TD and ASD Both TD and ASD groups performed poorly on difficulty groups in the linear trend analyses, for visualisation only. levels 7 and 8 (D7 and D8), (TD: D7—M = 0.59, SD = 0.15; Spherical ROIs with 6-mm radii centred on the local D8—M = 0.58, SD = 0.13; ASD: D7—M =0.53, SD =0.11; maxima of cohort difference maps were created from D8—M = 0.49, SD = 0.14). D7 and D8 were therefore ex- which average percent signal change and standard error cluded from the analyses, and the first four difficulty levels scores were derived. The average peak cluster signal (D3 to D6) were analysed. Averages across runs for each change for both the TD and ASD groups was plotted as group were generated for accuracy and response times at a function of difficulty to examine visually the verbal each difficulty level, which were analysed using two-way working memory activation patterns with increasing mixed ANOVAs with difficulty level (D3, D4, D5 and D6) cognitive load. as a within-subject factor and group (ASD and TD) as a between-subject factor. Results Behavioural data fMRI data analyses There was a weak but significant effect of group on accur- fMRI data were preprocessed using tools from FMRIB’s acy, F(1,55) = 4.06, p = 0.049, in which TD children per- Software Library: FSL  and AFNI . The initial formed slightly better than children with ASD. There was a three volumes were discarded from each run to ensure significant main effect of difficulty level on accuracy, with scanner stabilisation. 3dvolreg was used for interleaved accuracy decreasing as a function of difficulty, F(2.50, 165) slice-timing and McFlirt motion correction; the data = 80.26, p < 0.001 (Greenhouse-Geisser corrected degrees were smoothed in place using a 6-mm FWHM Gaussian of freedom). There was also a significant group × kernel and temporally filtered (0.01–0.2 Hz) then con- level interaction, F(2.50,165) = 2.98, p = 0.043 (Green- verted to percent signal change from baseline volumes. house-Geisser corrected), in which group differences Images were registered to the Montreal Neurological in performance became larger with increasing task Institute (MNI) 152 brain template. The maximum Eu- difficulty (see Fig. 2a). Post hoc t tests revealed that clidean displacement (MD) travelled by any brain voxel group performance did not differ on D3 (t(55) = 0.12, was calculated for each volume from the six rigid body p = 0.91) and D4 (t(55) = 1.45, p = 0.15), whereas TD transformation parameters. This MD metric was used to children performed somewhat better than children identify volumes with motion surpassing the minimum with ASDonD5(t(55) = 2.15, p = 0.04) and D6 (t(55) motion threshold. Each subject’s average MD was used = 2.14, p = 0.04). to examine group motion differences (TD: M = 0.47, SD There was no significant effect of group on response times, = 0.40; ASD: M = 0.50, SD = 0.47; t = 0.25, ns.). F(1,55) = 0.29, p = 0.59, or on group × level interaction effect, (51) Data analyses were performed using FSL fMRI expert F(1.81, 165) =2.70, p = 0.077 (Greenhouse-Geisser corrected analysis tool (FEAT) . The data were fit to a degrees of freedom). There was a significant effect of block-design general linear model combined with a load on response times, F(1.81, 165) = 83.82, p < 0.001 gamma function used to model haemodynamic changes, (Greenhouse-Geisser corrected), with response times with D3 to D6 task parameters. IQ and age were both increasing as a function of difficulty across groups assessed as confounding variables using FSL FEAT and (see Fig. 2b). were both found to have no significant impact on BOLD response during LMT. Linear trend analyses were per- Within-group fMRI results formed using levels D3 to D6 with fixed-effects higher Typically developing children showed significantly in- level modelling to examine areas that linearly modulated creasing activation as a function of increasing cognitive as a function of task difficulty. Linear trend analyses load (i.e. positive linear trend between BOLD signal and Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 6 of 12 Fig. 2 LMT behavioural performance. a Mean proportion correct for levels D3 to D6 with standard error bars. b Average reaction times for levels D3 to D6 with standard error bars difficulty level) in the occipital, parietal, fusiform, cingulate and frontal areas (Fig. 3;see Table 2 for a complete