Classifying autism in a clinical population based on motion synchrony: a proof-of-concept study using real-life diagnostic interviews.

Predictive modeling strategies are increasingly studied as a means to overcome clinical bottlenecks in the diagnostic classification of autism spectrum disorder. However, while some findings are promising in the light of diagnostic marker research, many of these approaches lack the scalability for adequate and effective translation to everyday clinical practice. In this study, our aim was to explore the use of objective computer vision video analysis of real-world autism diagnostic interviews in a clinical sample of children and young individuals in the transition to adulthood to predict diagnosis. Specifically, we trained a support vector machine learning model on interpersonal synchrony data recorded in Autism Diagnostic Observation Schedule (ADOS-2) interviews of patient-clinician dyads. Our model was able to classify dyads involving an autistic patient (n = 56) with a balanced accuracy of 63.4% against dyads including a patient with other psychiatric diagnoses (n = 38). Further analyses revealed no significant associations between our classification metrics with clinical ratings. We argue that, given the above-chance performance of our classifier in a highly heterogeneous sample both in age and diagnosis, with few adjustments this highly scalable approach presents a viable route for future diagnostic marker research in autism.

Descriptive Analysis of Social Interaction and Communication Skills of Autistic Children According to Sibling Status and Characteristics.

OBJECTIVE: Sibling relationships in early childhood can have a positive impact on children's social interaction and communication skills. Similarly, autistic children can benefit from interactions with their siblings, who can serve as readily available partners for social interaction. However, there is a lack of research on the effects of siblings based on specific characteristics of the sibling. Therefore, the aim of this study is to compare social interactions and communication skills of autistic children based on sibling status and characteristics. METHODS: We conducted a retrospective data review involving 895 autistic children and their siblings at Seoul National University Bundang Hospital. Variety of diagnostic assessments or questionnaires were administered. Based on the characteristics of the data, Quade's test for nonparametric analysis of covariance was used to compare autism-related symptoms and levels of functioning of the autistic child according to 1) sibling status, 2) birth order, 3) sex, and 4) diagnosis of the sibling. Pearson correlation was used to explore associations between the sibling age gap and different clinical scores. RESULTS: Having siblings was associated with fewer difficulties in restricted and repetitive behaviors. Based on the comparison of the Autism Diagnostic Interview-Revised scores, autistic children with multiple siblings demonstrated better nonverbal behaviors. Autistic children with autistic siblings experienced greater difficulties in social interactions and communications, such as peer relationships, sharing enjoyment, and engaging in social imitative play. CONCLUSION: The study revealed differences in social interactions and communication skills of autistic children based on sibling status, birth order, affected sibling, age gap, and sex.

Clinical characteristics of comorbid tic disorders in autism spectrum disorder: exploratory analysis.

BACKGROUND: The frequency, clinical characteristics, and associated symptoms of comorbid tic disorders in individuals with autism spectrum disorder (ASD) remain unclear. METHODS: We included subsets of individuals from a larger genetic study who were diagnosed with ASD (n = 679; age: 4-18 years) and completed the Yale Global Tic Severity Scale (YGTSS) questionnaire. Based on the YGTSS score, the individuals were divided into two groups: ASD only (n = 554) and ASD with tics (n = 125). Individuals were assessed using the verbal and non-verbal intelligence quotient (IQ), Vineland Adaptive Behavior Scale (VABS-2), Social Responsiveness Scale-2 (SRS-2), Child Behavior Checklists (CBCL), and Yale-Brown Obsessive-Compulsive Scale (YBOCS), followed by between-group comparisons. All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 26. RESULTS: Tic symptoms were observed in 125 (18.4%) participants; among them, most participants presented both motor and vocal tics (n = 40, 40.0%). The ASD with tics group had a significantly higher average age and full-scale IQ score than the ASD only group. After adjusting for age, the ASD with tics group had significantly higher scores in the SRS-2, CBCL, and YBOCS subdomains than the ASD only group. Furthermore, all variables except the non-verbal IQ and VABS-2 scores were positively correlated with the YGTSS total score. Finally, the proportion of tic symptoms was significantly higher among individuals with a higher IQ score (≥ 70). CONCLUSIONS: The IQ score was positively correlated with the proportion of tic symptoms among individuals with ASD. Moreover, the severity of the core and comorbid symptoms of ASD was associated with the occurrence and severity of tic disorders. Our findings suggest the need for appropriate clinical interventions for individuals with ASD. Trial registration This study retrospectively registered participants.

