Efficacy of a digital intervention on visual scan pattern for faces in school-aged children with autism

Efficacy of a digital intervention on visual scan pattern for faces in school-aged children with autism | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Efficacy of a digital intervention on visual scan pattern for faces in school-aged children with autism Weihua Zhao, Lan Zhang, Qi Liu, Linghong Huang, Yanmiao Yang, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8631184/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Integrating scalable digital interventions into routine care may expand access to support and reduce the clinical burden for children with autism spectrum disorder (ASD). To evaluate this potential, we first used eye-tracking to identify distinct visual scanning patterns using a Hidden Markov Model approach between school-aged children with ASD and typically developing (TD) controls. This analysis revealed more random visual scan patterns and a reduced preference for social versus nonsocial stimuli in the ASD group. We then conducted a randomized clinical trial assessing a targeted, app-based emotion recognition intervention. Children with ASD were assigned to either two months of emotion recognition training(n = 25, 6.88 ± 1.31 years)or an active control (memory training, n = 25, 7.05 ± 1.54 years), with a six-month follow-up. Importantly, compared to the active control group, the emotion training group showed significantly greater improvements in clinical symptoms, cognitive performance, and eye-tracking measures of face processing, social preference, and joint attention. Mechanistically, the intervention promoted more normative visual scanning, including increased attention to eyes and a greater preference for social stimuli. These findings demonstrate that a targeted digital intervention can modify core visual-behavioral mechanisms and improve symptoms in ASD. This not only underscores the centrality of emotion processing in therapeutic design but also demonstrates the potential of scalable digital tools to augment care of autistic children. TRIAL REGISTRATION ClinicalTrial.gov Identifier: NCT06421272 Health sciences/Diseases Health sciences/Health care Biological sciences/Neuroscience Biological sciences/Psychology Social science/Psychology app-based digital intervention autism spectrum disorder face processing Hidden Markov Model eye-tracking Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction The global prevalence of autism spectrum disorder (ASD) has risen significantly, with recent estimates indicating approximately 1 in 31 children in the United States 1 and 1 in 143 children in China 2 . Although ASD symptoms vary widely, challenges in social functioning, such as joint attention, sustained eye contact, and recognition of facial expressions, remain common targets for intervention 3 . Behavioral and educational approaches are widely recommended and have shown promise in improving ASD-related symptoms across the lifespan 4 , 5 . However, these evidence-based interventions are often resource-intensive; for example, applied behavioral analysis (ABA) programs may recommend up to 20 hours per week for at least 2 years 6 , resulting in high costs and substantial time commitments. This intensity translates into high direct costs for families, averaging $ 24,869 annually per child in China 7 , and projects staggering societal costs, estimated to reach $ 11.5 trillion in the United States by 2029 if prevalence remains unchanged 8 . This substantial resource burden creates a major barrier to timely and accessible intervention. Consequently, there is a growing imperative for scalable solutions, driving healthcare providers to turn increasingly to digital strategies 9 – 12 . Digital interventions delivered via smartphones, wearable devices, and application-based platforms have emerged as efficient and scalable alternatives to address this critical need 13 . These tools are particularly promising for improving healthcare equity through increased accessibility 3 . Furthermore, compared to face-to-face traditional models, digital approaches can enhance patient engagement 14 , enable real-time progress monitoring, and allow for greater personalization 10 , 15 . These tools hold particular potential for improving healthcare equity by increasing accessibility to support services 16 . Large-scale meta-analyses of randomized controlled trials have demonstrated that digital interventions can effectively treat mental health conditions, including depression 17 – 19 , anxiety 20 , 21 , psychosis 9 and suicidal ideation 22 . More importantly, one study found that integrating ABA therapy with a wearable digital intervention, namely, Superpower Glass, improved facial engagement and emotion recognition more than ABA alone 13 . Similarly, computer-based executive function training produced improvements in attention-deficit/hyperactivity disorder (ADHD) symptoms comparable to conventional training, with some evidence that children with milder symptoms may benefit even more from the digital format 23 . Further supporting these findings, recent meta-analyses studies indicated that gamified digital health interventions can enhance emotional, social 24 , executive and motor skills in autistic children and adolescents 25 , 26 . Taken together, these studies highlight the considerable potential of home-based digital therapy not only to supplement standard care but also to reduce economic burden and save time for families. However, it is important to note that the overall number of high-quality trials using digital intervention in ASD remains limited, highlighting a need for further rigorous research 27 , 28 . Therefore, this study aimed to: (1) evaluate the efficacy of a tablet app-based digital intervention for school-aged children with ASD aged from 4 to 9 years old; (2) compare five autism-sensitive objective eye-tracking measures, including paradigms for face emotion processing, dynamic social preference, joint attention, social bonding, and restricted and repetitive behavior between children with ASD and TD controls; and (3) assess the digital intervention’s specific impact on these eye-tracking outcomes. To address these aims, we recruited 52 children with ASD along with age-matched TD controls, and employed Hidden Markov Models 29 , 30 to analyze the effects of a 2-month digital intervention, comparing an experimental group receiving emotion recognition training with a control group receiving memory training. Finally, we discuss the intervention’s potential scalability, providing clearer evidence regarding its benefits, limitations, and practical applicability. Results Participants A total of 52 eligible autistic children aged from 4 to 9 years (7 girls, mean ± SD = 6.98 ± 1.46) were included in the study. Baseline visual scan patterns (cognitive style) in the ASD group were first compared with a sample of 14 typical developing children (2 girls, mean ± SD = 6.68 ± 1.24). Age (t = 0.88, p = 0.38) and gender (χ 2 = 0.01, p = 0.94) are matched. All participants completed five eye-tracking paradigms to assess cognitive style differences between the ASD and TD groups. Children with ASD were randomly assigned to one of two app-based intervention groups: emotion recognition training or memory training (see Fig. 1 A). Study procedure see Fig. 1 B. Eye-movement visual scan pattern between TD and ASD groups To explore whether there were differences in the visual scan patterns between ASD and TD individuals, a data-driven eye movement analysis was performed using Hidden Markov Models (hMM) by calculating the log-likelihood of eye-movement under the ASD hMM and under the TD hMM. T-tests were conducted twice: once using the data from individuals with ASD and once using data from TD individuals. The two hMMs were regarded as significantly different if two p values were less than 0.05. The results indicated significantly different visual scanning strategies between TD and ASD groups across all five eye-tracking paradigms (all ps 0.24, Fig. 2 A). Specifically: (1) In the face emotion processing paradigm, the ASD group exhibited diffuse scanning of the entire face, whereas the TD group focused significantly more on the core facial features (eyes, nose, and mouth). (2) In the dynamic social preference paradigm, the ASD group showed a preference for viewing non-social stimuli, in contrast to the TD group, who preferred viewing social stimuli more. (3) In the joint attention paradigm, the ASD group focused primarily on the characters' faces, while the TD group attended more to the interactive space between the characters and the target object. (4) In the social bonding paradigm, the ASD group attended more to the child alone and toys, whereas the TD group focused more on the parent-child interaction. (5) In the restricted and repetitive behaviors paradigm, the ASD group again focused more on spinning non-social stimuli, contrasting with the TD group's focus on spinning social stimuli (Fig. 2 B). The effects of app-based interventions in ASD Two children in the ASD group did not complete the post-intervention tasks and so the final analysis included a sample of 50 children (Emotion group: n = 25, 6.88 ± 1.31 years; Memory group: n = 25, 7.05 ± 1.54 years). Pre-intervention assessments confirmed no significant differences between the two groups in age, sex, or scores on clinical assessments and questionnaires (ps > 0.07, uncorrected, see Table 1 ). The average duration of training was similar between the two app-based interventions (Emotion group: 65.68 ± 10.60 days; Memory group: 65.24 ± 14.79 days), and this difference was not significant (t = 0.12, p = 0.90). Following training, the memory group showed improvements in autistic symptoms, as measured by the SRS (p = 0.04, effect size = 0.50), RBS (p = 0.0012, effect size = 0.59), and both the imitation (p = 0.05, effect size = 0.41) and affective expression ༈p = 0.04, effect size = 0.39༉subscales of the PEP-3, though these results did not survive Bonferroni multiple comparison correction. More importantly, the emotion recognition training led to significant improvements in core autistic symptoms. These included the ADOS-2 comparison score༈p = 0.05, effect size = 0.40༉, the SRS (surviving correction, p = 0.017, effect size = 0.50), and both the imitation༈p = 0.02, effect size = 0.47༉ and affective expression༈surviving correction, p = 0.004, effect size = 0.61༉ subscales of the PEP-3. Notably, both training regimens showed a trend toward improving Wechsler IQ scores. The improvement was statistically significant in the emotion group (p = 0.03, effect size = 0.48) but marginal in the memory group (p = 0.06, effect size = 0.49, more details see Table 2 ). This suggests that the emotion recognition training may have a broader impact, influencing not only its target domain but also general cognitive performance. Table 1 Demographic and baseline measures Baseline Measures Emotion(n = 25) Memory(n = 25) Statistic P value Age 6.88 ± 1.31 7.05 ± 1.54 0.42 0.67 Sex (female) 25(4) 25(3) 0.17 a 0.68 Training days 65.68 ± 10.60 65.24 ± 14.79 0.12 0.90 ADOS-2 Total 14.84 ± 4.36 13.72 ± 3.40 1.01 0.32 SA score 11.48 ± 4.25 10.76 ± 2.88 0.70 0.49 RRB score 3.36 ± 1.32 2.96 ± 1.59 0.97 0.34 CS score 7.00 ± 1.22 6.44 ± 0.87 1.86 0.07 SRS-2 Total 99.20 ± 21.68 95.36 ± 21.47 0.63 0.53 RBS-R 21.00 ± 14.46 22.24 ± 10.67 0.35 0.73 SCQ 16.64 ± 5.10 17.44 ± 6.83 0.47 0.64 CSQ 16.33 ± 4.56 15.27 ± 4.76 0.81 0.42 Wechsler IQ score 67.78 ± 19.04 67.44 ± 9.71 0.08 0.94 PEP-3 imitation 15.48 ± 3.44 15.84 ± 2.98 0.40 0.69 PEP-3 affective expression 12.32 ± 3.33 13.24 ± 2.74 1.07 0.29 SCAS 20.00 ± 10.42 21.80 ± 11.60 0.58 0.57 Table 2 Post-treatment effects on assessments and questionnaires Memory Intervention Pre-intervention Post-intervention Statistic P value Primary outcomes ADOS-2 Total 13.67 ± 3.46 13.04 ± 3.03 1.21 0.24 SA score 10.67 ± 2.89 10.13 ± 2.94 1.23 0.23 RRB score 3.00 ± 1.62 2.92 ± 1.56 0.34 0.74 CS score 6.42 ± 0.88 6.08 ± 0.97 1.62 0.19 SRS-2 Total 95.36 ± 21.47 89.76 ± 25.50 2.16 0.04 + Secondary outcomes RBS-R 22.24 ± 10.67 18.20 ± 12.22 2.71 0.012 + SCQ 17.62 ± 7.24 15.71 ± 7.81 1.30 0.21 CSQ 15.27 ± 4.76 14.23 ± 3.91 1.29 0.21 Wechsler IQ score 67.38 ± 10.53 70.67 ± 12.62 1.98 0.06 PEP-3 imitation 15.84 ± 2.98 16.84 ± 2.93 2.06 0.05 + PEP-3 affective expression 13.24 ± 2.74 14.60 ± 3.55 2.17 0.04 + SCAS 21.80 ± 11.60 20.32 ± 10.42 1.25 0.22 Emotion Intervention Primary outcomes ADOS-2 Total 14.84 ± 4.36 14.04 ± 3.83 1.73 0.096 SA score 11.48 ± 4.25 10.88 ± 3.47 1.52 0.14 RRB score 3.36 ± 1.32 3.16 ± 1.07 0.74 0.47 CS score 7.00 ± 1.22 6.56 ± 1.29 2.03 0.05 + SRS-2 Total 99.20 ± 21.68 91.80 ± 20.45 2.57 0.017 * Secondary outcomes RBS-R 21.00 ± 14.46 19.56 ± 13.63 0.70 0.49 SCQ 16.11 ± 5.07 16.94 ± 6.58 0.82 0.42 CSQ 16.33 ± 4.56 15.15 ± 4.62 1.70 0.10 Wechsler IQ score 67.78 ± 19.04 70.70 ± 17.75 2.30 0.03 + PEP-3 imitation 15.48 ± 3.44 16.96 ± 3.14 2.43 0.02 + PEP-3 affective expression 12.32 ± 3.33 14.08 ± 2.94 3.23 0.004 * SCAS 20.00 ± 10.42 19.88 ± 9.22 0.08 0.94 To evaluate the intervention effects, we compared post-intervention eye-tracking data between the emotion and memory training groups. Differences in hMMs were specific to three paradigms: the face emotion processing (p Memory = 0.03, p Emotion =0.0002), dynamic social (p Memory = 0.002, p Emotion =0.01), and joint attention paradigms (p Memory = 0.01, p Emotion =0.026, effect size were reported in Fig. 3 A). By analyzing gaze scan patterns within predefined regions of interest (ROIs), we found that the emotion intervention altered visual attention toward more socially relevant stimuli. Specifically, in the face emotion paradigm, the emotion group focused more on the eyes (11.24%) than the memory group, who focused more on the nose (12.95%). Similarly, in the dynamic social paradigm, the emotion