AI and neuroimaging in autism spectrum disorder: advances in diagnosis, methodological challenges and future directions

From brain scans to classifiers: A systematic review of ML-based autism diagnostic frameworks

Background Autism Spectrum Disorder (ASD) is a lifelong neurodevelopmental condition affecting social interaction, communication, and behavior, with traditional diagnosis relying on subjective and time-consuming behavioral assessments. Advances in neuroimaging have enhanced understanding of the brain mechanisms underlying ASD. Objective This systematic review aimed to comprehensively examine ASD classification datasets and recent advancements in ASD diagnosis using neuroimaging modalities, and to analyze machine learning techniques for ASD diagnosis to evaluate their diagnostic performance in terms of accuracy and Area Under the Curve (AUC). Methods The review followed PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. A comprehensive literature search (2021–2025) was conducted across major databases, including Web of Science, IEEE Xplore, ACM, ScienceDirect, MDPI, and Springer. Results Out of 2,329 initially identified records, 825 were screened for eligibility after title and abstract review. The final analysis included 107 studies, which predominantly used structural and functional Magnetic Resonance Imaging, Electroencephalography, and multimodal datasets for ASD classification. The most common classifiers were Convolutional Neural Networks, Support Vector Machines, Random Forests, and hybrid Deep Learning (DL) models. Studies reported performance metrics such as accuracy and AUC, with many showing promising diagnostic results. Key limitations included small sample sizes, lack of external validation, dataset imbalance, and limited generalizability across multi-site datasets. Conclusion Neuroimaging-based Machine Learning (ML) offers strong potential for improving ASD diagnosis but faces challenges in reproducibility, interpretability, dataset variability, and clinical translation. Future work should focus on multi-site validation, explainable AI, and standardized evaluation to ensure reliable, real-world applications.

NARRATIVE REVIEW OF AUTISM SPECTRUM DISORDER DIAGNOSTICS: CURRENT PERSPECTIVES AND CHALLENGES

Background Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition characterized by persistent impairments in social communication and social interaction, together with restricted and repetitive patterns of behavior, interests, or activities. Accurate diagnosis remains clinically challenging because ASD presentation varies according to age, language level, cognitive functioning, adaptive skills, sex-related characteristics, compensatory strategies, and psychiatric or neurodevelopmental comorbidities. Aims The aim of this narrative review was to summarize contemporary evidence on diagnostic approaches to ASD, with particular emphasis on the Autism Diagnostic Observation Schedule, Second Edition (ADOS-2). The review also considered complementary diagnostic instruments and the main diagnostic challenges in females, adults, high-functioning individuals, and patients with comorbid conditions. Methods A narrative literature review was conducted using PubMed. The search was performed on 23 March 2026 and focused on ASD diagnosis, ADOS-2, ADI-R, diagnostic accuracy, differential diagnosis, psychiatric comorbidity, intellectual disability, adult ASD assessment, female ASD presentation, masking, and camouflaging. A total of 44 sources were included in the final reference list and synthesized qualitatively. No formal systematic review protocol, formal risk of bias assessment, or quantitative meta-analysis was performed. Results The reviewed literature indicates that ADOS-2 remains an important standardized observational instrument for ASD assessment, but it should not be interpreted as an independent basis for diagnosis. Its diagnostic value depends on appropriate module selection, examiner training, clinical context, and integration with developmental history, informant-based data, language level, cognitive functioning, adaptive functioning, and comorbidity assessment. Diagnostic interpretation is particularly challenging in females, adults, high-functioning individuals, and patients with ADHD, anxiety symptoms, depressive symptoms, obsessive-compulsive symptoms, intellectual disability, language impairment, or camouflaging behaviors. Conclusions ASD diagnosis requires a multimodal and individualized approach based on comprehensive clinical judgement rather than on a single test result. ADOS-2, ADI-R, screening questionnaires, and adaptive functioning measures should be interpreted as complementary components of broader clinical assessment. Further research should improve diagnostic sensitivity across diverse populations and develop assessment models that better capture the heterogeneity of autism spectrum disorder.

