Examples of behaviours common to dyslexia and dysgraphia that can be found in a child's handwriting. — University at Buffalo
A team of researchers from the State University of New York at Buffalo has presented an innovative study on the use of artificial intelligence in the early detection of dyslexia and dysgraphia in children.
This study proposes a method for analysing handwriting, on paper or tablet, in order to identify early signs of this type of disorder. Its aim is to complement current screening tools, which are often costly, time-consuming and focused on a single disorder at a time.
To achieve this, the researchers developed an AI system capable of analyzing various aspects of children's handwriting, from letter formation to general writing organization, including spelling mistakes, syntax and even the use of margins. For the time being, the team has only been able to collect handwriting samples from pupils up to 5th grade in an elementary school in Reno, Nevada.
Still, initial results are encouraging, with nearly 90% of handwriting irregularities detected. By considerably expanding the data provided to train the AI model, this approach could very well revolutionize early detection of disorders such as dyslexia and dysgraphia, enabling rapid intervention to support the educational development of the children concerned.
The model has been designed for use by teachers as a screening tool in the classroom, by speech and language therapists to facilitate rapid intervention, and for parents to monitor their children's progress at home.
Currently, two AI tools are under development: one to identify children requiring a formal needs assessment; and the other to serve as a personalized virtual assistant based on each child's abilities.
This research is published in SN Computer Science. – AFP Relaxnews