list). Regions that showed decreasing activation as a function of Fig. 3 Group activation maps for the linear trend analyses in TD and load (i.e. negative linear relations between BOLD signal ASD groups during LMT. Significant activations using cluster-based and difficulty level) included the medial frontal, anterior thresholding determined by Z > |2.3| and a corrected cluster cingulate, bilateral temporal and parietal gyri and precu- significant threshold of p = 0.05. Areas in orange depict regions of increasing activation as a function of difficulty (positive linear neus and cingulate cortices (Table 2 and Fig. 3). relations between cortical activity and task difficulty level), and areas Children with ASD showed increasing activation with in blue depict regions of decreasing activation (negative linear greater WM load in the occipital gyri, fusiform, precu- relations between cortical activity and task difficulty level). ACC neus and inferior frontal gyrus (Fig. 3; see Table 3 for a anterior cingulate cortex, mPFC medial prefrontal cortex, PCC complete list). Areas that showed decreasing activation posterior cingulate cortex, Cing cingulate, AngG angular gyrus, Prec precuneus, dlPFC dorsolateral prefrontal cortex, IFG inferior frontal across task difficulty included the parietal lobule, middle gyrus, IPL inferior parietal gyrus, SMG superior medial gyrus, SFG temporal, cingulate, precuneus and frontal gyri (Fig. 3 superior frontal gyrus, STG superior temporal gyrus, mOG medial and Table 3). occipital gyrus, Cun cuneus, Cereb cerebellum, vmPFC ventromedial prefrontal cortex Between-group comparison TD children had significantly stronger positive linear re- lations between activation and cognitive load compared ASD group showed greater linear activation across task to children with ASD in the bilateral prefrontal cortex, difficulty than TD children. There were no significant precuneus and inferior parietal lobule (Table 4; Fig. 4). differences between groups in the patterns of deactiva- In these regions, TD children showed increasing activa- tion with increasing cognitive load. tion with increasing task difficulty, whereas the ASD group failed to show a positive linear trend (see Fig. 5 Discussion for graphs of the percent signal change of the ROIs of Protracted development of the frontal lobes in combin- cortical areas that had significant between-group differ- ation with vulnerability to neurodevelopmental distur- ences in linear patterns). There were no areas where the bances emphasises the need for deeper understanding of Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 7 of 12 Table 2 Linear trend analyses across difficulty levels for TD children Voxels MNI coordinates Z p value Hem. Region value xy z −25 Regions where activation increases with 14,922 − 28 − 76 − 4 4.81 6.31 × 10 L Middle occipital gyrus difficulty (increasing BOLD signal) x − 24 − 62 52 4.73 L Superior parietal lobule x − 26 − 72 − 10 4.51 L Fusiform gyrus x − 20 − 92 16 4.5 L Cuneus x28 − 58 58 4.44 R Inferior parietal lobule −24 14,033 − 8 18 42 5.41 6.92 × 10 L Cingulate gyrus x − 40 2 30 4.96 L Inferior frontal gyrus x 10 22 38 4.83 R Cingulate gyrus x − 4 28 28 4.76 L Anterior cingulate cortex −9 3447 50 32 32 4.57 9.29 × 10 R Middle frontal gyrus x 32 18 3 4.47 R Insula −12 Regions where activation decreases with 5355 2 38 − 22 5.38 4.32 × 10 R Medial frontal gyrus difficulty (decreasing BOLD signal) x −640 − 12 4.74 L Medial frontal gyrus x − 14 34 − 16 4.51 L Inferior frontal gyrus x6 62 − 6 4.5 R Superior frontal gyrus x2 14 − 6 4.41 R Anterior cingulate cortex −12 5165 − 60 − 26 − 18 4.72 8.85 × 10 L Middle temporal gyrus x − 40 − 78 40 4.54 L Precuneus/angular gyrus x − 56 − 62 30 4.46 L Angular gyrus −10 4472 60 − 22 20 4.84 1.31 × 10 R Supramarginal gyrus x66 − 30 34 4.56 R Inferior parietal lobule x60 − 56 2 4.45 R Middle temporal gyrus x68 − 38 12 4.24 R Superior temporal gyrus −9 3615 − 16 − 48 36 4.25 4.5 × 10 L Cingulate gyrus x2 − 44 34 4.04 R Cingulate gyrus the function of these regions with typical and atypical activation patterns, we observed a larger spread of activa- development. Previous research has been centred on tion in children in this study compared to adults from WM in adolescents and adults with ASD, leaving a gap Vogan et al.’s study. This is consistent with the study in our understanding of WM in