Summary: According to researchers, 10% of young children have trouble learning to write. EPFL researchers have developed new software that can analyze a child’s writing disability and the cause with precision.
Trouble learning how to write, called dysgraphia, affects some 10% of schoolchildren. This learning disability is often associated with dyslexia and can appear in children to varying degrees, with causes that can differ from one child to the next. A team of researchers at EPFL’s Computer-Human Interaction in Learning and Instruction Laboratory (CHILI) has developed software that enables doctors to make highly detailed, personalized assessments of this disability and to accurately identify the letters and numbers that are most difficult and are thus the most discriminative. Their research has just been published in Nature Digital Medicine.
Writing is an essential skill for schoolchildren, but it requires an adroit combination of careful concentration, well-developed motor skills and good language comprehension – something that not all children possess. And although writing problems may appear trivial at first, if they are not treated early on, they could quickly snowball into more serious conditions like a lack of confidence, low self-esteem, trouble learning other skills, a high level of fatigue and even behavioral problems. Early detection is therefore key.
In French-speaking countries, dysgraphia is currently diagnosed using a standardized writing test called BHK. Ergotherapists or psychomotor therapists use the test to assess a child’s handwriting according to 13 criteria. But according to Thibault Asselborn, a PhD student at CHILI and lead author of the study, the BHK test is limited. “It relies entirely on a therapist’s own observation and is therefore subjective. And it can be six months or more between when a teacher first becomes concerned about a child’s writing skills and when the child is finally taken to a specialist.”
Analyzing over 50 different characteristics
The test developed at EPFL, called Tegami, which is run using a tablet computer, represents a major step forward in terms of analytical precision and accuracy of input. It was developed from a database of writing samples from 300 children, around 25% of whom suffered from dysgraphia. The program was able to detect the learning disability 98% of the time.
The big advantage of Tegami is that it can help pinpoint the cause of a child’s dysgraphia because it analyzes no fewer than 53 different characteristics of a child’s writing, which are measured up to 200 times per second. These characteristics include the angle of the pen, the amount of pressure the child applies to the tablet, how fast the child writes and any changes in that speed, whether the child’s hand trembles and if so, with what frequency, and which letters or characters are most discriminative.
A dynamic assessment
“Our software brings a dynamic aspect to the evaluation of a child’s writing. The BHK test lets therapists evaluate a writing sample only after it’s been written. But with Tegami, therapists can analyze the entire writing process and get a clear, comprehensive picture of all of a child’s movements,” says Thomas Gargot, a child psychiatrist, expert in cognitive science, PhD student in computer science at Pitié Salpêtrière teaching hospital in Paris and one of the study’s authors.
According to Gargot, the software also paves the way to classifying different kinds of dysgraphia. The new types of data it collects will enable pediatricians to determine whether there are writing disabilities associated with autism, hyperactivity or attention deficit disorder, and to better understand how teaching methods can be adapted accordingly.
Tegami should also help children with writing disabilities get more targeted treatment. The EPFL researchers are now working with psycho motor therapists and speech therapists to outline remedial measures; for example, if a child shows too much variation in the pressure he puts on his pen, his doctor can prescribe motor-skill exercises. The researchers are also looking into how they can combine Tegami with another program developed at CHILI called CoWriter, where children improve their skills and self-confidence by teaching a robot how to write.
Source: Sarah Perrin – EPFL
Publisher: Organized by NeuroscienceNews.com.
Image Source: NeuroscienceNews.com image is credited to EPFL.
Original Research: Open access research for “Automated human-level diagnosis of dysgraphia using a consumer tablet” by Thibault Asselborn, Thomas Gargot, Łukasz Kidziński, Wafa Johal, David Cohen, Caroline Jolly & Pierre Dillenbourg in Nature Digital Medicine. Published August 31 2018.
Automated human-level diagnosis of dysgraphia using a consumer tablet
The academic and behavioral progress of children is associated with the timely development of reading and writing skills. Dysgraphia, characterized as a handwriting learning disability, is usually associated with dyslexia, developmental coordination disorder (dyspraxia), or attention deficit disorder, which are all neuro-developmental disorders. Dysgraphia can seriously impair children in their everyday life and require therapeutic care. Early detection of handwriting difficulties is, therefore, of great importance in pediatrics. Since the beginning of the 20th century, numerous handwriting scales have been developed to assess the quality of handwriting. However, these tests usually involve an expert investigating visually sentences written by a subject on paper, and, therefore, they are subjective, expensive, and scale poorly. Moreover, they ignore potentially important characteristics of motor control such as writing dynamics, pen pressure, or pen tilt. However, with the increasing availability of digital tablets, features to measure these ignored characteristics are now potentially available at scale and very low cost. In this work, we developed a diagnostic tool requiring only a commodity tablet. To this end, we modeled data of 298 children, including 56 with dysgraphia. Children performed the BHK test on a digital tablet covered with a sheet of paper. We extracted 53 handwriting features describing various aspects of handwriting, and used the Random Forest classifier to diagnose dysgraphia. Our method achieved 96.6% sensibility and 99.2% specificity. Given the intra-rater and inter-rater levels of agreement in the BHK test, our technique has comparable accuracy for experts and can be deployed directly as a diagnostics tool.