Improving adaptive behaviors for autistic adults without intellectual disability through executive function training.

Executive functioning (EF) is a cognitive process used to perform various daily activities throughout one's lifespan. Autistic adults without intellectual disabilities (ID) also experience difficulties with EF, which is closely associated with adaptive behaviors. For this reason, it is important to improve adaptive behaviors through enhanced use of EF for autistic adults to transition into adulthood successfully. This study aims to conduct a randomized controlled trial to evaluate the effectiveness of a newly developed and modified intervention program. Thirty autistic adults without ID were randomly assigned to the treatment or waitlist group. The participants and caregivers completed various assessments and self-report questionnaires to measure everyday EF and adaptive behaviors. We performed linear mixed-effect modeling to compare the two groups. Data collected at pre-, middle, post-, and follow-up based on participants who completed the program were used to explore changes across time. While there were significant differences in the EF utility-scale (F=5.46, p = .027) between the treatment and waitlist groups, no group x time interactions were detected in other measures. Everyday EF and adaptive behaviors improved when comparing measurements at different time points (p < .001). Our program is Korea's first evidence-based intervention to improve everyday EF and adaptive behaviors for autistic adults without ID.

Loss of Acquired Skills: Regression in Young Children With Autism Spectrum Disorders.

Objectives: Regression, while not a core symptom of autism spectrum disorder (ASD), has been suggested to be a distinct subtype by previous studies. Therefore, this study aimed to explore the prevalence and clinical differences between those with and without regression in children with ASD. Methods: This study includes data from toddlers and young children aged 2-7 years acquired from other projects at Seoul National University Bundang Hospital. The presence and characteristics of regression were explored using question items #11-28 from the Autism Diagnostic Interview-Revised. Chi-square and independent t-tests were used to compare various clinical measurements such as autistic symptoms, adaptative behavior, intelligence, and perinatal factors. Results: Data from 1438 young children (1020 with ASD) were analyzed. The overall prevalence rate of regression, which was mainly related to language-related skills, was 10.2% in the ASD group, with an onset age of 24 months. Regarding clinical characteristics, patients with ASD and regression experienced ASD symptoms, especially restricted and repetitive interests and behaviors, with greater severity than those without regression. Furthermore, there were significant associations between regression and hypertension/placenta previa. Conclusion: In-depth surveillance and proactive interventions targeted at young children with ASD and regression should focus on autistic symptoms and other areas of functioning.

Diagnostic validity of Autism Diagnostic Observation Schedule, second edition (K-ADOS-2) in the Korean population.

BACKGROUND: Although the Korean version of the Autism Diagnostic Observation Schedule-2 (K-ADOS-2) is widely being used to diagnose autism spectrum disorder (ASD) in South Korea, no previous study has examined the validity and reliability of all modules of K-ADOS-2 across a wide age range, particularly older children, adolescents, and adults. METHOD: Data from 2,158 participants were included (mean age = 79.7 months; 73.6% male): 1473 participants with ASD and 685 participants without ASD (Toddler Module, n = 289; Module 1, n = 642; Module 2 n = 574; Module 3 n = 411; Module 4, n = 242). Participants completed a battery of tests, including the K-ADOS or K-ADOS-2 and other existing diagnostic instruments. Sensitivity, specificity, area under the receiver operating characteristic (ROC) curve, positive predictive value (PPV), negative predictive value (NPV), Cohen's kappa (k), and agreement with existing diagnostic instruments were computed. Cronbach's α values were also calculated. RESULTS: All developmental cells of the K-ADOS-2 showed sufficient ranges of sensitivity 85.4-100.0%; specificity, 80.4-96.8%; area under the ROC curve, .90-.97; PPV, 77.8-99.3%; NPV, 80.6-100.0%; and k values, .83-.92. The kappa agreements of developmental cells with existing diagnostic instruments ranged from .20 to .90. Cronbach's α values ranged from .82 to .91 across all developmental cells. LIMITATION: The best-estimate clinical diagnoses made in this study were not independent of the K-ADOS-2 scores. Some modules did not include balanced numbers of participants in terms of gender and diagnostic status. CONCLUSION: The K-ADOS-2 is a valid and reliable instrument in diagnosing ASD in South Korea. Future studies exploring the effectiveness of the K-ADOS-2 in capturing restricted, repetitive behaviors and differentiating ASD from other developmental disabilities are needed.

