Artificial Intelligence in Dyslexia Research and Education: A Scoping Review

Abstract

Advancements in artificial intelligence have enabled its integration into research and educational settings, particularly in supporting students with dyslexia. This scoping review explores the transformative role of artificial intelligence in dyslexia research and education, outlining its applications, opportunities, and challenges. A search across Google Scholar, Scopus, Web of Science, PsycINFO, and ERIC yielded 683 unique articles, of which 84 met the inclusion criteria. The findings reveal four primary applications of AI in dyslexia education and research: early detection and diagnosis, personalised learning and interventions, speech and language processing as well as brain imaging and neuroimaging analysis. Additionally, five significant challenges were identified: data quality and quantity, accessibility of artificial intelligence-powered support, bias and validity in assessments, balancing artificial intelligence assistance with human expertise, and transparency in artificial intelligence decision-making. Actionable steps are recommended to enhance dyslexia research and education to address these challenges. Key approaches include improving artificial intelligence tools for early detection, ensuring inclusivity in affordable systems, addressing ethical concerns, fostering interdisciplinary collaboration, training educators and developing culturally relevant tools. Future research should focus on the creation of tools for non-English-speaking and multicultural contexts, advancing research in neuroimaging and personalised learning, longitudinal studies on artificial intelligence interventions, and the establishment of ethical frameworks for the responsible deployment of artificial intelligence in dyslexia education. By navigating these complexities of artificial intelligence integration, we can unlock its full potential to revolutionise dyslexia education and create more inclusive, effective, and ethical learning environments for all.