children with ASD. This by Geier et al. , who performed a visual spatial work- is the first study to investigate the neural correlates of ing memory oculomotor delayed-response task with verbal WM in children and young adolescents with ASD adults, adolescents and children, and found that while all compared to TD youth and examine the impact of cog- three age groups showed recruitment of a common net- nitive load. work including the frontal, parietal and temporal regions, The TD group showed increasing recruitment of the children and adolescents showed a wider distribution in brain areas classically linked to WM as a function of in- addition to that network. creasing cognitive demand and decreasing activation in The behavioural data showed comparable performance regions associated with the DMN. The group with ASD, on D3 and D4 between the TD and ASD groups; how- however, did not show this opposing system of cognitive ever, TD children performed with a higher accuracy on processing. Specifically, TD children recruited the D5 and D6. These between-group differences emerging prefrontal and parietal cortical regions, areas directly at higher cognitive loads are consistent with the litera- correlated with verbal WM [3, 32, 38, 41, 42, 57], as a ture suggesting that WM in children with ASD, when function of cognitive load significantly more than children compared with TD children, is similar for simpler tasks with ASD who only demonstrated load-dependent deacti- but deficient for more complex tasks or those with vation in DMN regions. In a qualitative examination of greater cognitive demand [8, 22–25, 27, 29]. Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 8 of 12 Table 3 Linear trend analyses across difficulty levels for children with ASD Voxels MNI coordinates Z p value Hem. Region value xy z −6 Regions where activation increases with 2170 − 28 − 76 − 6 4.25 3.64 × 10 L Middle occipital gyrus difficulty (increasing BOLD signal) x − 24 − 68 − 6 4.09 L Lingual gyrus/fusiform gyrus x − 28 − 54 − 6 3.91 L Lingual gyrus x − 20 − 86 12 3.85 L Cuneus/middle occipital gyrus x − 22 − 86 22 3.82 L Precuneus −3 1120 − 2 − 74 − 26 4.19 1.31 × 10 L Cerebellar vermis x28 − 72 − 10 3.62 R Fusiform/lingual gyrus x0 − 58 − 32 3.51 L Culmen/cerebellar vermis x − 10 − 64 − 30 3.16 L Cerebellum −3 960 6 − 28 − 12 4.29 3.68 × 10 R Thalamus x6 − 34 − 20 4.01 R Culmen x − 8 − 24 − 12 3.43 L Thalamus x0 − 34 − 46 3.42 L Brain-stem x − 2 − 34 − 30 3.41 L Culmen −2 610 − 42 − 4 26 4.04 4.4 × 10 L Inferior frontal gyrus x − 58 18 32 3.89 L Middle frontal gyrus x − 38 22 18 3.05 L Insula/inferior frontal gyrus −9 Regions where activation decreases with 3908 42 − 68 42 4.92 1.31 × 10 R Inferior parietal/angular gyrus difficulty (decreasing BOLD signal) x48 − 58 40 4.52 R Inferior parietal lobule x48 − 50 36 4.49 R Angular gyrus x48 − 46 36 4.46 R Supramarginal gyrus x46 − 44 0 4.1 R Middle temporal gyrus −9 3816 2 − 32 44 4.38 1.92 × 10 R Cingulate gyrus x − 6 − 38 38 4.19 L Cingulate gyrus x6 − 68 36 4.02 R Precuneus x12 − 46 30 3.79 R Cingulate gyrus x − 16 − 24 44 3.73 L Cingulate gyrus −5 1810 − 64 − 40 30 4.37 2.38 × 10 L Inferior parietal lobule x − 56 − 56 34 3.97 L Angular gyrus −3 843 − 12 62 12 4.07 8.14 × 10 L Medial frontal gyrus x − 22 66 14 3.77 L Middle frontal gyrus x − 18 52 8 3.58 L Superior frontal gyrus x 16 58 4 3.33 R Superior frontal gyrus x − 20 42 4 3.32 L anterior cingulate −2 604 22 58 18 3.44 4.6 × 10 R superior frontal gyrus x 24 52 32 3.29 R middle frontal gyrus Results from linear trend analyses from D3 to D6 for children with ASD. Areas that increased as a function of difficulty level are associated largely with visual processing, whereas areas that decreased as a function of difficulty level are associated with the default mode network. MNI coordinates represent the peak Z value of the cluster, X peak local maximas within cluster The ASD youth did not show comparable increasing was expected. The further activity that was greater in activity in frontal-parietal regions with increased mem- the TD group than the ASD group in the cingulate and ory load, as the TD group. The frontal areas (BA 9) and precuneus could be due to increased recruitment of cog- inferior parietal lobe are classic areas for working mem- nitive control mechanisms due to task difficulty, as both ory , and activity in this WM task in