Patterns of Restricted and Repetitive Behaviors in Toddlers and Young Children with Autism Spectrum Disorder.

Objectives: Restricted and repetitive behaviors (RRBs) are a core symptom in the diagnosis of autism spectrum disorder (ASD). The complexity of behavioral patterns has called for the creation of phenotypically homogeneous subgroups among individuals with ASD. The purpose of this study was 1) to investigate the different types of RRBs and 2) to explore whether subgroups created by RRBs would show unique levels of functioning in toddlers and young children with ASD. Methods: A total of 313 children with ASD, aged 12-42 months were included in the analysis. The Autism Diagnostic Interview-Revised was used to obtain information on the different types of RRBs by grouping 15 items into six categories. The Vineland Adaptive Behaviors Scale, a parent-reported questionnaire, was used to measure adaptive functioning. A portion of the children were analyzed separately for verbal-related RRBs based on their expressive language level. Two-step cluster analysis using RRB groups as features was used to create subgroups. Analysis of covariance while covarying for age and language was performed to explore the clinical characteristics of each cluster group. Results: Sensory-related RRBs were the most prevalent, followed by circumscribed interests, interest in objects, resistance to change, and repetitive body movements. A subset of the children was analyzed separately to explore verbal-related RRBs. Four cluster groups were created based on reported RRBs, with multiple RRBs demonstrating significant delays in adaptive functioning. Conclusion: Heterogeneity of RRBs emerges at a young age. The different patterns of RRBs can be used as valuable information to determine developmental trajectories with better implications for treatment approaches.

End-to-End Model-Based Detection of Infants with Autism Spectrum Disorder Using a Pretrained Model.

In this paper, we propose an end-to-end (E2E) neural network model to detect autism spectrum disorder (ASD) from children's voices without explicitly extracting the deterministic features. In order to obtain the decisions for discriminating between the voices of children with ASD and those with typical development (TD), we combined two different feature-extraction models and a bidirectional long short-term memory (BLSTM)-based classifier to obtain the ASD/TD classification in the form of probability. We realized one of the feature extractors as the bottleneck feature from an autoencoder using the extended version of the Geneva minimalistic acoustic parameter set (eGeMAPS) input. The other feature extractor is the context vector from a pretrained wav2vec2.0-based model directly applied to the waveform input. In addition, we optimized the E2E models in two different ways: (1) fine-tuning and (2) joint optimization. To evaluate the performance of the proposed E2E models, we prepared two datasets from video recordings of ASD diagnoses collected between 2016 and 2018 at Seoul National University Bundang Hospital (SNUBH), and between 2019 and 2021 at a Living Lab. According to the experimental results, the proposed wav2vec2.0-based E2E model with joint optimization achieved significant improvements in the accuracy and unweighted average recall, from 64.74% to 71.66% and from 65.04% to 70.81%, respectively, compared with a conventional model using autoencoder-based BLSTM and the deterministic features of the eGeMAPS.

A Randomized Controlled Trial of the Korean Version of the Program for the Education and Enrichment of Relational Skills for Young Adults (PEERS®-YA-K) With Autism Spectrum Disorder: A Pilot Study.

