Penn Engineering AI Tool Uses Motor Imitation to Accelerate Autism Screening

Researchers from Penn Engineering and the Kennedy Krieger Institute have developed an AI-driven screening tool designed to streamline the identification of autism in children. The system, known as the Computerized Assessment of Motor Imitation (CAMI), or CAMI-2DNet, analyzes how children imitate simple body movements to detect key behavioral markers associated with autism spectrum disorder (ASD).

In a clinical study involving 183 children between the ages of 6 and 13, the system demonstrated a diagnostic accuracy rate between 80% and 85%. Unlike previous methods that often required specialized and costly motion-capture hardware, CAMI utilizes standard video recordings from ordinary cameras to perform its analysis. The findings from this research were published in IEEE Transactions on Biomedical Engineering.

The Role of Motor Imitation

Traditional autism evaluations are often labor-intensive, relying heavily on parent questionnaires and manual behavioral observations by clinicians. A critical component of these evaluations is the assessment of motor imitation, which Dr. Stewart H. Mostofsky of the Kennedy Krieger Institute describes as a "critical building block for social development." While these behaviors are recognized markers for ASD, the manual review process is time-consuming and can create bottlenecks in the diagnostic pipeline.

René Vidal, a professor at Penn Engineering, noted that the CAMI system provides a "much more fine-grained motor assessment than is usually done in the current standardized test." By automating this process through computer vision, the tool removes the technical and financial barriers associated with high-end motion-capture equipment.

Impact on Clinical Access

The shift toward intelligent computer vision allows objective behavioral screening to move out of specialized labs and into community settings. According to Vidal, replacing specialized hardware and manual analysis with AI helps make these assessments more accessible to both clinicians and families.

While the researchers emphasize that CAMI is not a replacement for a comprehensive clinical diagnosis, it serves as a scalable, objective complement to existing protocols. By providing a faster initial screening mechanism, the tool can reduce the time families wait for evaluations and accelerate the path to timely intervention.

Future Outlook

As the tool moves toward broader application, the focus remains on its role as a supportive instrument for medical professionals. The ability to use standard cameras means the system could potentially be deployed in a wider variety of clinics, though the researchers continue to refine the system's precision. Observers will be watching to see if the tool's accuracy holds across broader age groups or different clinical environments beyond the initial study group.