the TD group the anterior cingulate cortex (ACC) and precuneus are Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 9 of 12 Table 4 Regions of significant differences between TD and ASD groups Voxels MNI coordinates Z p value Hem. Region value xyz −4 1341 − 10 40 26 3.38 3.37 × 10 L Medial frontal gyrus x − 14 8 50 3.32 L Cingulate gyrus x − 16 6 62 3.26 L Superior frontal gyrus −4 1175 38 − 36 44 3.58 9.25 × 10 R Inferior parietal lobule −3 1095 14 − 52 46 3.7 1.53 × 10 R Precuneus x − 6 − 62 52 3.32 L Precuneus −3 833 54 36 22 3.73 8.73 × 10 R Middle frontal gyrus x 28 42 38 3.16 R Superior frontal gyrus Results from between group comparisons of the linear trend analyses from D3 to D6. All regions reported are areas where TD children showed greater positive linear relations between cortical activity and difficulty level (increasing BOLD signal with increasing task difficulty) than children with ASD. There were no areas where children with ASD showed greater linear relations between cortical activity and difficulty level than TD children. MNI coordinates represent the peak Z values of the cluster; X peak local maximas within cluster key hubs in cognitive networks. The differences between periods and show decreased BOLD signals during tasks the groups were despite both completing the task suc- , particularly tasks that are cognitively demanding. cessfully. Although the accuracy of the ASD group was This modulation is believed to contribute to more effi- lower than the TD group at the two higher load levels, cient cognitive processing, and DMN regions are ex- they were performing the task similarly at D3 and D4 pected to deactivate with increasing task difficulty. The and were still at acceptable levels for D5 and D6. This fact that here we saw no difference between the ASD suggests that the ASD group had unconventional utilisa- and the TD groups in the decreasing activation of DMN tion of the brain areas for the WM task. This is concord- regions with increasing task load could be due to a ant with the model that activation is more idiosyncratic slightly older age range than previous studies, suggesting in those with ASD, as reported elsewhere . This that DMN modulation may ‘catch up’ in children with leads to the usual regions not being seen in the ASD ASD as they move into the teenage years. This is sup- group analysis, and the more typical regions emerging as ported by a similar longitudinal protocol with somewhat more active in the TD group in the group comparison. older cohort , where the DMN modulation increased With a larger study, idiosyncratic patterns could be in- compared to 2 years earlier. These combined results in- vestigated specifically to determine if there are ASD sub- dicate that the DMN modulation develops in ASD, albeit groups with distinct alternative strategies. later than in the TD group, while the working memory Following on from this notion, future larger studies processes remain distinct. should also determine the role of other cognitive steps A limitation of the current study was the use of only or strategies that may differ between TD and ASD linear models in the analyses. This was chosen as this is groups that could influence WM performance. We in- a subsequent study from our normative series , and cluded irrelevant aspects in the stimuli, to allow better a sister study to two other papers using a colour match- determination of WM , but irrelevant details may ing task [55, 56] all of which used the same analytic also impact selective filtering and attention, which has procedures, and we wanted to be able to relate the find- been linked to working memory capacity . As some ings across the studies. Other approaches could be used researchers have found heightened visual processing in in the future that investigate non-linear changes as a those with ASD [60, 61] particularly in relation to local function of group (e.g. ) or with WM load, such as features , this visual strategy may emerge more com- logarithmic changes that would be seen as rapidly in- monly in an ASD group and potentially impact strategy, creasing activation and then a plateau. Another limita- and hence underlying neural recruitment. Future work tion is that we had to exclude children who could not could include assessments of visual processing skills (see stay still in the scanner and who did not perform ad- ) and