Grasping movements could offer simpler autism diagnosis path, finds new study

Researchers at the University of York and the University of Haifa have shown that the automatic learning analysis of naturalistic hands of hands during simple grip tasks can classify autism with approximately 85 percent precision. This advance suggests that subtle motor differences could operate valuable diagnostic markers for autistic spectrum disorder, potentially allowing the previous intervention and better results.

New diagnostic approach takes advantage of motor markers in autism
An innovative study published in the Magazine Autism Research on May 5, 2025 has identified a potentially simpler approach to diagnose autistic spectrum disorder through the analysis of hand movements duration of daily grip tasks.

The International Research Team, led by the associated professor Erez Freud of the Department of Psychology at the University of York and the Vision Research Center, used automatic learning for movement analysis of naturalistic fingers duration in autistic and non -autistic young adults.

“Our models were able to classify autism with approximately 85 % precision, suggestion of this approach could sacrifice simpler and scalable tools for diagnosis,” says Freud.

Motor abnormalities provide early diagnostic opportunities
Autistic spectrum disorder affects approximately one in 50 Canadian children and is generally diagnosed through behavioral evaluations that focus on social communication challenges and retitutional behaviors. However, characteristic thesis markers appear relatively late in development.

The research team indicates that motor abnormalities, which are widely documented in autism, are often manifested in early childhood and could provide previous diagnostic signals, something that is not yet widely liver in the clinical price.

“The main markers of behaviors for the diagnosis focus on those with a relatively late onset and the engines that can be captured very early in childhood can reduce the diagnostic age,” explains Professor Batheva Hady of the University of Haifa.

Study methodology focused on natural movements
The researchers recruited 31 autistic young adults and 28 non -autistic with normal IQ scores. Participants were asked to perform a simple grip task: use their thumbs and index fingers to understand, lift and replace blocks of variable sizes, while the tracking markers attached to their fingers captured precise movement data.

When focusing on young adults instead of children, the researchers assured that the observed differences could not be attributed to delays in development, but INSEAD reflected fundamental differences in motor control.

The research team used five different automatic learning classifiers to analyze the data, achieving a consistent classification accuracy above 84% in all models. When examining the area under the curve (AUC), a measure of classification reliability, achieved scores greater than 0.95 in the analysis in terms and above 0.85 in the test analysis.

Possible classification with minimal characteristics
A particularly promising finding was that the classifiers coined high precision even when a small set of characteristics was used. With only eight carefully selected and correlated, correlated with multiple domains, including domains, including experimental condition, time information, speed data and location information, the classifiers achieved an accuracy of 82-86%.

“These findings suggest that subtle motor control differences can be effective, offering a promising approach to the development of accessible and reliable diagnostic tools for autism,” the authors point out in their document.

Implications for previous diagnosis and intervention
The authors highlight the potential clinical importance of their work in the conclusion of the document: “The current study provides strong evidence that grip movements are a strongly autism diagnosis, and that the ML techniques that the robustness of the robustness of the tasks of robustness and minimum data inputs can be obtained by Bey and the minimum data inputs, our approach sacrifices a promising plane to develop objectives, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible, accessible. relevant to the diagnosis of diagnosis for autism for autism for autism for autism for the reason for an approach. “

This approach could complement the existing diagnostic methods and allow an earlier intervention, which is known to improve the results for autistic people. Researchers suggest that additional studies should explore simolarmatic markers that can be observed in younger populations, specially in early childhood when the visuomotor system is still being developed.

Reference:
Freud, E., Ahmad, Z., Shelef, E., et. Alabama. (2025). Effective autism classification through grip kinematics. Autism Research, 0 (0), 1-12. https://doi.org/10.1002/aur.70049