Effective Autism Classification Through Grasping Kinematics.
Erez Freud, Zoha Ahmad, Eitan Shelef and 1 others
PMID 40323705WHAT IT FOUND
A machine learning model classified autistic from non-autistic young adults with 84% to 89% accuracy using only thumb and finger movement data.
This suggests grasping kinematics could serve as an objective diagnostic marker, though it requires validation in children.
Key findings
01Classifiers achieved 84% to 89% accuracy in distinguishing autistic from non-autistic young adults using kinematic data from two passive markers on the thumb and index finger.
02Autistic individuals showed significantly longer movement times during grasping tasks compared to non-autistic individuals.
03High classification accuracy was maintained with as few as eight non-redundant kinematic features, spanning spatial, temporal, and velocity domains.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for OTsWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study only included young adults with normal IQ, so the results do not apply to children or autistic individuals with cognitive impairments. The classification was performed on data from the same participants who performed the task, which may not generalize to new populations or different testing environments. The study did not examine autism subtypes, which are known to have different motor profiles. The high accuracy relies on aggregating data across many trials; single-trial classification is less reliable (AUC 0.85-0.88).
Declared interests
The authors declare no conflicts of interest. The study was funded by the United States-Israel Binational Science Foundation, the Israel Science Foundation, the Canada First Research Excellence Fund, and the Natural Sciences and Engineering Research Council of Canada.
The easy way to misread this
Do not interpret the high classification accuracy as evidence that grasping kinematics can currently diagnose autism in clinical practice. The study used a small sample of young adults with normal IQ in a controlled lab setting, and the method has not been validated for children or individuals with intellectual disabilities, who represent a significant portion of the autistic population.