Spectrul autist și particularitățile asociate.
496 de articoleThis is a peer-reviewed comment on research evaluating augmentation methods in machine learning classification of autism spectrum disorder using 3D convolutional neural networks applied to fMRI data. The commentary discusses integration of medical imaging, behavioral analysis, and AI systems for improved ASD diagnostic approaches.
This eye-tracking study compared how 5-to 8-year-old Chinese children with ASD and typically developing peers understood reciprocal social gestures like giving and taking. Children with ASD showed intact action prediction abilities but demonstrated selective difficulties with processing the social meaning of giving gestures, suggesting deficits in top-down social processing rather than core action understanding.
This study examined whether the Japanese version of the AQ could differentiate between autism spectrum disorder, social communication disorder, and neurotypical groups using ROC analysis across 127 ASD patients, 52 SCD patients, and 49 controls. Optimal AQ cutoff scores were identified for distinguishing ASD from neurotypical individuals (23 points, AUC=0.96) and SCD from neurotypical individuals (22 points, AUC=0.89), demonstrating the instrument's clinical utility for screening.
This study validated the Social Communication Questionnaire (SCQ) as an autism screening tool in a sample of 205 Nigerian adolescents using the ADOS-2 as the diagnostic standard. The SCQ demonstrated good psychometric properties with sensitivity of 0.81 and specificity of 0.88 at a cut-off score of 10, supporting its use for ASD identification in this population.
This study examined how bilingual exposure affects language development in autistic children from Hong Kong, comparing bilingual (Cantonese-English) autistic children with monolingual autistic peers and typically developing children matched on key variables. Results showed bilingual autistic children performed comparably to typically developing children and better than monolingual autistic children on linguistic tasks, suggesting a potential bilingual advantage rather than burden for autistic language learners.
This study compared satisfaction with emergency department care between mothers of children with ASD (n=59) and without ASD (n=69) in Israel, finding significantly lower satisfaction among ASD mothers regarding wait times, staff responsiveness, coordination, and organization. Communication difficulties in children were identified as a key predictor of maternal satisfaction, suggesting EDs need targeted awareness and training for supporting autistic children and their families.
This study presents a deep neural network approach using bimodal convolutional neural networks to detect ASD in children by analyzing gait patterns from video recordings. The method combines non-invasive gait analysis with deep learning to achieve objective, high-accuracy ASD identification.
This ERP neuroimaging study compared brain responses to faces and houses in preschoolers with ASD, fragile X syndrome (FXS), autism-risk siblings, and typically developing controls. The fragile X group showed larger P1 and N170 components than other groups, while ASD and at-risk siblings showed reduced N170 responses compared to controls and FXS groups, suggesting distinct neural patterns of face processing across these conditions.
This qualitative study examined mothers' experiences and opinions about how home environment characteristics influence autistic children, using focus groups and thematic analysis. Findings identified sensory factors, routines, and physical environment as key elements affecting autistic individuals' well-being and behavior in the home setting.
This study validates the RITA-T, an autism screening tool for 18-36 month-old children, in a Turkish population and confirms its good sensitivity and specificity for identifying autism spectrum disorder. The validation enables broader use of this reliable screening instrument to support early intervention programs across diverse populations.
This study examined the relationship between gender diversity and autism traits in 379 Chinese children aged 4-12 using validated questionnaires completed by parents. The findings confirm that increased gender diversity correlates with increased autism traits in this non-clinical, non-Western sample, with imagination and patterns subscales showing particularly strong associations.
This study examined self-reference effects (SRE) in memory among autistic preschoolers and their relationship to social functioning. Results showed autistic children demonstrated significant SREs, with larger effects associated with better parent-rated social functioning, particularly in children with fewer autism traits.
This study compared irritability symptoms in 21 children with high-functioning autism, 21 with disruptive mood dysregulation disorder (DMDD), and 14 healthy controls using standardized behavioral assessment scales. Results showed that autism and DMDD share similar irritability profiles but differ in aggression patterns, with implications for differential diagnosis and clinical management of mood dysregulation in autistic children.
This study compared irony comprehension abilities in 20 autistic children versus 25 typically developing children, examining the cognitive factors underlying these differences. Results indicated that theory of mind (particularly second-order ToM) significantly predicts irony comprehension deficits in autism and may serve as a cognitive marker for autism symptom severity.
This fMRI study identified ten white-matter functional networks in autistic and control participants, finding significantly decreased white matter connectivity in occipital, middle temporo-frontal, and deep networks in autism spectrum disorder. The research also documented reduced spontaneous activity in specific networks and altered white-matter-gray-matter connectivity patterns associated with ASD.
This study describes a tablet-based digital cognitive intervention software developed for young children with ASD using Applied Behavior Analysis principles, and evaluates its feasibility through preliminary testing with 6 children followed by a 3-month study with 12 additional children. Results indicate the software shows feasibility and preliminary effectiveness for cognitive intervention in ASD, with positive parent attitudes toward digital-assisted therapy.
This paper proposes that autism may represent one branch within a broader tree of neurodevelopmental disorders rather than a distinct category, based on evidence of phenotypic and genetic overlap with ADHD, schizophrenia spectrum, and bipolar disorders. The author suggests organizing these conditions under three unifying frameworks: Global Neurodevelopmental Disorder, Extreme Sensitivity to the Environment Syndrome, and Neurotransmission Systems Dysfunction Disorder.
This paper proposes a framework for an explainable AI care agent designed to assist with social interaction training for individuals with autism spectrum disorder. The work presents interaction protocols, training models, and discusses practical implementation considerations through representative social scenarios.