Evidence-based social skills interventions for young adults are limited, despite social difficulties in autism spectrum disorder (ASD) persisting after transition to adulthood. The Program for the Education and Enrichment of Relational Skills for Young Adults (PEERS®-YA) is an evidence-based intervention found to be effective in improving relational skills in young adults with ASD. To translate the original American version of the PEERS®-YA treatment manual into Korean, intensive interviews were performed. Based on results from interviews, several rules of dating etiquette and social activities were modified to be culturally sensitive and linguistically appropriate. Next, young adults diagnosed with ASD (18-35 years of age; IQ > 70) and their social coaches were recruited for the randomized controlled trial (RCT). Participants were randomly assigned either to a treatment group (TG; n = 19) or a delayed treatment group (DTG; n = 18). In the analysis of group differences in the TG and DTG, social skills knowledge was improved. The within group analyses showed positive effects of improving social skills knowledge on reducing depression and anxiety symptoms. After modest cultural adaptations focusing on dating and social activities, the implementation of the PEERS®-YA-K was found feasible for the Korean community. This is one of only a few cross-cultural validation trials establishing evidence-based treatment in young adults with ASD. Clinical Trial Registration: This trial was registered at ClinicalTrials.gov, identifier: NCT03310775.

Validation of Social Responsiveness Scale for Korean Preschool Children With Autism.

OBJECTIVE: This cross-cultural study aims to examine the psychometric properties of the Social Responsiveness Scale (SRS) and to determine the best-estimate-cut-off scores for the diagnosis of autism spectrum disorder (ASD) in Korean preschool children. METHODS: A total of 563 children was recruited from multiple sources, including Seoul National University Bundang Hospital. Participants were assessed by a multidisciplinary research team using multiple diagnostic tools based on DSM-5 diagnostic criteria. Discriminative validity was tested by comparing the difference in SRS scores between ASD and non-ASD groups. Convergent validity was tested by examining correlations between SRS scores with other diagnostic instruments. A receiver operation characteristic curve analysis was conducted to test the sensitivity and specificity of SRS and to determine the best-estimate-cut-off scores for screening ASD in Korean preschool children. RESULTS: There were significant differences in the total SRS scores between the ASD (n=242) and non-ASD group (n=321) (p<0.01, 95% confidence interval [CI]: 14.08-17.24). The differences were significant even after adjusting for age. SRS scores were significantly correlated with other prescreening diagnostic tools for ASD. The best-estimate cut-off score to screen ASD was 55 (area under curve=0.88, sensitivity 78.1%, specificity 86.6%). CONCLUSION: The SRS is a valid and reliable instrument to screen and aid in the diagnosis of ASD in Korean preschool children. The adjusted cut-off scores, notably lower than in the original U.S. version, may yield a more accurate diagnosis by reflecting transcultural differences.

Short caregiver interview and play observation for early screening of autism spectrum disorder: Behavior development screening for toddlers (BeDevel).