group showed a greater focus on social stimuli (36.90%) compared to the memory group's focus on non-social stimuli (40.18%). No such differences were observed in the joint attention paradigm, where both groups focused equally on the target objects (emotion: 33.06%, memory: 32.88%, see Fig. 3 B). Additionally, a correlation between the mean log-likelihood (mLL) for face processing paradigm and ADOS-2 total scores revealed a significant negative relationship in the memory group (r = -0.613, p = 0.001) but not in the emotion group (r = 0.036, p = 0.863). Fisher's z-test confirmed that these two correlation coefficients were significantly different from each other (z = -2.46, p = 0.01; Fig. 4 ). Six-month follow-up post training intervention Six months post-intervention the caregivers of the children in the trial were asked to complete the SRS-2 and SCQ questionnaires again as a follow up. A total of 21 children in the memory group and 19 in the face emotion recognition group had this assessment completed. Table 3 shows that for this reduced sub-group of participants there was no overall significant difference across pre- post- and follow up scores indicating that further improvements in SRS-2 scores post-treatment were not found. Table 3 Follow-up (6 months) treatment effects on questionnaires Follow-up effects T1-pre T2-post T3-follow up Statistic P value Memory(N = 21) SRS-2 Total 97.14 ± 20.88 94.00 ± 22.09 90.90 ± 28.17 1.21 0.31 SCQ 17.62 ± 7.24 15.71 ± 7.81 15.33 ± 7.50 2.13 0.16 Emotion (N = 19) SRS-2 Total 97.89 ± 22.54 92.00 ± 19.82 97.53 ± 23.89 2.73 0.08 SCQ 16.11 ± 5.07 16.94 ± 6.58 16.67 ± 6.40 0.54 0.47 SRS-2: Social Responsivity Scale-2 score. SCQ: Social communication questionnaire Discussion The current study first revealed differences in visual scan patterns between school aged individuals with ASD and TD controls across five autism-sensitive paradigms (face processing, dynamic social preference, joint attention, social bonding, and RRB). Following this, we evaluated a 2-month, home-administered digital intervention. Participants receiving emotion recognition training showed significantly greater improvement than those receiving memory training, both in clinical symptoms (ADOS-2 comparison scores) and cognitive performance (IQ scores) and on three key eye-tracking paradigms: face processing, dynamic social preference, and joint attention. Mechanistically, the emotion-based intervention promoted more normative visual scanning, including increased attention to the eyes and a preference for social stimuli. The memory intervention group, conversely, showed a reduction of probability in a specific scan pattern linked to severe symptoms, an effect not observed in the emotion group. Collectively, a targeted digital intervention can modify visual scanning behavior and improve symptoms in ASD, highlighting the central role of emotion processing in therapeutic design. The spontaneous ability to attend to the social overtures and activities of others is essential for the development of social communication. Individuals with ASD, for whom core symptoms involve early and persistent difficulties in social interaction and communication, often exhibit challenges in establishing eye contact, processing facial information, and interpreting others’ intentions. Consequently, researchers have been particularly interested in how individuals with ASD orient to and visually explore faces, and whether they utilize gaze information 39 . Eye-tracking measurements of social visual engagement have shown robust performance in early diagnosis of autistic children compared to typically developing 40 , 41 . However, findings do not support a generalized deficit in social orienting and face engagement using eye-tracking 42 , as these patterns appear modulated by factors such as age 43 and context 44 . To better understand the nuanced nature of these social difficulties, researchers are increasingly integrating advanced data acquisition and analytical methods 42 . In the present study, applying a Hidden Markov Model approach across five autism-sensitive paradigms, we observed that the ASD children exhibited: diffuse scanning of the entire face rather than focusing on key regions; a preference for non-social over social stimuli; and attention directed primarily to characters' faces rather than to interactive behavior 45 . These findings indicate a distinct visual scanning pattern for social information in ASD compared to TD individuals. To investigate whether digital intervention could modify the distinct visual attention patterns observed in ASD, we conducted a 2-month parallel-design clinical trial. Participants with ASD were randomly assigned to either an emotion recognition training program, which focused on time spent identifying emotions, or an active control condition involving memory training games. Compared to memory training, the emotion-based intervention led to greater improvements in core clinical symptoms and cognitive performance, specifically in ADOS-2 comparison scores and IQ. Both groups showed improvement in SRS and PEP-3 scores. Given established difficulties in facial emotion recognition in ASD 46 , such training represents a promising intervention target 3 . Notably, while IQ was not a primary intervention target, IQ scores showed an improving trend in both groups, reaching significance only in the emotion-based group. Since higher intelligence is associated with better social cognition in ASD 47 , this indirect effect suggests that emotion-based training may offer broader benefits for cognitive and clinical profiles. Furthermore, only the improvement in the affective subscale of the PEP-3 48 survived Bonferroni correction. This result indicates that emotion-focused digital intervention could function as an effective psychoeducational tool for enhancing affective function in school-aged children with ASD 49 . Furthermore, analysis of eye-tracking data revealed that, compared to the memory training group, participants in the emotion-based intervention demonstrated significantly increased attention to the eye region and a stronger preference for social over non-social stimuli. This result is consistent with effects observed in other interventions, such as oxytocin administration 32 . This shift was observed consistently across several key paradigms: face processing, dynamic social preference, and joint attention paradigms. Importantly, in the memory training group, a significant negative correlation was observed between the mean log-likelihood for the face processing paradigm and ADOS-2 total scores, implying that a lesser probability of a certain gaze pattern was linked to greater symptom severity. This correlation was absent in the emotion training group. This dissociation suggests that the experimental intervention may have fundamentally altered the functional significance of gaze behavior, thereby decoupling this visual pattern from core clinical severity. Furthermore, qualitative feedback from caregivers underscored a critical practical benefit for the current digital intervention study. They reported consistently high levels of motivation and engagement from their children throughout the training. This sustained adherence highlights a significant advantage for home-based implementation, suggesting that such interventions can be both effective and feasible, potentially increasing accessibility and long-term compliance 13 . Given the established heterogeneity of autism 50 , 51 , future research should employ larger samples to identify ASD subtypes most responsive to this emotion-based digital intervention, using eye-tracking measures for stratification 52 . Additionally, as the combination of ABA and digital intervention has shown benefits for socialization 13 , integrating the current emotion-based training with ABA principles could enhance its efficacy and promote lasting effects. Finally, extending the intervention period (e.g., to six months) would help evaluate long-term efficacy, and the home-based format facilitates accessibility and adherence. This is of particular importance since our preliminary 6-month follow-up assessment suggested that improvements in symptoms seen immediately after a 2-month intervention were not obviously maintained and that longer or combination treatments might be needed to improve longer-term outcomes. In summary, this study first identified a distinct visual scanning pattern during social paradigms in children with ASD compared to TD. It further demonstrated that an emotion recognition-based digital intervention can improve clinical symptoms, cognitive performance, and adaptive visual attention patterns more effectively than active control training. These findings underscore the potential of home-based digital interventions for school-aged children with ASD, offering a scalable approach that maintains engagement and motivation. Methods Trial Design This randomized clinical trial was conducted to compare the effects of a self-help app-based emotion recognition intervention(see Supplementary Figure S1 A ) with a control app-based memory intervention༈ Figure S1 B , developed by Infinite Brain Technology, Beijing, China༉ in individuals with ASD. The study included assessments at baseline (study entry), post-treatment (2 months after baseline), and a 6-month follow-up. Additionally, baseline data from participants with ASD were compared with those of TD children to assess potential differences in visual scan pattern (cognitive style) between the two groups. The study received ethical approval from the Institutional Review Board at the University of Electronic Science and Technology. Data were collected between May 2024 and May 2025, with written informed consent obtained from all participants’ caregivers. This report adheres to the Consolidated Standards of Reporting Trials (CONSORT) guidelines (Fig. 1 A). The trial was also pre-registered: ClinicalTrial.gov Identifier: NCT06421272. Participants The study included 52 autistic children aged 4–9 years (7 girls; mean age ± SD = 6.98 ± 1.46). Baseline visual scan patterns (cognitive style) were compared with a sample of 14 TD children (2 girls; mean age ± SD = 6.68 ± 1.24). The groups were matched on age (t = 0.88, p = 0.38) and gender (χ² = 0.01, p = 0.94). All participants completed five eye-tracking paradigms to assess cognitive style differences. Subsequently, the children with ASD were randomly assigned to one of two app-based intervention groups: emotion recognition training (n = 25, 6.88 ± 1.31 years)or an active control (memory training, n = 25, 7.05 ± 1.54 years). The study was conducted at Chengdu Maternal and Children’s Central Hospital (CMCCH) in compliance with Good Clinical Practice guidelines and the Declaration of Helsinki. Eligible children, recruited from outpatient clinics at CMCCH, met the following criteria: (1) DSM-5 ASD diagnosis confirmed by Autism Diagnostic Observation Schedule-2 31 administration from experienced clinicians, (2) age 4–10 years, and (3) had prior experience with iPads or mobile phones. Exclusion criteria included genetic/chromosomal abnormalities (e.g., Fragile X or Rett syndrome), neurological disorders (e.g., epilepsy, cerebral palsy), psychiatric disorders other than ASD, concurrent psychotropic medication use, or severe sensory/respiratory impairments. Of 56 screened children, 52 met eligibility (see Fig. 1 A). Assessments and outcomes For the trial protocol see Fig. 1 B. Following informed consent by a caregiver, eligible children visited CMCCH for recording pre-intervention outcome measures including assessment, questionnaires and eye-tracking paradigms. Primary outcome measures were the gold-standard clinical assessment using ADOS-2 scores (Comparison and Total scores) and the SRS-2 total score in line with previous trials 32 and conducted by trained research reliable individuals. The comparison ADOS-2 score was included since modules 1, 2 and 3 were used, although each individual was always assessed using the same ADOS-2 module. Secondary outcomes were assessed using five autism-sensitive eye-tracking paradigms focused on social affect and restricted and repetitive behaviors. These paradigms detailed in Figure S2 , were as follows: (1) Face emotion processing paradigm: Participants were required to observe happy, angry, fearful or neutral static facial expressions, respectively (in total 16 trials; 2s each and 4 models including adult man, adult female, boy and girl); (2) Dynamic social preference paradigm: Participants viewed simultaneous pairs of social (e.g. people dancing) alongside non-social geometric videos (20s each, 2 social and 2 non-social video clips); (3) Joint attention paradigm: Participants watched videos depicting individuals using eye-gaze or a combination of eye-gaze and finger pointing to direct their attention towards target objects (8 trials; 5s each; object location balanced); (4) Social bonding paradigm: Participants observed videos of parent-child social interactions presented alongside static images of toys (8 trials; 5s each; stimulus location balanced); (5) Restricted and repetitive social vs. non-social (RRS) paradigm: Participants viewed concurrent videos of stereotypical social behaviors (e.g., a child spinning) and repetitive non-social motions (e.g., a spinning fan) (8 trials; 5s each; location balanced) ( see Figures S2 ). All stimuli were counterbalanced and non-overlapping sets were used for pre- and post-intervention assessments. During these eye-tracking paradigms, younger participants sat comfortably on a parent’s lap, while older participants sat independently. All were free to explore and gaze at the display screen based on their interest. Eye‑gaze data were collected using a Tobii TX300 system (Tobii, Danderyd, Sweden) with a sampling rate of 300 Hz and a gaze accuracy of 0.4°. Recording, stimulus presentation, and analysis were