ASD Toddlers Exhibit Impaired Development of Neural Systems That Respond to and Guide Mother-Child Interactions

Motherese is an experience-expectant, human-specic and innate form of parent speech that enhances social and language learning, and affect and emotion development in infants. An early sign of ASD is the child’s lack of responding to motherese and reduced social mother-child interactions. To learn why, we devised a novel experiment quantifying (a) neural responses to motherese and other emotion speech with sleep fMRI and (b) active behavioral preference for motherese with eye tracking in ASD and TD toddlers. We combined the power of diverse neural and clinical data types using Similarity Network Fusion to reveal four neural-clinical clusters. The ASD cluster with the weakest neural responses to motherese and the poorest social and language abilities had the lowest eye tracking attention to motherese, while the TD cluster with the strongest neural response to motherese showed the opposite effects. We conclude that the ASD child’s impairment in engaging in social mother-child interactions is due to impaired development of innate neural systems that normally respond to and guide behavior that maintains mother-child interactions.


Introduction
Social and language development are inextricably linked in typical infants 1 as a result of experience during mother-infant interactions. In these interactions, it is theorized that motherese or "infant-directed speech" is an experience-expectant form of speech that enhances social and language learning as well as affect and emotion development in babies and infants 2 . It has the human-unique and special characteristics of higher pitch, slower tempo, and exaggerated speech contours accompanied by heightened positive affect. The prevailing view, then, is that motherese utterances create and maintain a mother-infant social and language interaction which promotes learning. Implicit is that the infant's neural response to motherese mediates this. Motherese occurs in all cultures and is thought to have a genetic basis that evolutionarily emerged in humans [3][4][5] . Typically developing (TD) infants attend to and prefer "motherese" over other forms of adult speech 2, [6][7][8][9] , and a small number of behavioral and neuroimaging studies suggest TD infants may process motherese differently from non-speech sounds [10][11][12][13][14] . However, the general theory that neural responsiveness to motherese speech in TD infants relates to variation in behavioral attention towards motherese and social and language learning and development remains largely untested. Impairments in social and language learning and development are early-age signs of autism spectrum disorder (ASD), and the causes are essential to understand [15][16][17] . While behavioral literature has thoroughly described mother-infant joint social interaction de cits in ASD 18,19 , there is remarkably scant evidence about how these de cits arise and whether a key causal factor may be abnormally reduced neural and behavioral responsiveness of ASD infants to motherese. There are no studies that characterize impairments in neural processing of motherese in ASD. If neural responses to motherese in ASD infants are reduced, then behavioral responses to motherese should likewise be reduced. This would then disconnect the innate mother-infant interaction loop and result in reduced social and language learning and ability.
Here we hypothesize that ASD infants and toddlers exhibit impaired development of innately-driven neural mechanisms that respond to motherese, which underlies reduced behavioral preference for and attentiveness to motherese. To investigate this hypothesis, we measured functional magnetic resonance imaging (fMRI) responses to motherese speech as well as to non-motherese speech that has less affective prosody; assessed social and language development; and quanti ed gaze-contingent behavioral responses to females speaking motherese versus non-speech computer sounds and images. We rst tested how TD and ASD differ in their neural response to speech, and also examined how TD and ASD brain response relates to social and language development as well as behavioral responses to motherese. Our prediction was that individual differences in early-age social and language development would be associated with a child's neural responses. Thus, for example, ASD infants and toddlers with the poorest neural responses to motherese, were predicted to have the most severe social symptoms, poorest language outcomes, and most impairment of behavioral preference and attention towards motherese. The converse was predicted for TD toddlers with the strongest neural responses to motherese; as a strong test of this, we included TD toddlers with a wide range of language abilities. Therefore, we expected across-subject heterogeneity as well as within-subject concordance across multiple modalities, including neural, symptom, and behavioral data.
Importantly, our assessment of the covariation of brain and behavioral traits included an examination of whether there were homogeneous subgroups within our sample of ASD and TD toddlers that differed with respect to attentional preference for motherese speech. We used Similarity Network Fusion (SNF), a novel unbiased, data-driven approach 20,21 , not previously used in the ASD eld, to identify patterns of neural and clinical characteristics that underlie greater and lesser eye tracking behavioral preference for and attention to females speaking motherese. The SNF approach integrates multiple and very different kinds of measures (e.g., fMRI, clinical) from individuals to identify clusters of subjects whose measures are maximally similar to each other and different from subjects in other clusters. In this approach, rather than determining diagnostic categories a priori through clinical measures, the similarity in patterns of different kinds of measures drives clusters independent of diagnosis and in this way reveals bio-behavioral dimensionality. The target outcome of interest, in this case, i.e., a child's preference and interest in motherese utterances and other affective prosody stimuli, can then be objectively and independently examined and tested for each neural-clinical cluster/subgroup.