This study validates behavior development screening for toddlers (BeDevel), which utilizes a combination of short caregiver interviews (BeDevel-I) and semistructured play observations (BeDevel-P). The data of 431 toddlers (male 66.2%; mean age (SD) = 29.11 (8.59) months; ASD, n = 201; developmental delay, n = 46; typically developing, n = 184), aged 18 ~ 42 months, were included in the validation of BeDevel. The best clinical estimate diagnosis, screening rate, validity, sensitivity, and reliability of BeDevel were determined based on data cross-sectionally collected using BeDevel and existing diagnostic/screening instruments: autism diagnostic observation schedule (ADOS), autism diagnostic interview (ADI-R), Vineland adaptive behavior scales-II (VABS-II), social response scales (SRS), sequenced language scale for infants (SELSI), Korean childhood autism rating scale (K-CARS), and Korean social communication questionnaire (K-SCQ). The k values of BeDevel-I and BeDevel-P were 0.055 ~ 0.732 and 0.291 ~ 0.752, respectively. Items related to social referencing in BeDevel-P had a particularly high diagnostic validity (k = 0.483 ~ 0.684). Reliabilities of BeDevel-I and BeDevel-P were sufficient (Cronbach's alpha = 0.86 ~ 0.88 and 0.92 ~ 0.95, respectively). BeDevel-I and BeDevel-P showed high sensitivity (BeDevel-I: 85.00 ~ 89.29%; BeDevel-P: 85.00 ~ 91.75%), specificity (BeDevel-I: 77.55 ~ 89.55%; BeDevel-P: 85.09 ~ 97.01%), PPV (BeDevel-I: 70.83 ~ 88.54%; BeDevel-P: 81.52 ~ 94.68%), and NPV (BeDevel-I: 76.00 ~ 95.24%; BeDevel-P: 84.62 ~ 95.45%). The agreement between the composite BeDevel score and ADOS, ADI-R, K-CARS, and K-SCQ was >67.6% (range = 67.6 ~ 90.8%). Combining a short caregiver interview and direct play observation is a valid and reliable screening process. More studies on social referencing as an important early marker are needed. BeDevel can be utilized as a secondary screening instrument before diagnostic confirmation in clinical and community settings. LAY SUMMARY: BeDevel, which consists of a short caregiver interview and direct play observation, is a valid and reliable screening instrument for autism spectrum disorder (ASD). We suggest that BeDevel can be utilized as a secondary instrument before administering diagnostic assessments in clinical and community settings. More studies examining social referencing as a potential behavioral marker of ASD are needed.

State and Trait Anxiety of Adolescents with Autism Spectrum Disorders.

OBJECTIVE: This study examined how state and trait anxiety of adolescents with autism spectrum disorders (ASD) are associated with their demographic characteristics, repetitive and restricted behaviors (RRBs), and internalizing and externalizing problem behaviors. METHODS: A total of 96 participants with ASD (mean age=14.30 years; 91 males) completed a battery of tests including the State/Trait Anxiety Inventory (STAI), the Autism Diagnostic Interview-Revised, the Social Responsiveness Scale (SRS), and a cognitive test measuring intelligence quotient (IQ). Participants' parents completed the Child Behavior Checklist (CBCL). Pearson's correlations among age, IQ, two subscales of the STAI (i.e., STAIS and STAIT, measuring self-reported state and trait anxiety, respectively), and the Anxiety subscale of CBCL (i.e., CBCL-Anxiety, measuring parent-reported trait anxiety) were computed. Subsequently, Pearson's correlations were computed among the three anxiety measures, RRBs, and problem behaviors, while controlling for participants' age and IQ. RESULTS: The STAIS and CBCL-Anxiety were both significantly correlated with higher age, sensory sensitivity, depressive symptoms, somatic complaints, and aggressive behaviors. All three anxiety variables were significantly and positively correlated with total SRS RRB scores. Additionally, the STAIS and STAIT were significantly associated with more severe Compulsion/Adherence behaviors, and the CBCL-Anxiety was also significantly associated with more severe Rule-breaking Behaviors. CONCLUSION: Self-reported state anxiety showed association patterns similar to those of parent-reported trait anxiety. Future studies investigating the precise operationalization of different anxiety instruments are needed to accurately measure the anxiety of adolescents with ASD.

Validating the Autism Diagnostic Interview-Revised in the Korean Population.

OBJECTIVE: This study aimed to examine the validity of the Korean version of the Autism Diagnostic Interview-Revised (K-ADI-R) and determine its efficacy in identifying individuals with autism spectrum disorder (ASD). METHODS: Data were pooled from several past and ongoing studies as well as clinical records acquired at Seoul National University Bundang Hospital from 2008 to 2017. The K-ADI-R were administered and scored by trained research reliable examiners. Measurements to investigate the validity of the K-ADI-R was through sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and Cohen's kappa. RESULTS: A total of 1,271 (age 88.9±62.42 months, male=927) participants were included. The K-ADI-R yielded strong psychometric properties with high sensitivity (86.06-99.27%), specificity (84.75-99.55%), PPV (92.33-99.72%), and NPV (79.43-98.64%). There were significant differences in item scores across the K-ADI-R diagnostic algorithm regardless of age and sex (p<0.001). Agreement between the K-ADI-R and other ASD related measurements ranged between levels of good to excellent. CONCLUSION: Despite language or cultural boundaries, the K-ADI-R demonstrated high levels of sensitivity, specificity, PPV, and NPV within a wide range of participants; hence, suggesting promising usage as a valuable diagnostic instrument for individuals with ASD.

