Cracking the Code

Dyslexia and dysgraphia can significantly impact a child’s ability to read and write—making early detection crucial. Screening tools, however, are costly and time-consuming and often limited to identifying only one condition at a time.

A new study led by the University at Buffalo shows how AI-powered handwriting analysis could change that.

The ongoing work, part of the National AI Institute for Exceptional Education, which is based at UB, was recently presented in the journal SN Computer Science. It aims to augment current screening tools and help manage the effects of a nationwide shortage of speech-language pathologists and occupational therapists, who play a crucial role in diagnosing these disorders.

“Catching these neurodevelopmental disorders early is critically important to ensuring that children receive the help they need before it negatively impacts their learning and socio-emotional development,” said the study’s corresponding author, Venu Govindaraju, SUNY Distinguished Professor in the Department of Computer Science and Engineering at UB.

Decades ago, Govindaraju and colleagues did groundbreaking work developing AI that analyzes and reads handwriting, an advancement the U.S. Postal Service still uses to automate the sorting of mail.

The new study now proposes a similar framework and methodologies to identify spelling issues, poor letter formation, writing organization problems and other indicators of dyslexia and dysgraphia.

UB computer scientists are gathering insights from teachers, speech-language pathologists and other professionals to help ensure the AI models they’re developing are viable in the classroom and other settings.

The research team partnered with study co-author Abbie Olszewski, associate professor in literacy studies at the University of Nevada, Reno, who co-developed the Dysgraphia and Dyslexia Behavioral Indicator Checklist, to identify symptoms overlapping between dyslexia and dysgraphia.

Using paper and tablet writing samples from students in kindergarten through fifth grade, the team is now training AI models to complete the screening and compare how effective the models are compared to people administering the test.

“Our ultimate goal is to streamline and improve early screening for dyslexia and dysgraphia, and make these tools more widely available, especially in underserved areas,” said Govindaraju.