NEUROLEARN: DEEP LEARNING PLATFORM FOR PREDICTING DISLEXIA AND DYSGRAPHIA

ABSTRACT Dyslexia and dysgraphia are common neurodevelopmental disorders that significantly hinder a child’s ability to read, write, and perform fine motor tasks, directly impacting academic performance and self- esteem. Early detection and intervention are crucial for mitigating long-term educational challenges, yet conventional diagnostic methods often involve time-consuming assessments by specialists and are not easily accessible to all. This project introduces an innovative, cost-effective, and accessible solution through the development of a gamified learning platform integrated with machine learning techniques aimed at the early identification of dyslexia and dysgraphia. The support learning and to collect behavioural and performance data that is indicative of potential learning difficulties. In addition, users are prompted to submit handwriting samples, which are analysed using a deep learning model based on the Res Net architecture to detect spatial and structural irregularities associated with dysgraphia. The collected data from handwriting analysis is processed using supervised machine learning models to identify early signs of dyslexia and dysgraphia with a high degree of accuracy. Results are presented in a user-friendly dashboard, offering educators and parents actionable insights for personalized intervention. By merging engaging digital tools with advanced predictive analytics, this approach enhances the scalability, accessibility, and effectiveness of learning disability screening, paving the way for timely support and improved educational outcomes for children at risk.