ML Techniques Reviewed for Autism Diagnosis and Treatment

Rafael Mu\~noz-Terol, Jes\'us Peral, Sandra Amador, David Gil· August 20, 2026 View original

Key takeaways

  • Machine learning is increasingly vital for early diagnosis and personalized treatment of ASD.
  • Supervised learning dominates, but deep learning and hybrid methods are gaining traction.
  • Integrating multimodal data, including genetic and clinical information, is crucial for accuracy.
  • Wearable devices offer potential for continuous, non-intrusive monitoring of ASD.

Who benefits

HealthcarePharmaceuticalsEdTechSocial ServicesWearable Technology

Summary

This systematic review analyzes 55 studies from 2017-2023 on machine learning applications for Autism Spectrum Disorder (ASD), identifying trends, challenges, and opportunities in diagnosis and treatment. It highlights the dominance of supervised learning, the growing role of deep learning, and the need for multimodal data integration.

Autism Spectrum Disorder (ASD) presents significant challenges in diagnosis and treatment due to its complex and unclear causes. This systematic review examines 55 studies published between 2017 and 2023, focusing on how machine learning (ML) techniques are being applied to address ASD. The primary goal was to identify current trends, effective techniques, and datasets that enhance both diagnosis and treatment outcomes. The review found that supervised learning methods are predominantly used, aligning well with the diagnostic requirements for ASD. However, the increasing availability of data is expanding the role of deep learning in this field. Future opportunities lie in emerging hybrid methods that could combine unsupervised learning, deep learning, and fuzzy logic to tackle the complexities of ASD more effectively. Key challenges highlighted include the necessity for ML models to integrate diverse and complex data types, such as genetic and clinical information, to improve diagnostic accuracy and personalize treatment. The review also emphasizes the potential of innovative data sources like wearable devices and biometric sensors for continuous, non-intrusive monitoring, offering a more holistic understanding of ASD. Addressing these challenges will require interdisciplinary collaboration and the development of expanded, multimodal datasets specifically tailored for ASD research.

Why it matters

Healthcare professionals and researchers can gain insights into the most effective and promising AI/ML approaches for early and accurate ASD diagnosis and personalized treatment, potentially leading to better patient outcomes and resource allocation.

How to implement this in your domain

  1. 1Explore integrating ML-powered diagnostic tools into clinical workflows for early ASD screening.
  2. 2Collaborate with data scientists to develop models that combine genetic, clinical, and behavioral data for comprehensive patient profiles.
  3. 3Investigate the use of wearable devices and biometric sensors for continuous, non-intrusive monitoring of ASD indicators.
  4. 4Participate in interdisciplinary research initiatives to expand and standardize multimodal datasets for ASD.
  5. 5Evaluate the ethical implications and data privacy concerns associated with using ML in healthcare for sensitive conditions like ASD.

Original post by Rafael Mu\~noz-Terol, Jes\'us Peral, Sandra Amador, David Gil

"arXiv:2608.18188v1 Announce Type: new Abstract: Autism spectrum disorder (ASD) is a developmental disability characterized by challenges in social interaction and communication. As the causes of ASD remain unclear, identifying relevant features and hidden correlations is crucial…"

View on X

Originally posted by Rafael Mu\~noz-Terol, Jes\'us Peral, Sandra Amador, David Gil on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses