Sensitive Pen : An Open-Source And Low-Cost Digital Pen For Diagnosing Children With Dysgraphia
In IHM'25 - 36e Conférence Internationale Francophone sur l'Interaction Humain-Machine. 2025. à paraître.
Ana Phelippeau, Adrien Chaffangeon Caillet, Adrien Husson, Joël Chevrier
Résumé
Handwriting is a complex motor activity. Handwriting disorders, known as dysgraphia, have a considerable impact on an individual’s academic and professional success. To facilitate the diagnosis of dysgraphia, we propose an open-source and low-cost digital pen, called Sensitive Pen. In a first study, we evaluate its ease of use, usefulness and acceptability with psychomotor therapists. In a second study, we are testing the Sensitive Pen with children aged 6-10 to assess its ability to diagnose dysgraphia using machine learning. Our results show that psychomotor therapists would be ready and interested in using such a tool. Moreover, on a sample of 20 children, using the pen’s cinematic and angle data, we obtained a true positive rate, i.e. of dysgraphic children, of 100%, and an overall accuracy of 65%.