Document Type : Full Research Paper
Authors
1
M.Sc., Medical Engineering Group, Department of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
2
Associate Professor, Medical Engineering Group, Department of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
3
Assistant Professor, Institute for Cognitive Science Studies, Tehran, Iran
4
Assistant Professor, Department of Psychology, Imam Hossein University, Tehran, Iran
5
Ph.D. in Health Psychology, Tehran, Iran
6
M.Sc. of Science in Business Administration, Neurobusiness Lab, Deparment of Business Administration and Engineering, School of Management, Economics, and Progress Engineering, Iran University of Science and Technology (IUST), Tehran, Iran
Abstract
The pattern of abnormal gaze is observed in individuals with autism spectrum disorders. Studies of eye movements in people with autism have shown significant difference in the pattern of staring at the eyes and mouth compared to control groups. Yet, findings have been contradictory to date, and in spite of the fact that previous studies on eye dazzling in people with autism are expanding, the findings still do not appear to be consistent. Thus, we tracked eye movements in face processing for 25 teenagers with autism and 25 teenagers from the control group to examine any abnormal concentration in the facial areas. Experimental task used in this study includes standard images of the emotional states of the male and female faces (roundness of the face) in the state of anger, surprise, happiness, sadness and neutrality and subjects looked at these faces, while the eye tracker recorded their eye movements. In this task, they were required to select the displayed emotional state by the reply box. The selected Boosted Trees Ensemble classifier was able to use features related to the total data received from eye tracking in face segmentation into 8 areas (forehead, right and left eye, right and left cheek, nose, mouth and chin) with an accuracy of 83.31% in separating the two groups of autism and control. Moreover, in the study of facial components, left eye, left cheek, right cheek, and right eye, with 84.18%, 83.85%, 82.73% and 81.25% accuracy respectively, were able to make the most difference in the classification. Non-normal patterns in eye gaze can be very important because biomarkers indicate a condition that can be used for early diagnosis. It can also be a guide for researchers to design a game based on the results of this paper to improve the social interactions by strengthening eye contact for people with autism.
Keywords
Subjects