use that as a means of subgrouping partici- equately on the task to ensure brain behaviour-related pants by cognitive processing preferences. activation. By doing so, we were unable to include lower A number of studies have reported atypical DMN acti- functioning children with ASD, and thus, our results are vation in ASD [64–67], including an earlier investigation generalizable to higher functioning children only. There with the similar but visuo-spatial colour task (CMT) was also, on average, a lower IQ in the ASD youth and . The DMN is a well-established network of the greater IQ variability. This is typical of this population, brain regions that are active during rest or non-task but even when IQ was covaried, the effects remained, Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 10 of 12 Fig. 4 Results from between-group comparisons. Significant activations using cluster-based thresholding determined by Z > |2.3| and a corrected cluster significant threshold of p = 0.05. Areas in red/orange depict regions where the control children showed greater linear activation trends across difficulty level in the negative or positive direction than children with ASD. medPFC medial prefrontal cortex, Cing cingulate, Prec precuneus, IPL inferior parietal gyrus, mFG middle frontal gyrus suggesting that the effects were robust within the higher increasing verbal information. Impaired verbal working IQ range, despite the group differences in IQ. Further memory in ASD would have important academic and social fMRI investigations are required with less demanding implications. Specially, verbal WM difficulties could inter- protocols to understand verbal WM function in low fere with children’s ability to recall verbal information from functioning children with ASD, who may present unique conversations and social interactions, as well as to neural profiles. Finally, we had a wide age range in the learn verbal material from classroom lessons or follow study. We matched groups on age and age did not con- instructions. Determining the neural deficits of WM tribute to group effects. Nevertheless, smaller age ranges in children with ASD will help us understand the ori- are ideal, and with a larger sample, age-related effects gins of the behaviours associated with ASD. Brain could be explored. functional abnormalities in ASD may drive behav- ioural symptoms and give rise to cognitive impair- Conclusions ments. Thus, exploring the neural correlates of WM The results from this study have several important implica- contributes to knowledge of the ASD behavioural pheno- tions. Our findings that children with ASD, relative to TD types. Finally, our study helps determine the nature of children, demonstrate inadequate modulation of neural atypical neurodevelopment, which could help establish or capacity suggest that they could become overwhelmed with monitor interventions for WM function in ASD. Fig. 5 Mean peak cluster percent signal changes and standard errors plotted as a function of difficulty level. Areas where children with ASD differed significantly from TD children in the linear trend analyses Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 11 of 12 Abbreviations 5. Hill EL. Executive dysfunction in autism. Trends Cogn Sci. 2004;8:26–32. ADOS: Autism Diagnostic Observation Schedule; ASD: Autism spectrum 6. Joseph RM. Neuropsychological frameworks for understanding autism. Int disorder; BOLD: Brain-oxygen-level dependent; D3, D4 etc.: Difficulty level 3, Rev Psychiatry. 1999;11:309–24. difficulty level 4, etc.; dlPFC: Dorsolateral prefrontal cortex; DMN: Default 7. Luna B, Minshew NJ, Garver KE, Lazar NA, Thulborn KR, Eddy WF, Sweeney JA. mode network; fMRI: Functional magnetic resonance imaging; FWHM: Full Neocortical system abnormalities in autism: an fMRI study of spatial working width half maximum; LMT: Letter matching task; MD: Maximum Euclidean memory.Neurology.2002;59:834–40. displacement; MNI: Montreal Neurological Institute; PFC: Prefrontal cortex; 8. Russo N, Flanagan T, Iarocci G, Berringer D, Zelazo PD, Burack JA. ROIs: Regions of interest; TD: Typically developing; WM: Working memory Deconstructing executive deficits among persons with autism: implications for cognitive neuroscience. Brain Cogn. 2007;65:77–86. Acknowledgements 9. Koshino H, Carpenter PA, Minshew NJ, Cherkassky VL, Keller TA, Just MA. We would like to thank first the families and children for their Functional connectivity in an fMRI working memory task in high-functioning participation. Many thanks to Rachel Leung for administering the autism. Neuroimage. 