Results
Similar activation patterns in TD adults and TD toddlers. We collected a total of 200 fMRI datasets from 71 toddlers and 41 fMRI datasets from 14 adults, separately (Supplementary Table 1). We rst analyzed task activation data collected from sleeping TD toddlers and awake TD adults. For each language paradigm, whole-brain activation maps showed similar patterns in sleeping TD toddlers and those in passively listening adults (Fig. 1a), though, as expected, the percent signal changes in bilateral temporal regions of interest (ROIs)-from meta-analytic activation map in Neurosynth (https://neurosynth.org/) with the term "language"-were signi cantly lower in sleeping TD toddlers than awake TD adults (ROIs in Fig. 2; plots of toddler vs. adult differences in Supplementary Fig. 1). These results validate that activation during sleep in TD toddlers re ects similar language patterns to those in awake adults albeit with reduced activation strength.
In addition, we evaluated the test-retest reliability of brain activation to language paradigms with varying levels of affective prosody within individual toddlers across time. The test-retest scans were divided into two groups based on intervals between initial and retest scans: short-term retest (1-4 months after initial scans) and long-term retest (12-15 months after initial scans). The overall test-retest reliability (initial scans versus all retest scans) was quanti ed with intraclass correlation coe cients, which showed moderate to good reliability for Karen language and Motherese but poor reliability for Story language, especially in the right temporal region (Fig. 1b).
Reduced neural response to speech with varying prosodic levels in ASD. We found robust and signi cant activation in temporal language regions in TD toddlers but reduced activation in ASD toddlers across all levels of affective prosody (Fig. 3a). Although there were no signi cant differences in whole-brain activation between TD and ASD toddlers, a ROI-based analysis demonstrated medium or large effect sizes (Cohen's d) for group comparisons across the three language paradigms in both left and right temporal ROIs ( Fig. 3b and 3c), such that activation was lower in ASD than TD.
Correlations of neural response to speech with a toddler's social and communication abilities.
We further investigated correlations between a toddler's neural response to speech and his/her social and communication abilities. The results of mixed effects models showed signi cant correlations between fMRI activation and social and communication scores across subjects and language paradigms (left temporal ROI -communication scores: p = .021; left temporal ROI -social scores: p = .009; right temporal ROI -communication scores: p = .013; right temporal ROI -social scores: p = .002) (Supplementary Table 2; Supplementary Fig. 2).
Clustering results. We found ve fMRI-clinical phenotypically distinct clusters, including four TD and ASD clusters, and one small mixed cluster of 2 TD and 3 ASD (Fig. 4a). At the individual subject level, 90% of TD toddlers fell into two clusters, TD Cluster 1 and Cluster 2; and 80% of ASD toddlers also fell into two clusters, ASD Cluster 3 and Cluster 4. Within each of four TD and ASD clusters, toddlers share similar fMRI-clinical attributes ( Fig. 4b and 4c; for plots of all ve clusters, see Supplementary Fig. 3). Across Clusters 1 to 4, right temporal responses to Motherese ranged from very high to virtually absent and, in parallel social, communication, language, cognitive and symptom severity scores also neatly ranged from high to very low. The very highest Motherese neural responses occur in those with the highest social and cognitive abilities, and the very lowest Motherese neural responses occur in those with the very lowest social and cognitive abilities and most severe ASD social symptoms. The other prosodic language paradigms show slightly different patterns between clusters. Relative to Motherese, there is less distinction between the two TD clusters for Story language and less distinction between the two ASD clusters for Karen language. Thus, in the usual TD vs. ASD analysis that does not consider withindiagnosis heterogeneity, it is unsurprising that the TD vs ASD effect sizes were higher for the Story and Karen language than for the Motherese paradigm, which shows more differentiated activation between phenotypic subgroups.
