Detection of Anxiety Disorders in Adolescents Based on Reading Patterns

Eye movement as a biomarker and clinical manifestation represents unconscious brain processes that can objectively disclose abnormal eye fixation of Anxiety disorders (AD). With the aid of eye-tracking technology, plentiful methods that identify AD based on eye movements have been developed, but there are rarely works specifically focusing on the correspondence between eye movement trajectories and viewing content. In this paper, we propose an AD detection method, which aims to learn the actual viewing content of adolescents through eye movement trajectories in answering the Mental Health Testing (MHT) scale. Based the areas of interest (AOIs), we extract the reading orders of adolescents from eye movement trajectories that recorded when they answer the scale. We propose a two-branch model, one branch processes the obtained texts with viewing order, and the other branch processes the position information of each text corresponding to the original question. We compare our model to other models in the dataset we established, and the recognition performance of ours is higher than others. Experimental results verify the effectiveness of our proposed method in anxiety disorders recognition.