2005;24:810–21. ADOS-R and ADOS-2 and Dr. Jessica Brian for reviewing the assessments. 10. Koshino H, Kana RK, Keller TA, Cherkassky VL, Minshew NJ, Just MA. fMRI Warmest thanks to our MRI technologists, Tammy Rayner and Ruth investigation of working memory for faces in autism: visual coding and Weiss, forall theirincrediblehelpin data acquisition. Lastly, thank you underconnectivity with frontal areas. Cereb Cortex. 2008;18:289–300. to Crescent School, Toronto, for their support and participation in this 11. O’Hearn KAM, Ordaz S, Luna B. Neurodevelopment and executive function project. in autism. Dev Psychopathol. 2008;20:1103–32. 12. Silk TJ, Rinehart N, Bradshaw JL, Tonge B, Egan G, O’Boyle MW, Cunnington Funding R. Visuospatial processing and the function of prefrontal-parietal networks This research was funded by the Canadian Institutes of Health Research in autism spectrum disorders: a functional MRI study. Am J Psychiatry. 2006; (CIHR) grants MOP-106582 and MOP-142379. 163:1440–3. 13. Powell KB, Voeller KK. Prefrontal executive function syndromes in children. J Availability of data and materials Child Neurol. 2004;19:785–97. The datasets used and analysed during the current study are available from 14. Sowell ER, Thompson PM, Leonard CM, Welcome SE, Kan E, Toga AW. the corresponding author on reasonable request. Longitudinal mapping of cortical thickness and brain growth in normal children. J Neurosci. 2004;24:8223–31. Authors’ contributions 15. O’Hare ED, Lu LH, Houston SM, Bookheimer SY, Sowell ER. VV is responsible for the recruitment, data acquisition, analysis and Neurodevelopmental changes in verbal working memory load-dependency: interpretation and writing and editing of the manuscript. KF also contributed an fMRI investigation. Neuroimage. 2008;42:1678–85. significantly to the data analyses, interpretation and writing and editing of 16. Baddeley A. Working memory. Science. 1992;255:556–9. the manuscript. BM participated in the fMRI analysis and revising the 17. Case R. Exploring the Conceptual Underpinnings of Children’s Thought and manuscript. MS advised on patient testing, study design and revising the Knowledge. Hillsdale: Erlbaum; 1992. manuscript. MT initiated the study and participated in the design, analyses 18. Engle RW, Tuholski SW, Laughlin JE, Conway AR. Working memory, short- and writing and revising of the manuscript. All authors read and approved term memory, and general fluid intelligence: a latent-variable approach. J the final manuscript. Exp Psychol Gen. 1999;128:309–31. 19. Graf P, Uttl B, Dixon R. Prospective and retrospective memory in adulthood. Ethics approval and consent to participate In: Graf P, Ohta N, editors. Lifespan development of human memory. All the experimental procedures used were approved by SickKids Research Cambridge: MIT Press; 2002. p. 257-82. Ethics Board. All participants gave informed verbal assent, and the parent or 20. TP A. Working memory, but not IQ, predicts subsequent learning in children legal guardian of all participants gave informed written consent to with learning difficulties. Eur J Psychol Assess. 2009;25:92–8. participate. 21. Dennis M, Agostino A, Roncadin C, Levin H. Theory of mind depends on domain-general executive functions of working memory and cognitive Competing interests inhibition in children with traumatic brain injury. J Clin Exp Neuropsychol. The authors declare that they have no competing interests. 2009;31:835–47. 22. Ozonoff S, Strayer DL. Further evidence of intact working memory in autism. J Autism Dev Disord. 2001;31:257–63. Publisher’sNote 23. Russell J, Jarrold C, Henry L. Working memory in children with autism Springer Nature remains neutral with regard to jurisdictional claims in and with moderate learning difficulties. J Child Psychol Psychiatry. 1996; published maps and institutional affiliations. 37:673–86. 24. Steele SD, Minshew NJ, Luna B, Sweeney JA. Spatial working memory Author details deficits in autism. J Autism Dev Disord. 