performed using Tobii Pro Studio, E‑Prime 2.0, and E‑Prime Extensions for Tobii (Psychology Software Tools, Pittsburgh, PA). Two matched stimulus sets were presented in random order. Other secondary outcome measures included more detailed specific assessments of social communication using Social Communication Questionnaire (SCQ) 33 , repetitive behaviors via the Repetitive Behavior Scale-Revised (RBS-R) 34 , psychoeducational profile (PEP-3, third edition) 35 , the Spence Children’s Anxiety Scale (SCAS) 36 and the Wechsler Intelligence Scale 37 . Finally, the strain experienced by caregivers themselves was assessed using the caregiver strain questionnaire (CSQ) 38 . Statistical Analysis A data-driven eye movement analysis was performed using Hidden Markov Models (hMM) implemented with the EMHMM toolbox ( http://visal.cs.cityu.edu.hk/research/emhmm/ ) in Matlab. For each participant, an individual hMM was estimated from the fixation data using a variational Bayesian algorithm. This approach automatically selected the optimal number of regions of interest (ROIs, where K = 1, 2, or 3) and estimated all hyperparameters by identifying the model with the highest log-likelihood (LL) for the given data. The best model (i.e., the one with the highest LL) was selected separately for the ASD and TD groups using the baseline data across all five paradigms. The average difference in log-likelihoods serves as an estimate of the Kullback-Leibler (KL) divergence, a measure of dissimilarity between probability distributions. A KL divergence of zero indicates that the two hMMs represent identical distributions. Furthermore, to identify intervention-specific visual patterns, the effects of the two app-based interventions were compared using hMMs within predefined ROIs. Additionally, Pearson’s correlation was used to assess the relationship between the mean log-likelihood (mLL) of specific visual scan patterns and clinical symptoms (ADOS-2 scores). The resulting correlation coefficients were compared using Fisher’s z-transformation. Here, a lower mLL value indicates a lower probability of the corresponding visual scan pattern. Declarations Competing Interests The authors declare no competing interest. Funding This work was supported by the Humanity and Social Science Foundation of Ministry of Education of China [grant number 24YJC190046-WHZ] and National Natural Science Foundation of China (NSFC) [grant number 82301732- JL]. Author Contribution WZ: Conceptualization, Methodology, Formal analysis, Investigation, Data Curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.LZ: Resources, Writing – review & editing.QL: Methodology, Writing – original draft.LH: Data Curation, Investigation.YY: Formal analysis.DX: Data Curation.MZ: Data Curation.WY: Resources.WS: Supervision.JL: Supervision, Writing – review & editing.KMK: Conceptualization, Supervision, Writing – review & editing. Acknowledgement The authors would like to thank Infinite Brain Technology for their technological support, including software maintenance and user support. Code availability Data were analyzed using hMM tools available via http://visal.cs.cityu.edu.hk/research/emhmm/ . Ethics and Declarations All written informed consent was provided by participants’ parents or legal guardians. The trial was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board at the University of Electronic Science and Technology (approval number #1061420220115007). It is registered at ClinicalTrials.gov under trial ID NCT06421272 (Registration Date: May 2024). References Shaw, K. A. Prevalence and Early Identification of Autism Spectrum Disorder Among Children Aged 4 and 8 Years — Autism and Developmental Disabilities Monitoring Network, 16 Sites, United States, 2022. MMWR Surveill. Summ. 74, (2025). Jiang, X., Chen, X., Su, J. & Liu, N. Prevalence of autism spectrum disorder in mainland china over the past 6 years: a systematic review and meta-analysis. BMC Psychiatry 24, 404 (2024). Zhang, Q. et al. Facial emotion training as an intervention in autism spectrum disorder: A meta-analysis of randomized controlled trials. Autism Res. 14, 2169–2182 (2021). Gosling, C. J. et al. Efficacy of psychosocial interventions for Autism spectrum disorder: an umbrella review. Mol. Psychiatry 27, 3647–3656 (2022). Sandbank, M. et al. Project AIM: Autism Intervention Meta-Analysis for Studies of Young Children. Psychol. Bull. 146, 1–29 (2020). Dawson, G. Early behavioral intervention, brain plasticity, and the prevention of autism spectrum disorder. Dev. Psychopathol. 20, 775–803 (2008). Zhao, Y., Luo, Y., Zhang, R. & Zheng, X. Direct and indirect costs for families of children with autism spectrum disorder in China. Autism 27, 2337–2347 (2023). Cakir, J., Frye, R. E. & Walker, S. J. The lifetime social cost of autism: 1990–2029. Res. Autism Spectr. Disord. 72, 101502 (2020). Bucci, S. et al. Actissist: Proof-of-Concept Trial of a Theory-Driven Digital Intervention for Psychosis. Schizophr. Bull. 44, 1070–1080 (2018). Michie, S., Yardley, L., West, R., Patrick, K. & Greaves, F. Developing and Evaluating Digital Interventions to Promote Behavior Change in Health and Health Care: Recommendations Resulting From an International Workshop. J. Med. Internet Res. 19, e232 (2017). Riadi, I. et al. Digital interventions for depression and anxiety in older adults: a systematic review of randomised controlled trials. Lancet Healthy Longev. 3, e558–e571 (2022). Shah, A., Kraemer, K. R., Won, C. R., Black, S. & Hasenbein, W. Developing Digital Intervention Games for Mental Disorders: A Review. Games Health J. 7, 213–224 (2018). Voss, C. et al. Effect of Wearable Digital Intervention for Improving Socialization in Children With Autism Spectrum Disorder: A Randomized Clinical Trial. JAMA Pediatr. 173, 446–454 (2019). Nahum-Shani, I., Shaw, S. D., Carpenter, S. M., Murphy, S. A. & Yoon, C. Engagement in digital interventions. Am. Psychol. 77, 836–852 (2022). Gega, L. et al. Digital interventions in mental health: evidence syntheses and economic modelling. Health Technol. Assess. 26, 1–182 (2022). Chatterjee, A., Prinz, A., Gerdes, M. & Martinez, S. Digital Interventions on Healthy Lifestyle Management: Systematic Review. J. Med. Internet Res. 23, e26931 (2021). Firth, J. et al. The efficacy of smartphone-based mental health interventions for depressive symptoms: a meta-analysis of randomized controlled trials. World Psychiatry 16, 287–298 (2017). Moshe, I. et al. Digital interventions for the treatment of depression: A meta-analytic review. Psychol. Bull. 147, 749–786 (2021). Osborn, T. L. et al. Single-session digital intervention for adolescent depression, anxiety, and well-being: Outcomes of a randomized controlled trial with Kenyan adolescents. J. Consult. Clin. Psychol. 88, 657–668 (2020). Firth, J. et al. Can smartphone mental health interventions reduce symptoms of anxiety? A meta-analysis of randomized controlled trials. J. Affect. Disord. 218, 15–22 (2017). Pauley, D., Cuijpers, P., Papola, D., Miguel, C. & Karyotaki, E. Two decades of digital interventions for anxiety disorders: a systematic review and meta-analysis of treatment effectiveness. Psychol. Med. 53, 567–579 (2023). Torok, M. et al. Suicide prevention using self-guided digital interventions: a systematic review and meta-analysis of randomised controlled trials. Lancet Digit. Health 2, e25–e36 (2020). Wu, Y. et al. Computer-based multiple component cognitive training in children with ADHD: a pilot study. Child Adolesc. Psychiatry Ment. Health 17, 9 (2023). Xu, F. et al. The Use of Digital Interventions for Children and Adolescents with Autism Spectrum Disorder—A Meta-Analysis. J. Autism Dev. Disord. https://doi.org/10.1007/s10803-024-06563-4 (2024) doi:10.1007/s10803-024-06563-4. Liu, Y. et al. Efficacy of gamified digital health interventions for children and adolescents with autism spectrum disorder: a systematic review and meta-analysis. Child Adolesc. Psychiatry Ment. Health https://doi.org/10.1186/s13034-025-01009-w (2025) doi:10.1186/s13034-025-01009-w. Wang, T. et al. Digital interventions for autism spectrum disorders: A systematic review and meta-analysis. Pediatr. Investig. 8, 224–236 (2024). Mukherjee, D. et al. Digital tools for direct assessment of autism risk during early childhood: A systematic review. Autism 28, 6–31 (2024). Sandgreen, H., Frederiksen, L. H. & Bilenberg, N. Digital Interventions for Autism Spectrum Disorder: A Meta-analysis. J. Autism Dev. Disord. 51, 3138–3152 (2021). Bottos, S. & Balasingam, B. Tracking the Progression of Reading Using Eye-Gaze Point Measurements and Hidden Markov Models. IEEE Trans. Instrum. Meas. 69, 7857–7868 (2020). Chuk, T., Chan, A. B., Shimojo, S. & Hsiao, J. H. Eye movement analysis with switching hidden Markov models. Behav. Res. Methods 52, 1026–1043 (2020). Lord, C. et al. Autism Diagnostic Observation Schedule. 2nd ed. West. Psychol. Corp. (2012). Le, J. et al. Infrequent Intranasal Oxytocin Followed by Positive Social Interaction Improves Symptoms in Autistic Children: A Pilot Randomized Clinical Trial. Psychother. Psychosom. 91, 335–347 (2022). Chandler, S. et al. Validation of the Social Communication Questionnaire in a Population Cohort of Children With Autism Spectrum Disorders. J. Am. Acad. Child Adolesc. Psychiatry 46, 1324–1332 (2007). Mirenda, P. et al. Validating the Repetitive Behavior Scale-Revised in Young Children with Autism Spectrum Disorder. J. Autism Dev. Disord. 40, 1521–1530 (2010). Fulton, M. L. & D’Entremont, B. Utility of the Psychoeducational Profile-3 for Assessing Cognitive and Language Skills of Children with Autism Spectrum Disorders. J. Autism Dev. Disord. 43, 2460–2471 (2013). Magiati, I. et al. The measurement properties of the spence children’s anxiety scale-parent version in a large international pooled sample of young people with autism spectrum disorder. Autism Res. 10, 1629–1652 (2017). Sadeghi, A., Rabiee, M. & Abedi, M. R. Validation and reliability of the Wechsler Intelligence Scale for Children-IV. Dev. Psychol. J. Iran. Psychol. 7, 377–386 (2011). Khanna, R. et al. Psychometric properties of the Caregiver Strain Questionnaire (CGSQ) among caregivers of children with autism. Autism 16, 179–199 (2012). Chawarska, K., Macari, S. & Shic, F. Decreased Spontaneous Attention to Social Scenes in 6-Month-Old Infants Later Diagnosed with Autism Spectrum Disorders. Biol. Psychiatry 74, 195–203 (2013). Jones, W. et al. Eye-Tracking–Based Measurement of Social Visual Engagement Compared With Expert Clinical Diagnosis of Autism. JAMA 330, 854–865 (2023). Kou, J. et al. Comparison of three different eye-tracking tasks for distinguishing autistic from typically developing children and autistic symptom severity. Autism Res. 12, 1529–1540 (2019). Guillon, Q., Hadjikhani, N., Baduel, S. & Rogé, B. Visual social attention in autism spectrum disorder: Insights from eye tracking studies. Neurosci. Biobehav. Rev. 42, 279–297 (2014). Fujioka, T. et al. Developmental changes in attention to social information from childhood to adolescence in autism spectrum disorders: a comparative study. Mol. Autism 11, 24 (2020). Kaliukhovich, D. A. et al. Social attention to activities in children and adults with autism spectrum disorder: effects of context and age. Mol. Autism 11, 79 (2020). Robain, F. et al. The impact of social complexity on the visual exploration of others’ actions in preschoolers with autism spectrum disorder. BMC Psychol. 9, 50 (2021). Black, M. H. et al. Mechanisms of facial emotion recognition in autism spectrum disorders: Insights from eye tracking and electroencephalography. Neurosci. Biobehav. Rev. 80, 488–515 (2017). Hirosawa, T. et al. Different associations between intelligence and social cognition in children with and without autism spectrum disorders. PLOS ONE 15, e0235380 (2020). Portoghese, C. et al. The usefulness of the Revised Psychoeducational Profile for the assessment of preschool children with pervasive developmental disorders. Autism 13, 179–191 (2009). Davis, K. S., Kennedy, S. A., Dallavecchia, A., Skolasky, R. L. & Gordon, B. Psychoeducational Interventions for Adults With Level 3 Autism Spectrum Disorder: A 50-Year Systematic Review. Cogn. Behav. Neurol. 32, 139 (2019). Moore, A. et al. The geometric preference subtype in ASD: identifying a consistent, early-emerging phenomenon through eye tracking. Mol. Autism 9, 19 (2018). Zhao, W. et al. A clustering approach identifies an Autism Spectrum Disorder subtype more responsive to chronic oxytocin treatment. Transl. Psychiatry 14, 312 (2024). Parellada, M. et al. In Search of Biomarkers to Guide Interventions in Autism Spectrum Disorder: A Systematic Review. Am. J. Psychiatry https://doi.org/10.1176/appi.ajp.21100992 (2023) doi:10.1176/appi.ajp.21100992. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterial.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8631184","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":583237662,"identity":"c2fe6df7-d6b0-45e7-bbfe-814cbc447da0","order_by":0,"name":"Weihua Zhao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4klEQVRIiWNgGAWjYLACxgYGBn447wCxWiQbYKqJ1mJwgFgt8u1nH378ucNOzvhG8uHPH2oY5PhuJDB+LsBnQU+6sYTkmWRjsxtpaRIHjjEYS95IYJaegUcLM0Mag4RhG3Piths5ZgwHGxgSN9xIYGPmwaOFjf8Z84/EtvrEzTPyP38AaqknqIVHIo1N4mDb4cQNEjkMEkAtCQaEtEhIPGOzbGw7bixx5pmZxJljEoYzzzxslsanRb4/jfnmz7ZqOf725McfKmps5PmOJx/8jE8Lhq0MkGgaBaNgFIyCUUARAADWJkvekUlnbgAAAABJRU5ErkJggg==","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":true,"prefix":"","firstName":"Weihua","middleName":"","lastName":"Zhao","suffix":""},{"id":583237663,"identity":"872a3b70-7a5b-49d0-8e13-b5b12255d531","order_by":1,"name":"Lan Zhang","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Lan","middleName":"","lastName":"Zhang","suffix":""},{"id":583237664,"identity":"806d8ad7-b721-40ce-bb55-1241eff4d21d","order_by":2,"name":"Qi Liu","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Qi","middleName":"","lastName":"Liu","suffix":""},{"id":583237665,"identity":"a85b08df-f2b5-46e3-bf3a-1dd415c0098b","order_by":3,"name":"Linghong Huang","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Linghong","middleName":"","lastName":"Huang","suffix":""},{"id":583237666,"identity":"0023947f-04ba-4d5f-b842-ac293071c3ef","order_by":4,"name":"Yanmiao Yang","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Yanmiao","middleName":"","lastName":"Yang","suffix":""},{"id":583237667,"identity":"0330d1e9-e249-4fff-b949-764b18aad83f","order_by":5,"name":"Dan Xu","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Dan","middleName":"","lastName":"Xu","suffix":""},{"id":583237668,"identity":"35515e4a-ac88-4c37-8a35-cd4167c4fc41","order_by":6,"name":"Menghan Zhou","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Menghan","middleName":"","lastName":"Zhou","suffix":""},{"id":583237669,"identity":"b2271a56-6c42-40e3-ba38-fbac64988bc8","order_by":7,"name":"Wenxu Yang","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Wenxu","middleName":"","lastName":"Yang","suffix":""},{"id":583237670,"identity":"b35753b2-a52d-4ec8-a8f9-5d318ad4b44f","order_by":8,"name":"Wei Sun","email":"","orcid":"","institution":"Beijing Infinite Brain Technologies","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Sun","suffix":""},{"id":583237671,"identity":"06be9ceb-a1ea-4d77-b074-a6ffd89d05aa","order_by":9,"name":"Jiao Le","email":"","orcid":"","institution":"Chengdu University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jiao","middleName":"","lastName":"Le","suffix":""},{"id":583237672,"identity":"6a64c01d-1e65-4422-90a9-6138d7601747","order_by":10,"name":"Keith M. Kendrick","email":"","orcid":"","institution":"University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Keith","middleName":"M.","lastName":"Kendrick","suffix":""}],"badges":[],"createdAt":"2026-01-18 12:23:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8631184/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8631184/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102745600,"identity":"a1dd306a-170e-4ba3-a583-d3950c2eb552","added_by":"auto","created_at":"2026-02-16 08:52:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":80251,"visible":true,"origin":"","legend":"\u003cp\u003eConsort flow diagram (A) and trial protocol (B). ADOS-2: Autism Diagnostic Observation Schedule–2. SRS-2: Social Responsivity Scale-2. RBS-R: Repetitive behavior scale – revised. SCQ: Social communication quotient. CSQ: caregiver strain questionnaire. IQ: Wechsler Intelligence Quotient. PEP-3: Psychoeducational Profile, Third Edition. SCAS: Spence Children’s Anxiety Scale.\u003c/p\u003e","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/4c076179619d84191e22c143.png"},{"id":102427080,"identity":"c7315189-e4ef-465c-abcb-2f97f7bd95e3","added_by":"auto","created_at":"2026-02-11 14:46:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":121758,"visible":true,"origin":"","legend":"\u003cp\u003eDifferences in scan patterns between ASD and TD groups across five eye-tracking paradigms. (A) Effects size (Cohen’s d) for the log-likelihood of eye-movement comparing the fit of the ASD-specific and TD-specific hidden Markov models (hMMs) for each eye-tracking paradigm. (B) Representative visualizations of the scan patterns for each group and paradigm. Within regions of interest (ROIs), the red arrow indicates the ROI with the highest probability of fixations. ASD, autism spectrum disorder; TD, typical developing; hMM, hidden Markov Models; Face, face emotion processing paradigm; Dynamic, dynamic social preference paradigm; Attention, joint attention paradigm; Bonding, social bonding paradigm; RRS, restricted and repetitive social paradigm. *, p \u0026lt;0.05; **, p\u0026lt;0.01; ***, p\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/3eb888e2a1a7f5b0f6aae131.png"},{"id":102745427,"identity":"f445bf3b-35cf-465f-a036-2ede73bc4e87","added_by":"auto","created_at":"2026-02-16 08:49:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":127147,"visible":true,"origin":"","legend":"\u003cp\u003eThe effect of app-based interventions on visual scan patterns. (A) Effects size (Cohen’s d) for the log-likelihood of eye-movement comparing the fit of the memory-specific and emotion-specific hidden Markov models (hMMs) for each eye-tracking paradigm. (B) Visualizations for the memory and emotion intervention groups on paradigms showing significant differences. Within predefined regions of interest (ROIs), the red arrow indicates the ROI with the highest probability of fixations. ASD, autism spectrum disorder; TD, typical developing; hMM, hidden Markov Models; Face, face emotion processing paradigm; Dynamic, dynamic social preference paradigm; Attention, joint attention paradigm. *, p \u0026lt;0.05; **, p\u0026lt;0.01; ***, p\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/07051c4182f431f67cb349bb.png"},{"id":102427076,"identity":"ab1184b9-34dc-4588-ad0a-3a597da1c362","added_by":"auto","created_at":"2026-02-11 14:46:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":25732,"visible":true,"origin":"","legend":"\u003cp\u003eAssociations between the mean log-likelihood (mLL) of face emotion processing and clinical symptom severity assessed by ADOS-2 across two intervention groups. The correlation coefficients were compared by Fisher’s Z. *, p \u0026lt;0.05.\u003c/p\u003e","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/61ee0967bda1890ed0de6f2f.png"},{"id":104781462,"identity":"fa432ab6-eb4e-4f6b-9b04-6a043d780992","added_by":"auto","created_at":"2026-03-17 07:55:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1446358,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/a80283e9-81a0-401b-a4e7-54b6972b62c5.pdf"},{"id":102427079,"identity":"1ab607ea-e3c4-4eeb-9379-dacf0ca4542a","added_by":"auto","created_at":"2026-02-11 14:46:40","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1270530,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-8631184/v1/f4909f882982f3c988162189.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Efficacy of a digital intervention on visual scan pattern for faces in school-aged children with autism","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe global prevalence of autism spectrum disorder (ASD) has risen significantly, with recent estimates indicating approximately 1 in 31 children in the United States \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e and 1 in 143 children in China \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Although ASD symptoms vary widely, challenges in social functioning, such as joint attention, sustained eye contact, and recognition of facial expressions, remain common targets for intervention \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Behavioral and educational approaches are widely recommended and have shown promise in improving ASD-related symptoms across the lifespan \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. However, these evidence-based interventions are often resource-intensive; for example, applied behavioral analysis (ABA) programs may recommend up to 20 hours per week for at least 2 years \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, resulting in high costs and substantial time commitments. This intensity translates into high direct costs for families, averaging \u003cspan\u003e$\u003c/span\u003e24,869 annually per child in China \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, and projects staggering societal costs, estimated to reach \u003cspan\u003e$\u003c/span\u003e11.5 trillion in the United States by 2029 if prevalence remains unchanged \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This substantial resource burden creates a major barrier to timely and accessible intervention. Consequently, there is a growing imperative for scalable solutions, driving healthcare providers to turn increasingly to digital strategies \u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e–\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDigital interventions delivered via smartphones, wearable devices, and application-based platforms have emerged as efficient and scalable alternatives to address this critical need \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. These tools are particularly promising for improving healthcare equity through increased accessibility \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Furthermore, compared to face-to-face traditional models, digital approaches can enhance patient engagement \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, enable real-time progress monitoring, and allow for greater personalization \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. These tools hold particular potential for improving healthcare equity by increasing accessibility to support services \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eLarge-scale meta-analyses of randomized controlled trials have demonstrated that digital interventions can effectively treat mental health conditions, including depression \u003csup\u003e\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e–\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, anxiety \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, psychosis \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e and suicidal ideation \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. More importantly, one study found that integrating ABA therapy with a wearable digital intervention, namely, Superpower Glass, improved facial engagement and emotion recognition more than ABA alone \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Similarly, computer-based executive function training produced improvements in attention-deficit/hyperactivity disorder (ADHD) symptoms comparable to conventional training, with some evidence that children with milder symptoms may benefit even more from the digital format \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Further supporting these findings, recent meta-analyses studies indicated that gamified digital health interventions can enhance emotional, social \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, executive and motor skills in autistic children and adolescents \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Taken together, these studies highlight the considerable potential of home-based digital therapy not only to supplement standard care but also to reduce economic burden and save time for families. However, it is important to note that the overall number of high-quality trials using digital intervention in ASD remains limited, highlighting a need for further rigorous research \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTherefore, this study aimed to: (1) evaluate the efficacy of a tablet app-based digital intervention for school-aged children with ASD aged from 4 to 9 years old; (2) compare five autism-sensitive objective eye-tracking measures, including paradigms for face emotion processing, dynamic social preference, joint attention, social bonding, and restricted and repetitive behavior between children with ASD and TD controls; and (3) assess the digital intervention’s specific impact on these eye-tracking outcomes. To address these aims, we recruited 52 children with ASD along with age-matched TD controls, and employed Hidden Markov Models \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e to analyze the effects of a 2-month digital intervention, comparing an experimental group receiving emotion recognition training with a control group receiving memory training. Finally, we discuss the intervention’s potential scalability, providing clearer evidence regarding its benefits, limitations, and practical applicability.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eA total of 52 eligible autistic children aged from 4 to 9 years (7 girls, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u0026thinsp;=\u0026thinsp;6.98\u0026thinsp;\u0026plusmn;\u0026thinsp;1.46) were included in the study. Baseline visual scan patterns (cognitive style) in the ASD group were first compared with a sample of 14 typical developing children (2 girls, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u0026thinsp;=\u0026thinsp;6.68\u0026thinsp;\u0026plusmn;\u0026thinsp;1.24). Age (t\u0026thinsp;=\u0026thinsp;0.88, p\u0026thinsp;=\u0026thinsp;0.38) and gender (χ\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.01, p\u0026thinsp;=\u0026thinsp;0.94) are matched. All participants completed five eye-tracking paradigms to assess cognitive style differences between the ASD and TD groups. Children with ASD were randomly assigned to one of two app-based intervention groups: emotion recognition training or memory training (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Study procedure see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEye-movement visual scan pattern between TD and ASD groups\u003c/h3\u003e\n\u003cp\u003eTo explore whether there were differences in the visual scan patterns between ASD and TD individuals, a data-driven eye movement analysis was performed using Hidden Markov Models (hMM) by calculating the log-likelihood of eye-movement under the ASD hMM and under the TD hMM. T-tests were conducted twice: once using the data from individuals with ASD and once using data from TD individuals. The two hMMs were regarded as significantly different if two p values were less than 0.05. The results indicated significantly different visual scanning strategies between TD and ASD groups across all five eye-tracking paradigms (all ps\u0026thinsp;\u0026lt;\u0026thinsp;0.04, Cohen\u0026rsquo;s d\u0026thinsp;\u0026gt;\u0026thinsp;0.24, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Specifically: (1) In the face emotion processing paradigm, the ASD group exhibited diffuse scanning of the entire face, whereas the TD group focused significantly