Deep-Learning-Based Detection of Infants with Autism Spectrum Disorder Using Auto-Encoder Feature Representation.

Autism spectrum disorder (ASD) is a developmental disorder with a life-span disability. While diagnostic instruments have been developed and qualified based on the accuracy of the discrimination of children with ASD from typical development (TD) children, the stability of such procedures can be disrupted by limitations pertaining to time expenses and the subjectivity of clinicians. Consequently, automated diagnostic methods have been developed for acquiring objective measures of autism, and in various fields of research, vocal characteristics have not only been reported as distinctive characteristics by clinicians, but have also shown promising performance in several studies utilizing deep learning models based on the automated discrimination of children with ASD from children with TD. However, difficulties still exist in terms of the characteristics of the data, the complexity of the analysis, and the lack of arranged data caused by the low accessibility for diagnosis and the need to secure anonymity. In order to address these issues, we introduce a pre-trained feature extraction auto-encoder model and a joint optimization scheme, which can achieve robustness for widely distributed and unrefined data using a deep-learning-based method for the detection of autism that utilizes various models. By adopting this auto-encoder-based feature extraction and joint optimization in the extended version of the Geneva minimalistic acoustic parameter set (eGeMAPS) speech feature data set, we acquire improved performance in the detection of ASD in infants compared to the raw data set.

How Do Children with Autism Spectrum Disorder Encode and Reproduce Visuospatial Stimuli?: Investigation into Visuospatial Processing Abilities and Styles.

OBJECTIVE: The purpose of this study was to examine the visuospatial processing abilities of children with autism spectrum disorder (ASD) using the Rey Osterrieth Complex Figure (ROCF). METHODS: One-hundred thirty-four children with ASD [mean age (MA)=113.56 months], 150 siblings of children with ASD (MA= 111.67 months), and 55 typically developing (TD) children (MA=109.02 months) were included in this study. During their one-time visit, participants completed the ROCF, various autism diagnostic assessments, and the Korean-Leiter International Performance ScaleRevised. Repeated-measures Analysis of Covariance (ANCOVA) and post-hoc Tukey-Kramer comparisons were computed to compare the ROCF scores. Partial correlations and multiple regressions were computed to examine the association between ROCF scores and the severity of autistic symptoms, as measured by the Autism Diagnostic Interview-Revised (ADI-R) among children with ASD. RESULTS: There were significant main effects of the analysis group in Structural and Incidental Accuracy, Error, and Style. More siblings than TD children drew in a part-oriented way, but the performance of the sibling group was comparable or superior to that of the TD group in all parameters. Social Interaction scores of children with ASD were significantly associated with Organization scores in Copy condition. Whether or not a child drew in the Part-Oriented style significantly predicted his/her repetitive and restricted behavior scores. CONCLUSION: The findings add to the evidence for altered visuospatial processing patterns of ASD as a potential inherent and genetic trait and suggest that this particular cognitive style should not be considered as a deficit. Educational and theoretical implications are discussed.

The Feasibility and Validity of Autism Spectrum Disorder Screening Instrument: Behavior Development Screening for Toddlers (BeDevel)-A Pilot Study.

