Subtle hand movements during grasping tasks may help diagnose autism spectrum disorder (ASD) with high accuracy. Using machine learning, researchers analyzed how autistic and non-autistic individuals used their fingers to grasp objects, achieving approximately 85% classification accuracy.