more on the core facial features (eyes, nose, and mouth). (2) In the dynamic social preference paradigm, the ASD group showed a preference for viewing non-social stimuli, in contrast to the TD group, who preferred viewing social stimuli more. (3) In the joint attention paradigm, the ASD group focused primarily on the characters' faces, while the TD group attended more to the interactive space between the characters and the target object. (4) In the social bonding paradigm, the ASD group attended more to the child alone and toys, whereas the TD group focused more on the parent-child interaction. (5) In the restricted and repetitive behaviors paradigm, the ASD group again focused more on spinning non-social stimuli, contrasting with the TD group's focus on spinning social stimuli (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eThe effects of app-based interventions in ASD\u003c/h3\u003e\n\u003cp\u003eTwo children in the ASD group did not complete the post-intervention tasks and so the final analysis included a sample of 50 children (Emotion group: n\u0026thinsp;=\u0026thinsp;25, 6.88\u0026thinsp;\u0026plusmn;\u0026thinsp;1.31 years; Memory group: n\u0026thinsp;=\u0026thinsp;25, 7.05\u0026thinsp;\u0026plusmn;\u0026thinsp;1.54 years). Pre-intervention assessments confirmed no significant differences between the two groups in age, sex, or scores on clinical assessments and questionnaires (ps\u0026thinsp;\u0026gt;\u0026thinsp;0.07, uncorrected, see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The average duration of training was similar between the two app-based interventions (Emotion group: 65.68\u0026thinsp;\u0026plusmn;\u0026thinsp;10.60 days; Memory group: 65.24\u0026thinsp;\u0026plusmn;\u0026thinsp;14.79 days), and this difference was not significant (t\u0026thinsp;=\u0026thinsp;0.12, p\u0026thinsp;=\u0026thinsp;0.90). Following training, the memory group showed improvements in autistic symptoms, as measured by the SRS (p\u0026thinsp;=\u0026thinsp;0.04, effect size\u0026thinsp;=\u0026thinsp;0.50), RBS (p\u0026thinsp;=\u0026thinsp;0.0012, effect size\u0026thinsp;=\u0026thinsp;0.59), and both the imitation (p\u0026thinsp;=\u0026thinsp;0.05, effect size\u0026thinsp;=\u0026thinsp;0.41) and affective expression ༈p\u0026thinsp;=\u0026thinsp;0.04, effect size\u0026thinsp;=\u0026thinsp;0.39༉subscales of the PEP-3, though these results did not survive Bonferroni multiple comparison correction. More importantly, the emotion recognition training led to significant improvements in core autistic symptoms. These included the ADOS-2 comparison score༈p\u0026thinsp;=\u0026thinsp;0.05, effect size\u0026thinsp;=\u0026thinsp;0.40༉, the SRS (surviving correction, p\u0026thinsp;=\u0026thinsp;0.017, effect size\u0026thinsp;=\u0026thinsp;0.50), and both the imitation༈p\u0026thinsp;=\u0026thinsp;0.02, effect size\u0026thinsp;=\u0026thinsp;0.47༉ and affective expression༈surviving correction, p\u0026thinsp;=\u0026thinsp;0.004, effect size\u0026thinsp;=\u0026thinsp;0.61༉ subscales of the PEP-3. Notably, both training regimens showed a trend toward improving Wechsler IQ scores. The improvement was statistically significant in the emotion group (p\u0026thinsp;=\u0026thinsp;0.03, effect size\u0026thinsp;=\u0026thinsp;0.48) but marginal in the memory group (p\u0026thinsp;=\u0026thinsp;0.06, effect size\u0026thinsp;=\u0026thinsp;0.49, more details see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). This suggests that the emotion recognition training may have a broader impact, influencing not only its target domain but also general cognitive performance.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographic and baseline measures\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline Measures\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEmotion(n\u0026thinsp;=\u0026thinsp;25)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMemory(n\u0026thinsp;=\u0026thinsp;25)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.88\u0026thinsp;\u0026plusmn;\u0026thinsp;1.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.05\u0026thinsp;\u0026plusmn;\u0026thinsp;1.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (female)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25(4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25(3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.17\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.68\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTraining days\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e65.68\u0026thinsp;\u0026plusmn;\u0026thinsp;10.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e65.24\u0026thinsp;\u0026plusmn;\u0026thinsp;14.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.90\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eADOS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14.84\u0026thinsp;\u0026plusmn;\u0026thinsp;4.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.72\u0026thinsp;\u0026plusmn;\u0026thinsp;3.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSA score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.48\u0026thinsp;\u0026plusmn;\u0026thinsp;4.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10.76\u0026thinsp;\u0026plusmn;\u0026thinsp;2.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.49\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRRB score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.36\u0026thinsp;\u0026plusmn;\u0026thinsp;1.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.96\u0026thinsp;\u0026plusmn;\u0026thinsp;1.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.34\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.00\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e99.20\u0026thinsp;\u0026plusmn;\u0026thinsp;21.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95.36\u0026thinsp;\u0026plusmn;\u0026thinsp;21.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.53\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRBS-R\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21.00\u0026thinsp;\u0026plusmn;\u0026thinsp;14.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22.24\u0026thinsp;\u0026plusmn;\u0026thinsp;10.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16.64\u0026thinsp;\u0026plusmn;\u0026thinsp;5.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17.44\u0026thinsp;\u0026plusmn;\u0026thinsp;6.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.64\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCSQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16.33\u0026thinsp;\u0026plusmn;\u0026thinsp;4.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15.27\u0026thinsp;\u0026plusmn;\u0026thinsp;4.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWechsler IQ score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67.78\u0026thinsp;\u0026plusmn;\u0026thinsp;19.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e67.44\u0026thinsp;\u0026plusmn;\u0026thinsp;9.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 imitation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15.48\u0026thinsp;\u0026plusmn;\u0026thinsp;3.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15.84\u0026thinsp;\u0026plusmn;\u0026thinsp;2.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 affective expression\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.32\u0026thinsp;\u0026plusmn;\u0026thinsp;3.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.24\u0026thinsp;\u0026plusmn;\u0026thinsp;2.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.00\u0026thinsp;\u0026plusmn;\u0026thinsp;10.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21.80\u0026thinsp;\u0026plusmn;\u0026thinsp;11.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.57\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePost-treatment effects on assessments and questionnaires\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMemory Intervention\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePre-intervention\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePost-intervention\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrimary outcomes\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eADOS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e13.67\u0026thinsp;\u0026plusmn;\u0026thinsp;3.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e13.04\u0026thinsp;\u0026plusmn;\u0026thinsp;3.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSA score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e10.67\u0026thinsp;\u0026plusmn;\u0026thinsp;2.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e10.13\u0026thinsp;\u0026plusmn;\u0026thinsp;2.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.23\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRRB score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e3.00\u0026thinsp;\u0026plusmn;\u0026thinsp;1.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e2.92\u0026thinsp;\u0026plusmn;\u0026thinsp;1.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e6.42\u0026thinsp;\u0026plusmn;\u0026thinsp;0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e6.08\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e95.36\u0026thinsp;\u0026plusmn;\u0026thinsp;21.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e89.76\u0026thinsp;\u0026plusmn;\u0026thinsp;25.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.04\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSecondary outcomes\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRBS-R\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e22.24\u0026thinsp;\u0026plusmn;\u0026thinsp;10.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e18.20\u0026thinsp;\u0026plusmn;\u0026thinsp;12.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.012\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e17.62\u0026thinsp;\u0026plusmn;\u0026thinsp;7.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e15.71\u0026thinsp;\u0026plusmn;\u0026thinsp;7.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCSQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e15.27\u0026thinsp;\u0026plusmn;\u0026thinsp;4.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e14.23\u0026thinsp;\u0026plusmn;\u0026thinsp;3.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWechsler IQ score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e67.38\u0026thinsp;\u0026plusmn;\u0026thinsp;10.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e70.67\u0026thinsp;\u0026plusmn;\u0026thinsp;12.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 imitation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e15.84\u0026thinsp;\u0026plusmn;\u0026thinsp;2.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e16.84\u0026thinsp;\u0026plusmn;\u0026thinsp;2.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.05\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 affective expression\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e13.24\u0026thinsp;\u0026plusmn;\u0026thinsp;2.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e14.60\u0026thinsp;\u0026plusmn;\u0026thinsp;3.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.04\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e21.80\u0026thinsp;\u0026plusmn;\u0026thinsp;11.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e20.32\u0026thinsp;\u0026plusmn;\u0026thinsp;10.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEmotion Intervention\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePrimary outcomes\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eADOS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e14.84\u0026thinsp;\u0026plusmn;\u0026thinsp;4.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e14.04\u0026thinsp;\u0026plusmn;\u0026thinsp;3.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.096\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSA score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e11.48\u0026thinsp;\u0026plusmn;\u0026thinsp;4.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e10.88\u0026thinsp;\u0026plusmn;\u0026thinsp;3.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRRB score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e3.36\u0026thinsp;\u0026plusmn;\u0026thinsp;1.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e3.16\u0026thinsp;\u0026plusmn;\u0026thinsp;1.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e7.00\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e6.56\u0026thinsp;\u0026plusmn;\u0026thinsp;1.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.05\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e99.20\u0026thinsp;\u0026plusmn;\u0026thinsp;21.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e91.80\u0026thinsp;\u0026plusmn;\u0026thinsp;20.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.017\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSecondary outcomes\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRBS-R\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e21.00\u0026thinsp;\u0026plusmn;\u0026thinsp;14.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e19.56\u0026thinsp;\u0026plusmn;\u0026thinsp;13.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.49\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e16.11\u0026thinsp;\u0026plusmn;\u0026thinsp;5.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e16.94\u0026thinsp;\u0026plusmn;\u0026thinsp;6.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCSQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e16.33\u0026thinsp;\u0026plusmn;\u0026thinsp;4.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e15.15\u0026thinsp;\u0026plusmn;\u0026thinsp;4.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWechsler IQ score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e67.78\u0026thinsp;\u0026plusmn;\u0026thinsp;19.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e70.70\u0026thinsp;\u0026plusmn;\u0026thinsp;17.