Although early screening is critical for individuals with autism spectrum disorder (ASD) in order to receive early intervention and improve function later in life, screening is often delayed. Limitations of existing screening instruments, and the need for a culturally appropriate early screening tool in Korean children, led us to develop Behavior Development Screening for Toddlers (BeDevel). The BeDevel assessment consists of two parts: BeDevel-Interview, a structured interview measure for parents/primary caregivers; and BeDevel-Play, a play-based semi-structured observational measure in children. To examine the feasibility and validity of BeDevel, 155 children (N = 75 ASD, N = 55 typical development, N = 25 developmentally delayed) aged 18-42 months (M = 31.54 months, SD = 7.60) were examined through parent-reported screening questionnaires, BeDevel, and standard diagnostic assessments. When BeDevel items were analyzed using Cohen's kappa statistics, most items in BeDevel-Interview and all items in BeDevel-Play were reasonably consistent with diagnoses. We identified primary items, which were significantly interacted with actual diagnosis in the chi-squared test (P < 0.05, range = 0.000-0.032). Using cutoff numbers of items determined using the receiver operating characteristics curve, BeDevel showed satisfactory levels of sensitivity (83.33%-100%), specificity (81.25%-100%), positive predictive values (80.65%-100%), and negative predictive values (83.87%-100%), as well as high internal consistency (Cronbach's α = 0.866-959). The agreement between BeDevel and most other screening/diagnostic instruments was moderate (k = 0.419-1.000). These results suggest that BeDevel can be a useful instrument for early screening of ASD. Autism Res 2019, 12: 1112-1128. © 2019 International Society for Autism Research, Wiley Periodicals, Inc. LAY SUMMARY: Although early screening is critical for individuals with autism spectrum disorder (ASD) in order to receive early intervention and improve function later in life, screening is often delayed. Limitations of existing screening instruments and the need for a culturally appropriate early screening tool in Korean children led us to develop Behavior Development Screening for Toddlers (BeDevel). The BeDevel assessment consists of two parts: BeDevel-Interview, a structured interview measure for parents/primary caregivers; and BeDevel-Play, a play-based, semi-structured observational measure in children. In order to test the feasibility and validity of BeDevel, we analyzed preliminary data of total 155 children aged 18-42 months, examined through parent-reported screening questionnaires, BeDevel, and standard diagnostic assessments. When individual items were analyzed, responses of all BeDevel-Interview items and of most BeDevel-Play items well matched actual diagnoses, and we identified primary items, which were particularly useful in differentiating between the ASD group and the non-ASD group. With the optimal screening criteria determined, the BeDevel was able to identify individuals with a diagnosis of ASD and those without it, all at satisfactory levels. Lastly, BeDevel items were closely related as a set, and the BeDevel screening results were reasonably consistent with the results of most other screening/diagnostic instruments. These results suggest that BeDevel can be a useful instrument for early screening of ASD.

Age as a Moderator of Social Skills Intervention Response Among Korean Adolescents with Autism Spectrum Disorder.

Few studies have investigated the effect of age on treatment effectiveness of evidence-based social skills training in autism spectrum disorder (ASD) adolescents. This study evaluated the effectiveness of Program for the Education and Enrichment of Relational Skills (PEERS®) in different ages of ASD adolescents. A total of 110 ASD adolescents aged 11-19 years were divided into early, middle, and late adolescence groups, and the changes in outcome measures following intervention were compared among the three groups. We found that the improvements in social competence after PEERS® were significant, but the treatment effects were not significantly different among the three age groups. The findings suggest that PEERS® is beneficial for ASD adolescents regardless of patient age and adolescence stage.

The Use of Artificial Intelligence in Screening and Diagnosis of Autism Spectrum Disorder: A Literature Review.

