Artificial Intelligence Enhances Speed and Precision in Autism and ADHD Diagnosis

Recent advancements in artificial intelligence (AI) are revolutionizing the diagnosis process for neurodivergent conditions such as autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD). Traditionally, diagnosing these disorders can take up to 18 months due to the reliance on subjective assessments, clinical interviews, and behavioral surveys, which often lack standardized biological markers.

A groundbreaking study led by researchers at Indiana University introduces a data-driven approach that leverages AI to identify quantifiable biomarkers and biometric data, significantly reducing diagnostic time to approximately 15 minutes. This innovative method involves analyzing minute movement patterns with sensors attached to participants' hands while performing simple reaching tasks. These sensors capture high-definition kinematic data, including acceleration, rotation, and linear movements, which reveal subtle differences in motor control between neurotypical individuals and those with ASD or ADHD.

The research builds on earlier work from 2018, where the team discovered movement biomarkers imperceptible to the naked eye but measurable with sensors. Using advanced deep learning algorithms, they assessed movement irregularities that correlate with the severity of neurodivergent disorders. This approach offers a new set of objective biomarkers, enabling clinicians to evaluate how serious a condition is and monitor treatment effectiveness.

While the AI-powered diagnostics are designed to augment, not replace, professional clinical assessments, they could serve as valuable tools for early screening and triage in educational and clinical settings. For example, schools could utilize this technology to identify students who may need further evaluation and intervention, facilitating earlier support and care.

The development of these quantitative tools aims to provide a more standardized and reliable method for diagnosing autism and ADHD, ultimately leading to more personalized and effective treatments. By quantifying movement irregularities and employing deep learning, healthcare providers can better understand the severity of neurodivergent conditions and tailor interventions accordingly.

For more details, refer to the original study published in Scientific Reports: Khoshrav P. Doctor et al, Deep learning diagnosis plus kinematic severity assessments of neurodivergent disorders, 2025.