Deep learning diagnosis plus kinematic severity assessments of neurodivergent disorders.

Researchers tested whether the way people move could help diagnose autism and ADHD. They used special sensors to measure precise movements while participants reached for targets on a screen. Computer analysis of these movements could accurately identify who had autism, ADHD, both conditions, or neither. The researchers also found patterns in movement that might indicate how severe someone's condition is.

This study investigated whether movement patterns captured through high-definition kinematic sensors could aid in diagnosing neurodivergent disorders. Researchers used deep learning techniques to analyze raw kinematic data from participants performing reaching tasks on touchscreens, aiming to distinguish between autism spectrum disorder (ASD), ADHD, comorbid ASD+ADHD, and neurotypical development. The deep learning approach reportedly achieved high accuracy in predicting participants' conditions based on area under the receiver operator characteristics curve. Additionally, researchers developed biometric measures (Fano Factor and Shannon Entropy) from movement fluctuations that may relate to condition severity, potentially enabling quantitative subtyping of neurodivergent disorders.

This technology could potentially supplement traditional diagnostic approaches by providing objective, quantitative measures. However, extensive validation studies with larger samples and comparison to gold-standard assessments would be needed before clinical implementation.

Sample size not reported, limiting generalizability assessment. Study type unclear. No comparison to standard diagnostic methods. Unclear if findings replicate across different populations or settings. Limited detail on validation procedures.

Early diagnostic assessments of neurodivergent disorders (NDD), remains a major clinical challenge. We address this problem by pursuing the hypothesis that there is important cognitive information about NDD conditions contained in the way individuals move, when viewed at millisecond time scales. We approach the NDD assessment problem in two complementary ways. First, we applied supervised deep learning (DL) techniques to identify participants with autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), comorbid ASD + ADHD, and neurotypical (NT) development.

We measured linear and angular kinematic variables, using high-definition kinematic Bluetooth sensors, while participants performed the reaching protocol to targets appearing on a touch screen monitor. The DL technique was carried out only on the raw kinematic data. The area under the receiver operator characteristics curve suggests that we can predict, with high accuracy, NDD participant's conditions. Second, we filtered the high frequency electronic sensor noise in the recorded kinematic data leaving the participants' physiological characteristic random fluctuations.

We quantified these fluctuations by their biometric Fano Factor and Shannon Entropy from a histogram distribution built from the magnitude difference between consecutive extrema unique to each participant, suggesting a relationship to the severity of their condition. The DL may be used as complementary tools for early evaluation of new participants by providers and the new biometrics allow for quantitative subtyping of NDDs according to severity.