OBJECTIVES: The detection of autism spectrum disorder (ASD) is based on behavioral observations. To build a more objective datadriven method for screening and diagnosing ASD, many studies have attempted to incorporate artificial intelligence (AI) technologies. Therefore, the purpose of this literature review is to summarize the studies that used AI in the assessment process and examine whether other behavioral data could potentially be used to distinguish ASD characteristics. METHODS: Based on our search and exclusion criteria, we reviewed 13 studies. RESULTS: To improve the accuracy of outcomes, AI algorithms have been used to identify items in assessment instruments that are most predictive of ASD. Creating a smaller subset and therefore reducing the lengthy evaluation process, studies have tested the efficiency of identifying individuals with ASD from those without. Other studies have examined the feasibility of using other behavioral observational features as potential supportive data. CONCLUSION: While previous studies have shown high accuracy, sensitivity, and specificity in classifying ASD and non-ASD individuals, there remain many challenges regarding feasibility in the real-world that need to be resolved before AI methods can be fully integrated into the healthcare system as clinical decision support systems.

Re-adjusting the cut-off score of the Korean version of the Childhood Autism Rating Scale for high-functioning individuals with autism spectrum disorder.

AIM: The current cut-off score of the Korean version of the Childhood Autism Rating Scale (K-CARS) does not seem to be sensitive enough to precisely diagnose high-functioning autism. The aim of this study was to identify the optimal cut-off score of K-CARS for diagnosing high-functioning individuals with autism spectrum disorders (ASD). METHODS: A total of 329 participants were assessed by the Korean versions of the Autism Diagnostic Interview - Revised (K-ADI-R), Autism Diagnostic Observation Schedule (K-ADOS), and K-CARS. IQ and Social Maturity Scale scores were also obtained. RESULTS: The true positive and false negative rates of K-CARS were 77.2% and 22.8%, respectively. Verbal IQ (VIQ) and Social Quotient (SQ) were significant predictors of misclassification. The false negative rate increased to 36.0% from 19.8% when VIQ was >69.5, and the rate increased to 44.1% for participants with VIQ > 69.5 and SQ > 75.5. In addition, if SQ was >83.5, the false negative rate increased to 46.7%, even if the participant's VIQ was ≤69.5. Optimal cut-off scores were 28.5 (for VIQ ≤ 69.5 and SQ ≤ 75.5), 24.25 (for VIQ > 69.5 and SQ > 75.5), and 24.5 (for SQ > 83.5), respectively. CONCLUSION: The likelihood of a false negative error increases when K-CARS is used to diagnose high-functioning autism and Asperger's syndrome. For subjects with ASD and substantial verbal ability, the cut-off score for K-CARS should be re-adjusted and/or supplementary diagnostic tools might be needed to enhance diagnostic accuracy for ASD.

Social skills training for children with autism spectrum disorder using a robotic behavioral intervention system.

We designed a robot system that assisted in behavioral intervention programs of children with autism spectrum disorder (ASD). The eight-session intervention program was based on the discrete trial teaching protocol and focused on two basic social skills: eye contact and facial emotion recognition. The robotic interactions occurred in four modules: training element query, recognition of human activity, coping-mode selection, and follow-up action. Children with ASD who were between 4 and 7 years old and who had verbal IQ ≥ 60 were recruited and randomly assigned to the treatment group (TG, n = 8, 5.75 ± 0.89 years) or control group (CG, n = 7; 6.32 ± 1.23 years). The therapeutic robot facilitated the treatment intervention in the TG, and the human assistant facilitated the treatment intervention in the CG. The intervention procedures were identical in both groups. The primary outcome measures included parent-completed questionnaires, the Autism Diagnostic Observation Schedule (ADOS), and frequency of eye contact, which was measured with the partial interval recording method. After completing treatment, the eye contact percentages were significantly increased in both groups. For facial emotion recognition, the percentages of correct answers were increased in similar patterns in both groups compared to baseline (P > 0.05), with no difference between the TG and CG (P > 0.05). The subjects' ability to play, general behavioral and emotional symptoms were significantly diminished after treatment (p < 0.05). These results showed that the robot-facilitated and human-facilitated behavioral interventions had similar positive effects on eye contact and facial emotion recognition, which suggested that robots are useful mediators of social skills training for children with ASD. Autism Res 2017,. © 2017 International Society for Autism Research, Wiley Periodicals, Inc. Autism Res 2017, 10: 1306-1323. © 2017 International Society for Autism Research, Wiley Periodicals, Inc.