CPS: A Tool for Classification and Prediction of Autism with STA Using Eye-Tracking Data


Sutoglu E., Sunar S., Sevinc G., Dilbaz P., Eraslan S., Yesilada Y.

15th Turkish National Software Engineering Symposium, UYMS 2021, Virtual, Izmir, Turkey, 17 - 19 November 2021 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/uyms54260.2021.9659649
  • City: Virtual, Izmir
  • Country: Turkey
  • Keywords: autism detection, eye-tracking, scanpath, diagnosis tool
  • Middle East Technical University Affiliated: Yes

Abstract

© 2021 IEEE.Eye-Tracking data is shown to have potential in autism detection. Previous studies show that there are significant differences in the eye-Tracking data of autistic and neurotypical users, in particular when the users engage with web pages. STA (Scanpath Trend Analysis) was previously proposed to be used as a way to analyze eye movement sequences (i.e. scanpaths) of autistic and neurotypical users to identify whether an unknown user can be identified as autistic or not. In this paper, we present a novel web-based application called CPS that aims to guide that process. Performance tests show that the application has a reasonable response time and the preliminary usability test shows that the CPS's usability is satisfactory.