Motherese eye-tracking results. As compared to TD toddlers, ASD toddlers showed substantially and signi cantly reduced percentage xation towards motherese, preferring the computer "techno" sounds ( Fig. 5a). We examined the association of the four neural-clinical clusters and gaze preference for motherese and found that toddlers in clusters predominantly including TD subjects, i.e., Clusters 1 and 2, showed signi cantly higher percentage xation towards motherese as compared to toddlers in clusters dominated by ASD subjects, i.e., Clusters 4 ( Fig. 5b and 5c; for all ve clusters, see Supplementary Fig. 4). ASD Cluster 3, with better cognitive and developmental skills, in turn, had higher percentage xation towards motherese than ASD Cluster 4 (66.6% versus 41.4%); there was a large effect size (Cohen's d = 1.01) in motherese xation between ASD toddlers in Cluster 3 and those in Cluster 4.

Discussion
Our within-subject fMRI, clinical and eye-tracking design provides unique evidence for the long-standing behavior-based theory that a toddler's increased neural responses may mediate behavioral responses to motherese utterances and lead to increased social and language abilities at young ages. Using datadriven network fusion and clustering approaches, we disentangled TD and ASD heterogeneity into four main subgroups and showed that toddlers with greater neural responses to motherese and other affective prosody stimuli have greater attention towards and preference for motherese utterances and better social and language skills than those with lesser neural responsiveness. This effect was evident across TD and ASD toddlers, indicating that social preference and language development are intertwined across a wide spectrum of social and language ability and disability. Further, clustering results suggest high response TD and low response ASD toddlers stand at opposite ends of the neural-motherese-social-language spectrum, and overall, these results indicate that the biology and behavior of TD as well as ASD are dimensional.
Our study also points to the early-age neural origins of the core social de cits that rst emerge in ASD infants and toddlers. As such, the absence of normal neural response to motherese and other affective language stimuli at the early age of clinical onset in ASD may be a biomarker of foundational dysregulation of social-emotional neural development that could underpin development of associated social cognitive and behavioral functions. Indeed, one of the rst early signs of ASD in babies and infants is a sharply reduced or absent behavioral response to mother's speech 22,23 . The present study, conducted in sleeping ASD infants, provides evidence for the early-age neural basis of reduced behavioral preferences for motherese utterances. It is robustly evident in the ASD Cluster 4 with reduced neural responses to motherese, reduced behavioral attention to motherese and poor social and language abilities, and, because the reduced neural response is observed during natural sleep, it cannot be attributed to attention, arousal, momentary distractions, or competing motivations. We literally image a near-failure of this neural cortical response in superior temporal regions. Such a near-failure in the awake ASD baby or infant will necessarily disconnect the foundational mother-infant loop that innately drives experience-expectant growth of social, language, and cognitive abilities in infants. This foundational dysfunction undermines not only the early experience-expectant mother-infant learning loop, but likely undermines later behavioral therapy efforts to socially engage ASD toddlers. Many early interventions that show success for some individuals hinge critically upon the idea of changing this attribute of early ASD development [24][25][26] . The hope is that early intervention will increase engagement between the child and the social world and enable experience-expectant neuroplasticity to divert a child towards more typical developmental trajectories.
Motherese is the tool that tunes the human-special mother-infant interactive loop and is thought to have a genetic basis that evolutionarily emerged in humans; motherese speech is the same in all humans and cultures [3][4][5] . As such, the neural, behavioral and experience-expectant responses in the infant are likely to also be genetically driven and evolutionarily emergent in human infants. Thus, the markedly de cits in neural and behavioral responses we identify in ASD toddlers are very likely to have a genetic basis, a hypothesis that should be pursued with brain-genetic studies. ASD is highly genetic with heritability of at least 81% 27 . It is a prenatal multistage, multi-process disorder that begins in the rst trimester with disruption of proliferation and neurogenesis and continues throughout second and third trimesters with disorder of neurite outgrowth, synaptogenesis and neural network function 28,29 . Social and language impairments are the consequences. Indeed, genetic dysregulation in ASD is highly correlated with ASD social symptoms in infants and toddlers 30 . We also identi ed gene co-expression networks in ASD that are associated with neural hypoactivation to language stimuli and with abnormal cortical growth in ASD toddlers who had very poor language development 31,32 . Identi ed genes and networks enrich languagerelevant, ASD-associated, human-speci c, and prenatal genes, including genes involved in cell proliferation and excitatory neuron development. These several studies indicate that prenatal genetic dysregulations in ASD are a cause for early-age impairment in social and language development, with reduced neural responses to motherese speech being a pivotal impediment to socially engaging with mother. It is of major importance that future work directly test this critical theory.
