New Method to Diagnosis of Dyslexia Using 1D-CNN

• Published in: 2020 Medical Technologies Congress (TIPTEKNO) pp. 1 - 4
• Main Authors: Ileri, Ramis, Latifoglu, Fatma, Demirci, Esra
• Format: Conference Proceeding
• Language: English
• Published: IEEE 19-11-2020
• Subjects: Biomedical Signal Processing; Convolution; Convolutional Neural Network; Dyslexia; Electrooculogram; Electrooculography; Pediatrics; Sensitivity; Task analysis; Training; Writing
• ISSN: 2687-7783
• DOI: 10.1109/TIPTEKNO50054.2020.9299241

Dyslexia is a learning disability that can be characterized by reading difficulties. The EOG signals are widely used in biomedical applications such as Human Computer Interaction (HCI), and the use of the EOG signal in the diagnosis of neurodegenerative diseases is increasing. In this paper, we proposed a novel approach for diagnosis with dyslexia using one dimensional convolutional neural network (ID CNN) based on EOG signals. In the first stage of the study, EOG signals were during healthy and dyslexic children read four different texts. In the second stage, the EOG signals were filtered and segmented into frames. At the last stage, the EOG signals were classified the using ID CNN. According to obtained results, 73. 6128\pm 2.8155 % classification accuracy was performed in classifying the healthy and dyslexic group.