2007;37:605–12. Diagnostic Imaging & Research Institute, Hospital for Sick Children, 555 25. Bennetto L, Pennington BF, Rogers SJ. Intact and impaired memory University Avenue, Toronto, Ontario M5G 1X8, Canada. Department of functions in autism. Child Dev. 1996;67:1816–35. Applied Psychology and Human Development, Ontario Institute for Studies 26. Minshew NJ, Goldstein G. The pattern of intact and impaired memory in Education, University of Toronto, 252 Bloor Street West, Toronto, Ontario functions in autism. J Child Psychol Psychiatry. 2001;42:1095–101. M56 1V6, Canada. Department of Psychology, University of Toronto, 100 St. 27. Williams DL, Goldstein G, Carpenter PA, Minshew NJ. Verbal and spatial George St., Toronto, Ontario M5S 3G3, Canada. working memory in autism. J Autism Dev Disord. 2005;35:747–56. 28. Williams DL, Goldstein G, Minshew NJ. The profile of memory function in Received: 21 November 2017 Accepted: 22 May 2018 children with autism. Neuropsychology. 2006;20:21–9. 29. Gabig CS. Verbal working memory and story retelling in school-age children with autism. Lang Speech Hear Serv Sch. 2008;39:498–511. References 30. Southwick JS, Bigler ED, Froehlich A, Dubray MB, Alexander AL, Lange N, 1. American Psychiatric Association. Diagnostic and statistical manual of Lainhart JE. Memory functioning in children and adolescents with autism. mental disorders (5th ed.). Arlington: American Psychiatric Publishing; 2013. Neuropsychology. 2011;25:702–10. 2. Barnard L, Muldoon K, Hasan R, O’Brien G, Stewart M. Profiling executive dysfunction in adults with autism and comorbid learning disability. Autism. 31. Fletcher PC, Henson RN. Frontal lobes and human memory: insights from 2008;12:125–41. functional neuroimaging. Brain. 2001;124:849–81. 3. Greene CM, Braet W, Johnson KA, Bellgrove MA. Imaging the genetics of 32. Owen AM, McMillan KM, Laird AR, Bullmore E. N-back working memory executive function. Biol Psychol. 2008;79:30–42. paradigm: a meta-analysis of normative functional neuroimaging studies. 4. Happe F, Booth R, Charlton R, Hughes C. Executive function deficits in Hum Brain Mapp. 2005;25:46–59. autism spectrum disorders and attention-deficit/hyperactivity disorder: 33. Geier CF, Garver K, Terwilliger R, Luna B. Development of working memory examining profiles across domains and ages. Brain Cogn. 2006;61:25–39. maintenance. J Neurophysiol. 2009;101:84–99. Vogan et al. Journal of Neurodevelopmental Disorders (2018) 10:19 Page 12 of 12 34. Nagel BJ, Herting MM, Maxwell EC, Bruno R, Fair D. Hemispheric 58. Byrge L, Dubois J, Tyszka JM, Adolphs R, Kennedy DP. Idiosyncratic brain lateralization of verbal and spatial working memory during adolescence. activation patterns are associated with poor social comprehension in Brain Cogn. 2013;82:58–68. autism. J Neurosci. 2015;35:5837–50. 35. Thomason ME, Race E, Burrows B, Whitfield-Gabrieli S, Glover GH, Gabrieli 59. Vogel EK, McCollough AW, Machizawa MG. Neural measures reveal JD. Development of spatial and verbal working memory capacity in the individual differences in controlling access to working memory. Nature. human brain. J Cogn Neurosci. 2009;21:316–32. 2005;438:500–3. 36. Damarla SR, Keller TA, Kana RK, Cherkassky VL, Williams DL, Minshew NJ, 60. Mottron L, Burack JA, Iarocci G, Belleville S, Enns JT. Locally oriented Just MA. Cortical underconnectivity coupled with preserved visuospatial perception with intact global processing among adolescents with high- functioning autism: evidence from multiple paradigms. J Child Psychol cognition in autism: evidence from an fMRI study of an embedded Psychiatry. 2003;44:904–13. figures task. Autism Res. 2010;3:273–9. 61. Mottron L, Dawson M, Soulieres I, Hubert B, Burack J. Enhanced perceptual 37. Kana RK, Liu Y, Williams DL, Keller TA, Schipul SE, Minshew NJ, Just MA. The functioning in autism: an update, and eight principles of autistic perception. local, global, and neural aspects of visuospatial processing in autism J Autism Dev Disord. 2006;36:27–43. spectrum disorders. Neuropsychologia. 2013;51:2995–3003. 62. Plaisted K, Saksida L, Alcantara J, Weisblatt E. Towards an understanding of 38. Carlson S, Martinkauppi S, Rama P, Salli E, Korvenoja A, Aronen HJ. the mechanisms of weak central coherence effects: experiments in visual Distribution of cortical activation during visuospatial n-back tasks as configural learning and auditory perception. Philos Trans R Soc Lond Ser B revealed by functional magnetic resonance imaging. Cereb Cortex. Biol Sci. 2003;358:375–86. 1998;8:743–52. 