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.03\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 imitation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e15.48\u0026thinsp;\u0026plusmn;\u0026thinsp;3.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e16.96\u0026thinsp;\u0026plusmn;\u0026thinsp;3.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.02\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEP-3 affective expression\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e12.32\u0026thinsp;\u0026plusmn;\u0026thinsp;3.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e14.08\u0026thinsp;\u0026plusmn;\u0026thinsp;2.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.004\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e20.00\u0026thinsp;\u0026plusmn;\u0026thinsp;10.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e19.88\u0026thinsp;\u0026plusmn;\u0026thinsp;9.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo evaluate the intervention effects, we compared post-intervention eye-tracking data between the emotion and memory training groups. Differences in hMMs were specific to three paradigms: the face emotion processing (p\u003csub\u003eMemory\u003c/sub\u003e = 0.03, p\u003csub\u003eEmotion\u003c/sub\u003e =0.0002), dynamic social (p\u003csub\u003eMemory\u003c/sub\u003e = 0.002, p\u003csub\u003eEmotion\u003c/sub\u003e =0.01), and joint attention paradigms (p\u003csub\u003eMemory\u003c/sub\u003e = 0.01, p\u003csub\u003eEmotion\u003c/sub\u003e =0.026, effect size were reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). By analyzing gaze scan patterns within predefined regions of interest (ROIs), we found that the emotion intervention altered visual attention toward more socially relevant stimuli. Specifically, in the face emotion paradigm, the emotion group focused more on the eyes (11.24%) than the memory group, who focused more on the nose (12.95%). Similarly, in the dynamic social paradigm, the emotion group showed a greater focus on social stimuli (36.90%) compared to the memory group's focus on non-social stimuli (40.18%). No such differences were observed in the joint attention paradigm, where both groups focused equally on the target objects (emotion: 33.06%, memory: 32.88%, see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Additionally, a correlation between the mean log-likelihood (mLL) for face processing paradigm and ADOS-2 total scores revealed a significant negative relationship in the memory group (r = -0.613, p\u0026thinsp;=\u0026thinsp;0.001) but not in the emotion group (r\u0026thinsp;=\u0026thinsp;0.036, p\u0026thinsp;=\u0026thinsp;0.863). Fisher's z-test confirmed that these two correlation coefficients were significantly different from each other (z = -2.46, p\u0026thinsp;=\u0026thinsp;0.01; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eSix-month follow-up post training intervention\u003c/h3\u003e\n\u003cp\u003eSix months post-intervention the caregivers of the children in the trial were asked to complete the SRS-2 and SCQ questionnaires again as a follow up. A total of 21 children in the memory group and 19 in the face emotion recognition group had this assessment completed. Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows that for this reduced sub-group of participants there was no overall significant difference across pre- post- and follow up scores indicating that further improvements in SRS-2 scores post-treatment were not found.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eFollow-up (6 months) treatment effects on questionnaires\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFollow-up effects\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT1-pre\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eT2-post\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eT3-follow up\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMemory(N\u0026thinsp;=\u0026thinsp;21)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e97.14\u0026thinsp;\u0026plusmn;\u0026thinsp;20.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e94.00\u0026thinsp;\u0026plusmn;\u0026thinsp;22.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e90.90\u0026thinsp;\u0026plusmn;\u0026thinsp;28.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.31\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e17.62\u0026thinsp;\u0026plusmn;\u0026thinsp;7.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e15.71\u0026thinsp;\u0026plusmn;\u0026thinsp;7.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e15.33\u0026thinsp;\u0026plusmn;\u0026thinsp;7.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEmotion (N\u0026thinsp;=\u0026thinsp;19)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRS-2 Total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e97.89\u0026thinsp;\u0026plusmn;\u0026thinsp;22.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e92.00\u0026thinsp;\u0026plusmn;\u0026thinsp;19.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e97.53\u0026thinsp;\u0026plusmn;\u0026thinsp;23.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSCQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e16.11\u0026thinsp;\u0026plusmn;\u0026thinsp;5.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e16.94\u0026thinsp;\u0026plusmn;\u0026thinsp;6.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e16.67\u0026thinsp;\u0026plusmn;\u0026thinsp;6.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eSRS-2: Social Responsivity Scale-2 score. SCQ: Social communication questionnaire\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe current study first revealed differences in visual scan patterns between school aged individuals with ASD and TD controls across five autism-sensitive paradigms (face processing, dynamic social preference, joint attention, social bonding, and RRB). Following this, we evaluated a 2-month, home-administered digital intervention. Participants receiving emotion recognition training showed significantly greater improvement than those receiving memory training, both in clinical symptoms (ADOS-2 comparison scores) and cognitive performance (IQ scores) and on three key eye-tracking paradigms: face processing, dynamic social preference, and joint attention. Mechanistically, the emotion-based intervention promoted more normative visual scanning, including increased attention to the eyes and a preference for social stimuli. The memory intervention group, conversely, showed a reduction of probability in a specific scan pattern linked to severe symptoms, an effect not observed in the emotion group. Collectively, a targeted digital intervention can modify visual scanning behavior and improve symptoms in ASD, highlighting the central role of emotion processing in therapeutic design.\u003c/p\u003e \u003cp\u003eThe spontaneous ability to attend to the social overtures and activities of others is essential for the development of social communication. Individuals with ASD, for whom core symptoms involve early and persistent difficulties in social interaction and communication, often exhibit challenges in establishing eye contact, processing facial information, and interpreting others’ intentions. Consequently, researchers have been particularly interested in how individuals with ASD orient to and visually explore faces, and whether they utilize gaze information \u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Eye-tracking measurements of social visual engagement have shown robust performance in early diagnosis of autistic children compared to typically developing \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. However, findings do not support a generalized deficit in social orienting and face engagement using eye-tracking \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, as these patterns appear modulated by factors such as age \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e and context \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. To better understand the nuanced nature of these social difficulties, researchers are increasingly integrating advanced data acquisition and analytical methods \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. In the present study, applying a Hidden Markov Model approach across five autism-sensitive paradigms, we observed that the ASD children exhibited: diffuse scanning of the entire face rather than focusing on key regions; a preference for non-social over social stimuli; and attention directed primarily to characters' faces rather than to interactive behavior \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. These findings indicate a distinct visual scanning pattern for social information in ASD compared to TD individuals.\u003c/p\u003e \u003cp\u003eTo investigate whether digital intervention could modify the distinct visual attention patterns observed in ASD, we conducted a 2-month parallel-design clinical trial. Participants with ASD were randomly assigned to either an emotion recognition training program, which focused on time spent identifying emotions, or an active control condition involving memory training games. Compared to memory training, the emotion-based intervention led to greater improvements in core clinical symptoms and cognitive performance, specifically in ADOS-2 comparison scores and IQ. Both groups showed improvement in SRS and PEP-3 scores. Given established difficulties in facial emotion recognition in ASD \u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, such training represents a promising intervention target \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Notably, while IQ was not a primary intervention target, IQ scores showed an improving trend in both groups, reaching significance only in the emotion-based group. Since higher intelligence is associated with better social cognition in ASD \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, this indirect effect suggests that emotion-based training may offer broader benefits for cognitive and clinical profiles. Furthermore, only the improvement in the affective subscale of the PEP-3 \u003csup\u003e48\u003c/sup\u003e survived Bonferroni correction. This result indicates that emotion-focused digital intervention could function as an effective psychoeducational tool for enhancing affective function in school-aged children with ASD \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFurthermore, analysis of eye-tracking data revealed that, compared to the memory training group, participants in the emotion-based intervention demonstrated significantly increased attention to the eye region and a stronger preference for social over non-social stimuli. This result is consistent with effects observed in other interventions, such as oxytocin administration \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. This shift was observed consistently across several key paradigms: face processing, dynamic social preference, and joint attention paradigms. Importantly, in the memory training group, a significant negative correlation was observed between the mean log-likelihood for the face processing paradigm and ADOS-2 total scores, implying that a lesser probability of a certain gaze pattern was linked to greater symptom severity. This correlation was absent in the emotion training group. This dissociation suggests that the experimental intervention may have fundamentally altered the functional significance of gaze behavior, thereby decoupling this visual pattern from core clinical severity. Furthermore, qualitative feedback from caregivers underscored a critical practical benefit for the current digital intervention study. They reported consistently high levels of motivation and engagement from their children throughout the training. This sustained adherence highlights a significant advantage for home-based implementation, suggesting that such interventions can be both effective and feasible, potentially increasing accessibility and long-term compliance \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eGiven the established heterogeneity of autism \u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, future research should employ larger samples to identify ASD subtypes most responsive to this emotion-based digital intervention, using eye-tracking measures for stratification \u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Additionally, as the combination of ABA and digital intervention has shown benefits for socialization \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, integrating the current emotion-based training with ABA principles could enhance its efficacy and promote lasting effects. Finally, extending the intervention period (e.g., to six months) would help evaluate long-term efficacy, and the home-based format facilitates accessibility and adherence. This is of particular importance since our preliminary 6-month follow-up assessment suggested that improvements in symptoms seen immediately after a 2-month intervention were not obviously maintained and that longer or combination treatments might be needed to improve longer-term outcomes.\u003c/p\u003e \u003cp\u003eIn summary, this study first identified a distinct visual scanning pattern during social paradigms in children with ASD compared to TD. It further demonstrated that an emotion recognition-based digital intervention can improve clinical symptoms, cognitive performance, and adaptive visual attention patterns more effectively than active control training. These findings underscore the potential of home-based digital interventions for school-aged children with ASD, offering a scalable approach that maintains engagement and motivation.