Understanding early-age clinical and neural ASD heterogeneity is a major challenge. Heterogeneity among typically developing toddlers is equally important to address but is often overlooked leading to weakened power to detect and characterize differences among ASD and TD toddlers. Here we demonstrate the power of the unbiased data-driven SNF/clustering method used for the rst time in the ASD eld to resolve the neural bases of early-age social and language heterogeneity in both ASD and TD and revealed four main distinct subgroups. We used this method to additionally resolve, in a purely datadriven approach, heterogeneity among TD. The relevance of the neural-clinical subgroups to the behavior of interest-attention to and preference for motherese utterances-was then quantitatively measured and validated using gaze contingent eye-tracking assessments of toddlers choosing to view a female telling a story in motherese or computer "techno" sounds and images. The results show that social and language ability and behavioral preference for motherese are linked to how strongly temporal cortex responds to motherese and other prosodic speech in toddlers across the neurodevelopmental spectrum from typical to language and social impaired.
Lastly, another exciting nding is that our purely unbiased data-driven SNF/clustering method appears to have replicated the same two ASD subgroups we previously identi ed using a subjective strati cation approach 31,33 in a new sample of toddlers using new paradigms. Previously, we arbitrarily strati ed ASD based on a child being above ("ASD Good") or below (ASD Poor") one standard deviation on the Mullen expressive and receptive language scales. The same two ASD subgroups emerged from the purely unbiased SNF/clustering method while also resolving TD toddler heterogeneity. Therefore, these may be robust and reliable ASD subgroups which are both etiologically and biologically meaningful. As such, future work should investigate how they may open early-age diagnostic, prognostic and/or treatment avenues for biomarker discovery.
In conclusion, in a one of a kind fMRI study of ASD, we resolve ASD and TD neural-social-languagebehavioral heterogeneity into four main discrete subgroups and show robust and systematic differences in how the ASD and TD toddler brain responds to varying levels of emotional prosodic speech including motherese. These neural differences underlie distinctions in behavioral preference for social-emotional motherese utterances and relate to social and communication developmental differences. These ndings support the longstanding behavior-based theory that motherese is an essential and innately driven stimulus that evokes neural responses driving infants to engage with mother in social and language learning. Neural responsiveness may lead to such learning, while sharply reduced or absent neural responsiveness may impede or preclude it. It suggests neural and behavioral de cits together may be a biomarker of foundational dysregulation of social-emotional neural development and learning. As such, different ASD neural-clinical-behavioral subgroups were identi ed that may bene t from different treatment approaches.  Table 3). Clinical testing occurred at the University of California, San Diego Autism Center of Excellence. Adult participants were recruited by word of mouth. This study was approved by the University of California, San Diego Institutional Review Board. Informed consent was obtained from parents or guardians of toddlers and from adult participants.
Clinical scores and fMRI scans were collected from 71 toddlers (41 ASD/30 TD). Scans were conducted during natural sleep, which has been proven to yield robust activation in ASD and TD toddlers [37][38][39][40] . Toddlers were considered TD if their diagnosis at outcome was TD and their Mullen Early Learning Composite scores fell within 2 standard deviations of the group mean. This allows us to examine activation patterns along a continuum of language and cognitive abilities in TD children. A subset of toddlers (4 ASD/6 TD) had test-retest fMRI scans collected at intervals ranging from 1 -15 months after the initial scan. fMRI scans were also obtained from 14 TD adults (6M/8F, 20 -37 years old).
Language paradigms. We presented three types of language stimuli with varying levels of emotional valence, including the "Story" language paradigm we used in earlier studies 31,33,38,41,42 (i.e., low emotional valence); a newly-created "Karen" language paradigm (i.e., moderate emotional valence); and a Motherese paradigm (i.e., high emotional valence). Stimuli were created by recording female voices reading nursery stories or age-appropriate phrases either using neutral or infant-directed utterances (e.g., Motherese 2,8 ). Language paradigms were presented in a block design (20s stimulus/20s rest).
The Story language paradigm has been used previously to identify ASD language-related brain signatures 38,39 and to develop predictors of language outcome among ASD toddlers 31,33 . It consists of three types of speech stimuli (simple forward speech, complex forward speech, and backward speech; 9 cycles of speech + rest; 6min 25s). Forward and backward speech stimuli were combined, and our main contrast of interest was all speech versus rest 31,33 .
To expand our understanding of neurofunctional language and affect development in ASD, we developed two new social orienting language paradigms. The Karen language paradigm utilizes 18 different nursery story stimuli (12min 5s) with moderate levels of emotion and prosody. The Motherese paradigm includes 12 phrases (8min 5s) recorded using high-pitched, intonational, lyrical and sing-songy speech characteristic of motherese.