Early Detection Methods for Autism Spectrum Disorder: From Clinical Screening to Multimodal AI

Early detection of autism spectrum disorder (ASD) in young children is essential for timely referral, developmental monitoring, and access to early intervention. However, conventional screening and diagnostic pathways often depend on parent-report instruments, episodic clinical observation, and specialist-administered assessments, which may delay identification during the first years of life. This scoping review maps the methodological landscape of early ASD detection from traditional clinical screening to multimodal artificial intelligence (AI). A structured literature search was conducted across major biomedical, psychological, and engineering databases for studies published between January 2010 and May 2026. After screening and eligibility assessment, 65 evidence sources were included in the qualitative synthesis, with additional methodological guidelines used to support reporting and appraisal. The reviewed evidence shows that early ASD detection is increasingly shifting from single-session clinical assessment toward multidimensional risk characterization. Clinical and behavioral screening tools remain the foundation of early identification, while eye tracking, video-based motor analysis, acoustic and vocal biomarkers, electroencephalography (EEG), functional near-infrared spectroscopy (fNIRS), and molecular or genomic indicators provide complementary information across different developmental windows. AI-based methods, including machine learning, deep learning, Transformer architectures, multimodal fusion strategies, and foundation-model-based representation learning, may improve the objective quantification of gaze, movement, vocalization, neural activity, and biological risk. Nevertheless, most AI-assisted systems remain limited by small and heterogeneous datasets, insufficient external validation, population bias, privacy concerns, computational burden, and limited interpretability. This review argues that future early ASD detection systems should be developed as clinician-supervised decision-support tools rather than autonomous diagnostic instruments. Clinically meaningful progress will require robust external validation, privacy-preserving deployment, age-appropriate risk stratification, and intrinsically interpretable architectures that align model outputs with developmental and clinical knowledge.

A Scoping Review of Machine Learning and Deep Learning Methods for Autism Spectrum Disorder Diagnosis and Analysis.

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition characterized by diverse behavioral, cognitive, sensory, and communication profiles, making early diagnosis and personalized intervention challenging. Recent advances in machine learning (ML) and deep learning (DL) have enabled the development of computational tools for ASD screening, classification, severity assessment, and intervention monitoring. This review synthesizes findings from 50 recent studies that applied ML and DL techniques to ASD-related datasets, including electroencephalography (EEG), eye-tracking, behavioral video, microbiome, voice acoustic, demographic, and multimodal data. The review addresses three key questions: (i) which data modalities and computational approaches are most frequently used, (ii) how diagnostic performance is evaluated across different study designs, and (iii) what methodological challenges limit clinical translation. The literature is organized according to data modality, algorithmic approach, and clinical readiness. Approaches examined include conventional ML methods, convolutional neural networks, graph neural networks, hybrid deep learning architectures, federated learning, explainable artificial intelligence, topological data analysis, and multimodal fusion. The findings suggest that multimodal and graph-based approaches provide a more comprehensive representation of ASD phenotypes than single-modality methods. Explainability and privacy-preserving learning have also emerged as important considerations for clinical deployment. However, many reported high-performance models are based on small sample sizes, repeated use of the ABIDE dataset, class imbalance, single-site validation, or limited external testing, raising concerns regarding generalizability. Beyond diagnostic accuracy, this review evaluates model interpretability, calibration, scalability, validation rigor, and clinical applicability. Overall, the analysis highlights the need for standardized benchmarks, externally validated multimodal datasets, clinically relevant evaluation metrics, and decision-support systems that complement rather than replace expert clinical assessment in ASD diagnosis and management.

From Data to Early Diagnosis: Artificial Intelligence as a Tool to Support Screening and Detection of Autism Spectrum Disorder in Childhood

Advanced Computational Approaches for Early detection of Autism Spectrum Disorder using Machine and Deep Learning: Recent Trends and Perspective

Autism Spectrum Disorder (ASD) is a neurological and developmental condition characterized by challenges in social interaction, communication (both verbal and non-verbal), and repetitive behaviours. While genetics play a key role in its onset, early diagnosis remains essential for effective intervention. Machine learning (ML) offers a promising approach to streamline and accelerate ASD detection, making it faster and more cost-effective than traditional methods. This paper evaluates eight classification models to identify key ASD features and automate diagnosis. We compare their performance on large datasets to enhance predictive accuracy. ML has transformed healthcare by leveraging vast data volumes for analysis, with technological advances over the past decade improving diagnostic tools now standard in medical settings. ASD affects individuals variably, with symptoms typically appearing between 18 months and 3 years. Although genetic and environmental factors contribute, no single cause is confirmed. Traditional screenings rely heavily on clinician expertise, involving manual assessments and scoring, which can be subjective and time-consuming—even experts face uncertainties in predicting onset or severity. Parents seek rapid, reliable results. ML and deep learning (DL) address these gaps by analyzing complex patterns in data, enabling early prediction of ASD and its severity. This study implements diverse algorithms to support precise, automated screening, reducing diagnostic delays and improving outcomes.