\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eTrial Design\u003c/h2\u003e\u003cp\u003eThis randomized clinical trial was conducted to compare the effects of a self-help app-based emotion recognition intervention(see Supplementary \u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eA\u003c/b\u003e) with a control app-based memory intervention༈\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eB\u003c/b\u003e, developed by Infinite Brain Technology, Beijing, China༉ in individuals with ASD. The study included assessments at baseline (study entry), post-treatment (2 months after baseline), and a 6-month follow-up. Additionally, baseline data from participants with ASD were compared with those of TD children to assess potential differences in visual scan pattern (cognitive style) between the two groups. The study received ethical approval from the Institutional Review Board at the University of Electronic Science and Technology. Data were collected between May 2024 and May 2025, with written informed consent obtained from all participants’ caregivers. This report adheres to the Consolidated Standards of Reporting Trials (CONSORT) guidelines (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). The trial was also pre-registered: ClinicalTrial.gov Identifier: NCT06421272.\u003c/p\u003e\u003ch3\u003eParticipants\u003c/h3\u003e\u003cp\u003eThe study included 52 autistic children aged 4–9 years (7 girls; mean age ± SD = 6.98 ± 1.46). Baseline visual scan patterns (cognitive style) were compared with a sample of 14 TD children (2 girls; mean age ± SD = 6.68 ± 1.24). The groups were matched on age (t = 0.88, p = 0.38) and gender (χ² = 0.01, p = 0.94). All participants completed five eye-tracking paradigms to assess cognitive style differences. Subsequently, the children with ASD were randomly assigned to one of two app-based intervention groups: emotion recognition training (n = 25, 6.88 ± 1.31 years)or an active control (memory training, n = 25, 7.05 ± 1.54 years). The study was conducted at Chengdu Maternal and Children’s Central Hospital (CMCCH) in compliance with Good Clinical Practice guidelines and the Declaration of Helsinki.\u003c/p\u003e\u003cp\u003eEligible children, recruited from outpatient clinics at CMCCH, met the following criteria: (1) DSM-5 ASD diagnosis confirmed by Autism Diagnostic Observation Schedule-2 \u003csup\u003e31\u003c/sup\u003e administration from experienced clinicians, (2) age 4–10 years, and (3) had prior experience with iPads or mobile phones. Exclusion criteria included genetic/chromosomal abnormalities (e.g., Fragile X or Rett syndrome), neurological disorders (e.g., epilepsy, cerebral palsy), psychiatric disorders other than ASD, concurrent psychotropic medication use, or severe sensory/respiratory impairments. Of 56 screened children, 52 met eligibility (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA).\u003c/p\u003e\u003ch2\u003eAssessments and outcomes\u003c/h2\u003e\u003cp\u003eFor the trial protocol see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB. Following informed consent by a caregiver, eligible children visited CMCCH for recording pre-intervention outcome measures including assessment, questionnaires and eye-tracking paradigms. Primary outcome measures were the gold-standard clinical assessment using ADOS-2 scores (Comparison and Total scores) and the SRS-2 total score in line with previous trials \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e and conducted by trained research reliable individuals. The comparison ADOS-2 score was included since modules 1, 2 and 3 were used, although each individual was always assessed using the same ADOS-2 module.\u003c/p\u003e\u003cp\u003eSecondary outcomes were assessed using five autism-sensitive eye-tracking paradigms focused on social affect and restricted and repetitive behaviors. These paradigms detailed in \u003cb\u003eFigure S2\u003c/b\u003e, were as follows: (1) Face emotion processing paradigm: Participants were required to observe happy, angry, fearful or neutral static facial expressions, respectively (in total 16 trials; 2s each and 4 models including adult man, adult female, boy and girl); (2) Dynamic social preference paradigm: Participants viewed simultaneous pairs of social (e.g. people dancing) alongside non-social geometric videos (20s each, 2 social and 2 non-social video clips); (3) Joint attention paradigm: Participants watched videos depicting individuals using eye-gaze or a combination of eye-gaze and finger pointing to direct their attention towards target objects (8 trials; 5s each; object location balanced); (4) Social bonding paradigm: Participants observed videos of parent-child social interactions presented alongside static images of toys (8 trials; 5s each; stimulus location balanced); (5) Restricted and repetitive social vs. non-social (RRS) paradigm: Participants viewed concurrent videos of stereotypical social behaviors (e.g., a child spinning) and repetitive non-social motions (e.g., a spinning fan) (8 trials; 5s each; location balanced) (\u003cb\u003esee Figures S2\u003c/b\u003e). All stimuli were counterbalanced and non-overlapping sets were used for pre- and post-intervention assessments.\u003c/p\u003e\u003cp\u003eDuring these eye-tracking paradigms, younger participants sat comfortably on a parent’s lap, while older participants sat independently. All were free to explore and gaze at the display screen based on their interest. Eye‑gaze data were collected using a Tobii TX300 system (Tobii, Danderyd, Sweden) with a sampling rate of 300 Hz and a gaze accuracy of 0.4°. Recording, stimulus presentation, and analysis were performed using Tobii Pro Studio, E‑Prime 2.0, and E‑Prime Extensions for Tobii (Psychology Software Tools, Pittsburgh, PA). Two matched stimulus sets were presented in random order.\u003c/p\u003e\u003cp\u003eOther secondary outcome measures included more detailed specific assessments of social communication using Social Communication Questionnaire (SCQ) \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, repetitive behaviors via the Repetitive Behavior Scale-Revised (RBS-R) \u003csup\u003e34\u003c/sup\u003e, psychoeducational profile (PEP-3, third edition) \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, the Spence Children’s Anxiety Scale (SCAS) \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e and the Wechsler Intelligence Scale \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Finally, the strain experienced by caregivers themselves was assessed using the caregiver strain questionnaire (CSQ) \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eA data-driven eye movement analysis was performed using Hidden Markov Models (hMM) implemented with the EMHMM toolbox (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://visal.cs.cityu.edu.hk/research/emhmm/\u003c/span\u003e\u003cspan address=\"http://visal.cs.cityu.edu.hk/research/emhmm/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) in Matlab. For each participant, an individual hMM was estimated from the fixation data using a variational Bayesian algorithm. This approach automatically selected the optimal number of regions of interest (ROIs, where K = 1, 2, or 3) and estimated all hyperparameters by identifying the model with the highest log-likelihood (LL) for the given data. The best model (i.e., the one with the highest LL) was selected separately for the ASD and TD groups using the baseline data across all five paradigms. The average difference in log-likelihoods serves as an estimate of the Kullback-Leibler (KL) divergence, a measure of dissimilarity between probability distributions. A KL divergence of zero indicates that the two hMMs represent identical distributions. Furthermore, to identify intervention-specific visual patterns, the effects of the two app-based interventions were compared using hMMs within predefined ROIs. Additionally, Pearson’s correlation was used to assess the relationship between the mean log-likelihood (mLL) of specific visual scan patterns and clinical symptoms (ADOS-2 scores). The resulting correlation coefficients were compared using Fisher’s z-transformation. Here, a lower mLL value indicates a lower probability of the corresponding visual scan pattern.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis work was supported by the Humanity and Social Science Foundation of Ministry of Education of China [grant number 24YJC190046-WHZ] and National Natural Science Foundation of China (NSFC) [grant number 82301732- JL].\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eWZ: Conceptualization, Methodology, Formal analysis, Investigation, Data Curation, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing, Visualization, Supervision, Project administration.LZ: Resources, Writing \u0026ndash; review \u0026amp; editing.QL: Methodology, Writing \u0026ndash; original draft.LH: Data Curation, Investigation.YY: Formal analysis.DX: Data Curation.MZ: Data Curation.WY: Resources.WS: Supervision.JL: Supervision, Writing \u0026ndash; review \u0026amp; editing.KMK: Conceptualization, Supervision, Writing \u0026ndash; review \u0026amp; editing.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors would like to thank Infinite Brain Technology for their technological support, including software maintenance and user support.\u003c/p\u003e\u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eData were analyzed using hMM tools available via \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://visal.cs.cityu.edu.hk/research/emhmm/\u003c/span\u003e\u003cspan address=\"http://visal.cs.cityu.edu.hk/research/emhmm/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003ch2\u003eEthics and Declarations\u003c/h2\u003e\u003cp\u003e All written informed consent was provided by participants’ parents or legal guardians. The trial was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board at the University of Electronic Science and Technology (approval number #1061420220115007). It is registered at ClinicalTrials.gov under trial ID NCT06421272 (Registration Date: May 2024).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eShaw, K. A. Prevalence and Early Identification of Autism Spectrum Disorder Among Children Aged 4 and 8 Years \u0026mdash; Autism and Developmental Disabilities Monitoring Network, 16 Sites, United States, 2022. \u003cem\u003eMMWR Surveill. Summ.\u003c/em\u003e 74, (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, X., Chen, X., Su, J. \u0026amp; Liu, N. Prevalence of autism spectrum disorder in mainland china over the past 6 years: a systematic review and meta-analysis. \u003cem\u003eBMC Psychiatry\u003c/em\u003e 24, 404 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, Q. \u003cem\u003eet al.\u003c/em\u003e Facial emotion training as an intervention in autism spectrum disorder: A meta-analysis of randomized controlled trials. \u003cem\u003eAutism Res.\u003c/em\u003e 14, 2169\u0026ndash;2182 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGosling, C. J. \u003cem\u003eet al.\u003c/em\u003e Efficacy of psychosocial interventions for Autism spectrum disorder: an umbrella review. \u003cem\u003eMol. Psychiatry\u003c/em\u003e 27, 3647\u0026ndash;3656 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSandbank, M. \u003cem\u003eet al.\u003c/em\u003e Project AIM: Autism Intervention Meta-Analysis for Studies of Young Children. \u003cem\u003ePsychol. Bull.\u003c/em\u003e 146, 1\u0026ndash;29 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDawson, G. Early behavioral intervention, brain plasticity, and the prevention of autism spectrum disorder. \u003cem\u003eDev. Psychopathol.\u003c/em\u003e 20, 775\u0026ndash;803 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Y., Luo, Y., Zhang, R. \u0026amp; Zheng, X. Direct and indirect costs for families of children with autism spectrum disorder in China. \u003cem\u003eAutism\u003c/em\u003e 27, 2337\u0026ndash;2347 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCakir, J., Frye, R. E. \u0026amp; Walker, S. J. The lifetime social cost of autism: 1990\u0026ndash;2029. \u003cem\u003eRes. Autism Spectr. Disord.\u003c/em\u003e 72, 101502 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBucci, S. \u003cem\u003eet al.\u003c/em\u003e Actissist: Proof-of-Concept Trial of a Theory-Driven Digital Intervention for Psychosis. \u003cem\u003eSchizophr. Bull.\u003c/em\u003e 44, 1070\u0026ndash;1080 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMichie, S., Yardley, L., West, R., Patrick, K. \u0026amp; Greaves, F. Developing and Evaluating Digital Interventions to Promote Behavior Change in Health and Health Care: Recommendations Resulting From an International Workshop. \u003cem\u003eJ. Med. Internet Res.\u003c/em\u003e 19, e232 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiadi, I. \u003cem\u003eet al.\u003c/em\u003e Digital interventions for depression and anxiety in older adults: a systematic review of randomised controlled trials. \u003cem\u003eLancet Healthy Longev.\u003c/em\u003e 3, e558\u0026ndash;e571 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShah, A., Kraemer, K. R., Won, C. R., Black, S. \u0026amp; Hasenbein, W. Developing Digital Intervention Games for Mental Disorders: A Review. \u003cem\u003eGames Health J.\u003c/em\u003e 7, 213\u0026ndash;224 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVoss, C. \u003cem\u003eet al.\u003c/em\u003e Effect of Wearable Digital Intervention for Improving Socialization in Children With Autism Spectrum Disorder: A Randomized Clinical Trial. \u003cem\u003eJAMA Pediatr.\u003c/em\u003e 173, 446\u0026ndash;454 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNahum-Shani, I., Shaw, S. D., Carpenter, S. M., Murphy, S. A. \u0026amp; Yoon, C. Engagement in digital interventions. \u003cem\u003eAm. Psychol.\u003c/em\u003e 77, 836\u0026ndash;852 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGega, L. \u003cem\u003eet al.\u003c/em\u003e Digital interventions in mental health: evidence syntheses and economic modelling. \u003cem\u003eHealth Technol. Assess.