Leveraging AI for Dyslexia Detection: Exploring Multimodal Approaches with Machine Learning and Deep Learning

Developmental dyslexia is a learning disorder that typically begins in early childhood, making it difficult for affected individuals to read, write, and spell, even when they possess average or above-average intelligence. Children with dyslexia often experience negative emotions such as low selfesteem, frustration, and anger, highlighting the importance of early detection and intervention. Various methods have been proposed for detecting dyslexia, including reading and writing tests, eye tracking, facial image analysis, and advanced imaging techniques such as Magnetic Resonance Imaging (MRI) and Electroencephalography (EEG). However, one key challenge in dyslexia detection is the reliance on small, unbalanced datasets, which can lead to poor generalization in models. This review suggests that deep learning techniques, particularly those utilizing advanced data augmentation methods, could address this issue by enhancing the model's ability to learn from limited data, improving both accuracy and robustness in real-world applications. The review focuses on the use of Machine Learning (ML) and Deep Learning (DL) for the automatic detection of dyslexia, aiming to improve diagnostic accuracy and accessibility. It explores various ML approaches, such as decision trees and support vector machines, as well as DL architectures, including Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), which analyze text, speech, and eye-tracking patterns. The review provides a comparison of feature extraction methods, classification techniques, and performance metrics, with an emphasis on data enhancement and interpretability. Additionally, it addresses challenges such as dataset variability, ethical concerns, and the practical application of these methods in real-world settings. The goal of this review is to provide an overview of the most current methods for dyslexia detection and propose innovative ways to incorporate ML and DL into personalized learning plans and early intervention programs, ultimately helping children with dyslexia from the outset.