A Serious Game to Train Rhythmic Abilities in Children With Dyslexia: Feasibility and Usability Study

Abstract

Background: Rhythm perception and production are related to phonological awareness and reading performance, and rhythmic deficits have been reported in dyslexia. In addition, rhythm-based interventions can improve cognitive function, and there is consistent evidence suggesting that they are an efficient tool for training reading skills in dyslexia. Objective: This paper describes a rhythmic training protocol for children with dyslexia provided through a serious game (SG) called Mila-Learn and the methodology used to test its usability. Methods: We computed Mila-Learn, an SG that makes training remotely accessible and consistently reproducible and follows an educative agenda using Unity (Unity Technologies). The SG's development was informed by 2 studies conducted during the French COVID-19 lockdowns. Study 1 was a feasibility study evaluating the autonomous use of Mila-Learn with 2500 children with reading deficits. Data were analyzed from a subsample of 525 children who spontaneously played at least 15 (median 42) games. Study 2, following the same real-life setting as study 1, evaluated the usability of an enhanced version of Mila-Learn over 6 months in a sample of 3337 children. The analysis was carried out in 98 children with available diagnoses. Results: Benefiting from study 1 feedback, we improved Mila-Learn to enhance motivation and learning by adding specific features, including customization, storylines, humor, and increasing difficulty. Linear mixed models showed that performance improved over time. The scores were better for older children (P < .001), and children with dyslexia (P < .001) and slower in children with dyslexia (β=-.06; t3816=-5.08; P< .001). Conclusions: Given these encouraging results, future work will focus on the clinical evaluation of Mila-Learn through a large double-blind randomized controlled trial comparing Mila-Learn and a placebo game.