\u003c/em\u003e 26, 1\u0026ndash;182 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChatterjee, A., Prinz, A., Gerdes, M. \u0026amp; Martinez, S. Digital Interventions on Healthy Lifestyle Management: Systematic Review. \u003cem\u003eJ. Med. Internet Res.\u003c/em\u003e 23, e26931 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFirth, J. \u003cem\u003eet al.\u003c/em\u003e The efficacy of smartphone-based mental health interventions for depressive symptoms: a meta-analysis of randomized controlled trials. \u003cem\u003eWorld Psychiatry\u003c/em\u003e 16, 287\u0026ndash;298 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoshe, I. \u003cem\u003eet al.\u003c/em\u003e Digital interventions for the treatment of depression: A meta-analytic review. \u003cem\u003ePsychol. Bull.\u003c/em\u003e 147, 749\u0026ndash;786 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOsborn, T. L. \u003cem\u003eet al.\u003c/em\u003e Single-session digital intervention for adolescent depression, anxiety, and well-being: Outcomes of a randomized controlled trial with Kenyan adolescents. \u003cem\u003eJ. Consult. Clin. Psychol.\u003c/em\u003e 88, 657\u0026ndash;668 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFirth, J. \u003cem\u003eet al.\u003c/em\u003e Can smartphone mental health interventions reduce symptoms of anxiety? A meta-analysis of randomized controlled trials. \u003cem\u003eJ. Affect. Disord.\u003c/em\u003e 218, 15\u0026ndash;22 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePauley, D., Cuijpers, P., Papola, D., Miguel, C. \u0026amp; Karyotaki, E. Two decades of digital interventions for anxiety disorders: a systematic review and meta-analysis of treatment effectiveness. \u003cem\u003ePsychol. Med.\u003c/em\u003e 53, 567\u0026ndash;579 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTorok, M. \u003cem\u003eet al.\u003c/em\u003e Suicide prevention using self-guided digital interventions: a systematic review and meta-analysis of randomised controlled trials. \u003cem\u003eLancet Digit. Health\u003c/em\u003e 2, e25\u0026ndash;e36 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, Y. \u003cem\u003eet al.\u003c/em\u003e Computer-based multiple component cognitive training in children with ADHD: a pilot study. \u003cem\u003eChild Adolesc. Psychiatry Ment. Health\u003c/em\u003e 17, 9 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, F. \u003cem\u003eet al.\u003c/em\u003e The Use of Digital Interventions for Children and Adolescents with Autism Spectrum Disorder\u0026mdash;A Meta-Analysis. \u003cem\u003eJ. Autism Dev. Disord.\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10803-024-06563-4\u003c/span\u003e\u003cspan address=\"10.1007/s10803-024-06563-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024) doi:10.1007/s10803-024-06563-4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Y. \u003cem\u003eet al.\u003c/em\u003e Efficacy of gamified digital health interventions for children and adolescents with autism spectrum disorder: a systematic review and meta-analysis. \u003cem\u003eChild Adolesc. Psychiatry Ment. Health\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13034-025-01009-w\u003c/span\u003e\u003cspan address=\"10.1186/s13034-025-01009-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025) doi:10.1186/s13034-025-01009-w.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, T. \u003cem\u003eet al.\u003c/em\u003e Digital interventions for autism spectrum disorders: A systematic review and meta-analysis. \u003cem\u003ePediatr. Investig.\u003c/em\u003e 8, 224\u0026ndash;236 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMukherjee, D. \u003cem\u003eet al.\u003c/em\u003e Digital tools for direct assessment of autism risk during early childhood: A systematic review. \u003cem\u003eAutism\u003c/em\u003e 28, 6\u0026ndash;31 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSandgreen, H., Frederiksen, L. H. \u0026amp; Bilenberg, N. Digital Interventions for Autism Spectrum Disorder: A Meta-analysis. \u003cem\u003eJ. Autism Dev. Disord.\u003c/em\u003e 51, 3138\u0026ndash;3152 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBottos, S. \u0026amp; Balasingam, B. Tracking the Progression of Reading Using Eye-Gaze Point Measurements and Hidden Markov Models. \u003cem\u003eIEEE Trans. Instrum. Meas.\u003c/em\u003e 69, 7857\u0026ndash;7868 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChuk, T., Chan, A. B., Shimojo, S. \u0026amp; Hsiao, J. H. Eye movement analysis with switching hidden Markov models. \u003cem\u003eBehav. Res. Methods\u003c/em\u003e 52, 1026\u0026ndash;1043 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLord, C. \u003cem\u003eet al.\u003c/em\u003e Autism Diagnostic Observation Schedule. 2nd ed. \u003cem\u003eWest. Psychol. Corp.\u003c/em\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLe, J. \u003cem\u003eet al.\u003c/em\u003e Infrequent Intranasal Oxytocin Followed by Positive Social Interaction Improves Symptoms in Autistic Children: A Pilot Randomized Clinical Trial. \u003cem\u003ePsychother. Psychosom.\u003c/em\u003e 91, 335\u0026ndash;347 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChandler, S. \u003cem\u003eet al.\u003c/em\u003e Validation of the Social Communication Questionnaire in a Population Cohort of Children With Autism Spectrum Disorders. \u003cem\u003eJ. Am. Acad. Child Adolesc. Psychiatry\u003c/em\u003e 46, 1324\u0026ndash;1332 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMirenda, P. \u003cem\u003eet al.\u003c/em\u003e Validating the Repetitive Behavior Scale-Revised in Young Children with Autism Spectrum Disorder. \u003cem\u003eJ. Autism Dev. Disord.\u003c/em\u003e 40, 1521\u0026ndash;1530 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFulton, M. L. \u0026amp; D\u0026rsquo;Entremont, B. Utility of the Psychoeducational Profile-3 for Assessing Cognitive and Language Skills of Children with Autism Spectrum Disorders. \u003cem\u003eJ. Autism Dev. Disord.\u003c/em\u003e 43, 2460\u0026ndash;2471 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMagiati, I. \u003cem\u003eet al.\u003c/em\u003e The measurement properties of the spence children\u0026rsquo;s anxiety scale-parent version in a large international pooled sample of young people with autism spectrum disorder. \u003cem\u003eAutism Res.\u003c/em\u003e 10, 1629\u0026ndash;1652 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSadeghi, A., Rabiee, M. \u0026amp; Abedi, M. R. Validation and reliability of the Wechsler Intelligence Scale for Children-IV. \u003cem\u003eDev. Psychol. J. Iran. Psychol.\u003c/em\u003e 7, 377\u0026ndash;386 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhanna, R. \u003cem\u003eet al.\u003c/em\u003e Psychometric properties of the Caregiver Strain Questionnaire (CGSQ) among caregivers of children with autism. \u003cem\u003eAutism\u003c/em\u003e 16, 179\u0026ndash;199 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChawarska, K., Macari, S. \u0026amp; Shic, F. Decreased Spontaneous Attention to Social Scenes in 6-Month-Old Infants Later Diagnosed with Autism Spectrum Disorders. \u003cem\u003eBiol. Psychiatry\u003c/em\u003e 74, 195\u0026ndash;203 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, W. \u003cem\u003eet al.\u003c/em\u003e Eye-Tracking\u0026ndash;Based Measurement of Social Visual Engagement Compared With Expert Clinical Diagnosis of Autism. \u003cem\u003eJAMA\u003c/em\u003e 330, 854\u0026ndash;865 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKou, J. \u003cem\u003eet al.\u003c/em\u003e Comparison of three different eye-tracking tasks for distinguishing autistic from typically developing children and autistic symptom severity. \u003cem\u003eAutism Res.\u003c/em\u003e 12, 1529\u0026ndash;1540 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuillon, Q., Hadjikhani, N., Baduel, S. \u0026amp; Rog\u0026eacute;, B. Visual social attention in autism spectrum disorder: Insights from eye tracking studies. \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e 42, 279\u0026ndash;297 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFujioka, T. \u003cem\u003eet al.\u003c/em\u003e Developmental changes in attention to social information from childhood to adolescence in autism spectrum disorders: a comparative study. \u003cem\u003eMol. Autism\u003c/em\u003e 11, 24 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaliukhovich, D. A. \u003cem\u003eet al.\u003c/em\u003e Social attention to activities in children and adults with autism spectrum disorder: effects of context and age. \u003cem\u003eMol. Autism\u003c/em\u003e 11, 79 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobain, F. \u003cem\u003eet al.\u003c/em\u003e The impact of social complexity on the visual exploration of others\u0026rsquo; actions in preschoolers with autism spectrum disorder. \u003cem\u003eBMC Psychol.\u003c/em\u003e 9, 50 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlack, M. H. \u003cem\u003eet al.\u003c/em\u003e Mechanisms of facial emotion recognition in autism spectrum disorders: Insights from eye tracking and electroencephalography. \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e 80, 488\u0026ndash;515 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHirosawa, T. \u003cem\u003eet al.\u003c/em\u003e Different associations between intelligence and social cognition in children with and without autism spectrum disorders. \u003cem\u003ePLOS ONE\u003c/em\u003e 15, e0235380 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePortoghese, C. \u003cem\u003eet al.\u003c/em\u003e The usefulness of the Revised Psychoeducational Profile for the assessment of preschool children with pervasive developmental disorders. \u003cem\u003eAutism\u003c/em\u003e 13, 179\u0026ndash;191 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis, K. S., Kennedy, S. A., Dallavecchia, A., Skolasky, R. L. \u0026amp; Gordon, B. Psychoeducational Interventions for Adults With Level 3 Autism Spectrum Disorder: A 50-Year Systematic Review. \u003cem\u003eCogn. Behav. Neurol.\u003c/em\u003e 32, 139 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore, A. \u003cem\u003eet al.\u003c/em\u003e The geometric preference subtype in ASD: identifying a consistent, early-emerging phenomenon through eye tracking. \u003cem\u003eMol. Autism\u003c/em\u003e 9, 19 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, W. \u003cem\u003eet al.\u003c/em\u003e A clustering approach identifies an Autism Spectrum Disorder subtype more responsive to chronic oxytocin treatment. \u003cem\u003eTransl. Psychiatry\u003c/em\u003e 14, 312 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParellada, M. \u003cem\u003eet al.\u003c/em\u003e In Search of Biomarkers to Guide Interventions in Autism Spectrum Disorder: A Systematic Review. \u003cem\u003eAm. J. Psychiatry\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1176/appi.ajp.21100992\u003c/span\u003e\u003cspan address=\"10.1176/appi.ajp.21100992\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023) doi:10.1176/appi.ajp.21100992.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"app-based digital intervention, autism spectrum disorder, face processing, Hidden Markov Model, eye-tracking","lastPublishedDoi":"10.21203/rs.3.rs-8631184/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8631184/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIntegrating scalable digital interventions into routine care may expand access to support and reduce the clinical burden for children with autism spectrum disorder (ASD). To evaluate this potential, we first used eye-tracking to identify distinct visual scanning patterns using a Hidden Markov Model approach between school-aged children with ASD and typically developing (TD) controls. This analysis revealed more random visual scan patterns and a reduced preference for social versus nonsocial stimuli in the ASD group. We then conducted a randomized clinical trial assessing a targeted, app-based emotion recognition intervention. Children with ASD were assigned to either two months of emotion recognition training(n = 25, 6.88 ± 1.31 years)or an active control (memory training, n = 25, 7.05 ± 1.54 years), with a six-month follow-up. Importantly, compared to the active control group, the emotion training group showed significantly greater improvements in clinical symptoms, cognitive performance, and eye-tracking measures of face processing, social preference, and joint attention. Mechanistically, the intervention promoted more normative visual scanning, including increased attention to eyes and a greater preference for social stimuli. These findings demonstrate that a targeted digital intervention can modify core visual-behavioral mechanisms and improve symptoms in ASD. This not only underscores the centrality of emotion processing in therapeutic design but also demonstrates the potential of scalable digital tools to augment care of autistic children.\u003c/p\u003e\n\u003cp\u003eTRIAL REGISTRATION ClinicalTrial.gov Identifier: NCT06421272\u003c/p\u003e","manuscriptTitle":"Efficacy of a digital intervention on visual scan pattern for faces in school-aged children with autism","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-11 14:46:35","doi":"10.21203/rs.3.rs-8631184/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6ccacbed-a4b5-492b-9360-3f256c44fcb1","owner":[],"postedDate":"February 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":62056381,"name":"Health sciences/Diseases"},{"id":62056382,"name":"Health sciences/Health care"},{"id":62056383,"name":"Biological sciences/Neuroscience"},{"id":62056384,"name":"Biological sciences/Psychology"},{"id":62056385,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-03-13T00:09:08+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-11 14:46:35","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8631184","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8631184","identity":"rs-8631184","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}