63. Dakin S, Frith U. Vagaries of visual perception in autism. Neuron. 2005; 39. Cohen JD, Perlstein WM, Braver TS, Nystrom LE, Noll DC, Jonides J, Smith EE. 48:497–507. Temporal dynamics of brain activation during a working memory task. 64. Chen CP, Keown CL, Jahedi A, Nair A, Pflieger ME, Bailey BA, Muller RA. Nature. 1997;386:604–8. Diagnostic classification of intrinsic functional connectivity highlights 40. de Vries M, Geurts HM. Beyond individual differences: are working memory somatosensory, default mode, and visual regions in autism. Neuroimage and inhibition informative specifiers within ASD? J Neural Transm (Vienna). Clin. 2015;8:238–45. 2014;121:1183–98. 65. Kennedy DP, Redcay E, Courchesne E. Failing to deactivate: resting functional 41. Fuster JM. Prefrontal neurons in networks of executive memory. Brain Res abnormalities in autism. Proc Natl Acad Sci U S A. 2006;103:8275–80. Bull. 2000;52:331–6. 66. Lynch CJ, Uddin LQ, Supekar K, Khouzam A, Phillips J, Menon V. Default 42. Kwon H, Reiss AL, Menon V. Neural basis of protracted developmental mode network in childhood autism: posteromedial cortex heterogeneity changes in visuo-spatial working memory. Proc Natl Acad Sci U S A. 2002; and relationship with social deficits. Biol Psychiatry. 2013;74:212–9. 99:13336–41. 67. Washington SD, Gordon EM, Brar J, Warburton S, Sawyer AT, Wolfe A, 43. Nelson CA, Monk CS, Lin J, Carver LJ, Thomas KM, Truwit CL. Functional Mease-Ference ER, Girton L, Hailu A, Mbwana J, et al. Dysmaturation of the neuroanatomy of spatial working memory in children. Dev Psychol. 2000;36: default mode network in autism. Hum Brain Mapp. 2014;35:1284–96. 109–16. 68. Raichle ME. The brain’s default mode network. Annu Rev Neurosci. 44. Rahko JS, Vuontela VA, Carlson S, Nikkinen J, Hurtig TM, Kuusikko-Gauffin S, 2015;38:433–47. Mattila ML, Jussila KK, Remes JJ, Jansson-Verkasalo EM, et al. Attention and 69. Vakorin VA, Doesburg SM, Leung RC, Vogan VM, Anagnostou E, Taylor MJ. working memory in adolescents with autism spectrum disorder: a Developmental changes in neuromagnetic rhythms and network synchrony functional MRI study. Child Psychiatry Hum Dev. 2016;47:503–17. in autism. Ann Neurol. 2017;81:199–211. 45. Urbain CM, Pang EW, Taylor MJ. Atypical spatiotemporal signatures of working memory brain processes in autism. Transl Psychiatry. 2015;5:e617. 46. Arsalidou M, Pascual-Leone J, Johnson J, Morris D, Taylor MJ. A balancing act of the brain: activations and deactivations driven by cognitive load. Brain Behav. 2013;3:273–85. 47. Powell TL, Vogan VM, Arsalidou M, Taylor MJ. Controlled interference and assessments of developmental working memory capacity: Evidence from the letter and colour matching tasks. Child Dev Res. 2014;6(1):19. 48. Vogan VM, Morgan BR, Powell TL, Smith ML, Taylor MJ. The neurodevelopmental differences of increasing verbal working memory demand in children and adults. Dev Cogn Neurosci. 2016;17:19–27. 49. Wechsler D. Wechsler intelligence scale for children. San Antonia: Psychological Corporation; 2003. 50. Lord C, Risi S, Lambrecht L, Cook EH Jr, Leventhal BL, DiLavore PC, Pickles A, Rutter M. The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord. 2000;30:205–23. 51. P-LJ AM, Johnson J. Misleading cues improve developmental assessment of working memory capacity: the colour matching tasks. Cogn Dev. 2010;25:262–77. 52. Woolrich MW, Behrens TE, Beckmann CF, Jenkinson M, Smith SM. Multilevel linear modelling for FMRI group analysis using Bayesian inference. Neuroimage. 2004;21:1732–47. 53. Cox RW. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res. 1996;29:162–73. 54. Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T, Beckmann C, Jenkinson M, Smith SM. Bayesian analysis of neuroimaging data in FSL. Neuroimage. 2009;45:S173–86. 55. Vogan VM, Morgan BR, Lee W, Powell TL, Smith ML, Taylor MJ. The neural correlates of visuo-spatial working memory in children with autism spectrum disorder: effects of cognitive load. J Neurodev Disord. 2014;6:19. 56. Vogan VM, Morgan BR, Smith ML, Taylor MJ. Functional changes during visuo-spatial working memory in autism spectrum disorder: 2-year longitudinal fMRI study: Autism; 2018. In Press 57. Baddeley A. Working memory: looking back and looking forward. Nat Rev Neurosci. 2003;4:829–39.
Journal of Neurodevelopmental Disorders
– Springer Journals
Published: Jun 1, 2018