Can machine learning support autism screening? Researchers are exploring how these systems are able to survey large data sets and detect subtle behavior patterns. Eye-tracking studies are an example of this phenomenon that allow tools to record how people look at images and videos, as well as social scenes.
Eye-tracking research explains how systems supported by artificial intelligence (AI) can help measure where people focus their gaze during social situations. Those focal points can include faces, movement, or objects. Researchers can then compare attention patterns among large groups, searching for behaviors that might signal differences in development.
Clinicians will still rely on structured assessments, interviews, and a person’s developmental history to better understand an autistic individual’s strengths and support needs. However, this technology can further contribute additional, supportive information as part of a comprehensive autism assessment.
How Machine Learning Detects Behavioral Patterns
More and more often, machine-learning systems are used to evaluate large collections of data about behaviors. In eye-tracking studies, participants view social scenes or certain objects, and cameras record where they look. Researchers can then analyze patterns such as:
- Eye gaze direction
- Length of visual attention
- Shifts between faces and items
- Responses to social versus non-social images
According to recent studies, deep learning models can now evaluate these gaze trends. Research published in Frontiers in Medicine explains how intelligent eye-tracking systems can pair with deep learning models for an automated analysis that detects patterns associated with autism characteristics.
Another Frontiers in Medicine study explored the ways models trained on eye-movement data sets might be able to help identify autism-related visual attention frameworks.
Earlier research supports those findings. A Nature study found that visual attention to eyes can change during infancy among children later diagnosed with autism.
Why Comprehensive Evaluation Still Matters
To identify autism, clinicians combine multiple sources of information. They collect observations, caregiver reports, and standardized assessments to build a complete picture of an individual’s development.
A few well-known assessments that support this process include:
- ADOS-2 (Autism Diagnostic Observation Schedule, Second Edition): A structured observational assessment that examines social communication, interaction, and restricted or repetitive actions.
- MIGDAS-2 (Monteiro Interview Guidelines for Diagnosing Autism Spectrum, Second Edition): A qualitative interview and observation framework highlighting sensory experiences along with interests and everyday behaviors.
- SRS-2 (Social Responsiveness Scale, Second Edition): A rating scale that’s completed by caregivers or teachers to measure subjects’ social awareness, their communication abilities, and related behaviors.
Each of these assessments helps clinicians observe and compare behaviors with typical developmental expectations.
Speech-Language and Sensory Considerations
A comprehensive autism assessment often includes tests in the areas of speech.Language, for example, looks at expressive language, receptive language, and social communication skills, as well as sensory processing, which help clinicians understand how various individuals respond to sound, touch, movement, and other environmental input.
When combining these perspectives, practitioners can understand both strengths and areas where extra support is needed.
Implications for Practitioners
Research into machine learning and eye-tracking continues to expand the methods through which researchers can study autism. For instance, an Archives of General Psychiatry study examined visual fixation during social scenes, demonstrating how gaze patterns relate to social functioning.
Careful clinician evaluation remains a must as new technologies develop. Tools like ADOS‑2, MIGDAS‑2, and SRS‑2—when combined with speech-language and sensory assessments—help practitioners complete thoughtful and individualized autism assessments.
Every autism evaluation is unique. However, the right combination of observational techniques, rating scales, and interviews helps clinicians build a clearer picture of each individual’s strengths and needs. Take a moment to explore the various autism assessments available and learn more about what practitioners use to support comprehensive evaluations.