Emotionality level testing. Two computer-based surveys were administered to TD adults to test emotionality levels of language paradigms.
Each fMRI paradigm consists of unique language segments, i.e., 2 Story language, 18 Karen language, and 12 Motherese segments. For survey 1, each unique segment was presented in random order (same order for each participant). Subjects were instructed to listen to each segment and respond using a Likert scale of 1 -5 with a rating of 1 indicating the least amount of emotionality and a rating of 5 indicating the most. TD adults (n=19) rated Story language as the least emotional, followed by Karen language, and Motherese segments as the most emotional ( Supplementary Fig. 5a). Survey 2 consisted of 18 trials, each containing a Story language segment, a Karen language segment, and a Motherese segment. Presenting all three stimulus types allowed for evaluation of differences in emotionality across all language paradigms. TD adults (n=15) then rated each segment using a Likert scale of 1 through 3 with a rating of 1 indicating the least amount of emotionality, 2 -some emotionality, and 3 -very emotional. Participants rated Story language as the least emotional, followed by Karen language, and Motherese segments as the most emotional ( Supplementary Fig. 5b). fMRI data acquisition. All fMRI data were collected in a 3T GE scanner at the University of California, San Diego Center for Functional MRI. Functional images were acquired with a multi-echo EPI protocol (echo times (TE) = 15ms, 28ms, 42ms, 56ms; TR = 2500ms; ip angle = 78°; matrix size = 64 × 64; slice thickness = 4 mm; eld of view (FOV) = 256 mm; 34 slices). Structural images were acquired using a T1weighted MPRAGE sequence (FOV = 256 mm; TE = 3.172ms; TR= 8.142ms; Flip angle = 12°).
Imaging data preprocessing. Functional data were preprocessed using the ME-ICA analysis pipeline "meica.py" 43,44 implemented in AFNI 45 and Python. First, the rst 4 volumes of each run were discarded to allow for magnetization to reach steady state. Next, motion correction parameters were calculated based on the rst TE images (TE = 15 ms) using a rigid-body alignment procedure. Slice timing correction was implemented for functional images of each TE, which then were normalized to an age-matched infant template 46 . The time series of four TEs were combined into a single time series 47 . Both principal and independent component analyses were applied to denoise the data through isolation of thermal (i.e., random) noise from structured signals (i.e., BOLD and non-BOLD signals) and separation of BOLD and non-BOLD signals. Only the BOLD-like components were retained in the preprocessed images, which were then spatially smoothed with a 8 mm FWHM Gaussian kernel.
Head motion was quanti ed via framewise displacement (FD) 48 . For adults and sleeping toddlers, head motion was minimal (mean FD < 0.1 mm). Group differences between ASD and TD toddlers in mean FD were only seen in the Motherese paradigm; there were no group differences between adults and toddler groups (Supplementary Table 4).
Whole-brain analyses. First-level and second-level whole-brain activation analyses were conducted with the general linear model (GLM) in SPM12 (https://www. l.ion.ucl.ac.uk/spm/software/spm12/). Events in rst-level models were based on the canonical hemodynamic response function and its temporal derivative. Given that scans were collected at multiple time points, including retest scans, we ran secondlevel whole-brain analyses with mixed effects models using 3dMVM program in AFNI: In mixed effects models, brain activation to each language paradigm (i.e., speech versus rest contrast) served as a dependent variable. Individual subjects were treated as a random effect, which allows for xed effects (i.e., age, gender, and mean FD) to vary for each subject.
Using a similar approach, we conducted whole-brain analyses with adult data for each language paradigm. However, only within-group tests were performed as all adult participants had typical development.
Resulting activation maps were corrected for multiple comparisons with the family-wise error (FWE) approach using 3dClustSim program in AFNI (voxel wise p = 0.005 and cluster size > 138 voxels for adults and cluster size > 186 voxels for toddlers). This spatial cluster correction took into account spatial autocorrelation by using the '-acf' option in 3dClustSim.
Calculation of percent signal changes in temporal language cortex. Two language-relevant ROIs from the meta-analytic activation map in Neurosynth (https://neurosynth.org/) with the term "language", including left and right temporal regions, were used for ROI analysis (Fig. 2). These ROIs were identical to those used in previous papers 31,33 . Given that a toddler template was used for toddler samples, ROIs were coregistered to the toddler template using FSL's irt function 49,50 . For each language paradigm, percent signal changes were calculated with rst-level models in speech versus rest contrast for all toddlers and adults.
Examination of Stability and Validation of fMRI Activation in Toddlers: Toddler-Adult, Test-Retest, and Sleep-Awake. Given the challenges relating to implementing sleep imaging with toddlers, test-retest is rarely examined, but essential for determining the rigor of this approach. Additional key questions surround the degree to which functional activation patterns vary along the dimension of sleep and awake states, and developmental periods such as between toddlers and adults or in individuals across time.
Here we took steps towards lling these gaps and tested: 1) whether brain activation to language stimuli in sleeping TD toddlers is similar to that in passively listening adults; 2) whether brain activation patterns are stable and reproducible in TD and ASD individuals across time.
These questions were addressed by comparing percent signal change values between TD adults and TD toddlers, and by computing intraclass correlation coe cients of brain activation within individuals who were scanned multiple times at intervals of 1-15 months, respectively.
Group differences in temporal cortex activation between ASD and TD. We compared percent signal change values between ASD and TD toddlers in a priori temporal ROIs relevant to language processing. Given the repeated time points, mixed effect models with lmer function (lme4 package) in R 51 were used, in which age, gender, and mean FD were included as xed effects while subjects served as a random effect.
Brain-behavior correlation analysis. Using similar mixed effects models as aforementioned, but with social or communication scores as a predictor of interest (age, gender, and mean FD as control variables/ xed effects, subjects as a random effect), we investigated the relevance of brain activation to a child's social and communication abilities assessed by the Vineland Adaptive Behavior Scales 36 .
Clustering analysis using Similarity Network Fusion. SNF is a novel approach for capturing heterogeneity in multiple types of patient data and forming clusters or subgroups. The method reduces noise by aggregating across multiple types of data, detects common and complementary signals from different types of data, and reveals the importance of each data type to patient similarity. Validation and clinical values emanate from follow-up analyses testing whether different clusters predict important "held-back" clinical variables of interest. In the original description, ve SNF clusters formed from multiple types of heterogenous genomic and genetic variables were associated with signi cantly different patient survival times 20 , a nding substantially superior to other single modality cluster approaches.
To identify clusters of clinical features of ASD and TD toddlers linked with patterns of fMRI activation to speech with varying emotionality levels, we used SNF 20 to integrate fMRI brain activation in three language paradigms and clinical measures and then used Louvain algorithm 52 to detect clusters of the similarity network. For this analysis, we included a subset of 52 of the 71 toddlers who had successful scans of all three language paradigms. The analysis was performed with 6 ROI variables (left and right temporal activation for each of the three language paradigms) and 14 clinical variables (i.e., 3 ADOS variables, 6 Vineland variables, and 5 Mullen variables) in R with SNFtool package. First, ROI and clinical data were normalized separately. Next, pairwise distance matrices between subjects were calculated for ROI or clinical data. A nity matrices (networks) were computed based on distance matrices. Each a nity matrix is equivalent to a similarity network where nodes are samples (e.g., subjects) and weighted edges represent pairwise sample similarities. Network fusion that iteratively updates every network was then performed, making two networks more similar to each other with every iteration. After a few iterations, two networks converged to a single network. We constructed the network with the strongest 15% connecting partners of each subject and ran the clustering analysis with the Louvain community algorithm. The clusters were visualized with Cytoscape 53 . Motherese eye-tracking task. We used a novel eye tracking Motherese task that utilized gaze contingent technology wherein a toddler's gaze activates what he/she sees and hears. Toddlers can choose to watch a movie depicting an actress telling a story using motherese speech or computer "techno" sounds and images. Motherese utterances used in this task were among those included in fMRI experiments. Fiftyfour toddlers had moderate or good eye-tracking performance and total looking time > 50% and were therefore included in the analysis. For 38 toddlers, eye-tracking data were collected prior to fMRI while 16 toddlers completed the task after fMRI.
Eye tracking was conducted using Tobii software (Tobii Studio; Tobii Pro Lab), and xation data were collected using a velocity threshold of 0.42 pixels/ms (Tobii Studio Tobii Fixation Filter) or 0.03 degrees/ms (Tobii Pro Lab Tobii IV-T Fixation Filter). Preference for Motherese was characterized by comparing percent xation duration within Motherese versus computer "Techno" sounds and images, we tested Motherese preferences between ASD and TD and across clusters identi ed by clustering analysis.