| Contribution title |
Development and Validation of a Novel Gamified Visual Task Suite for Eye-gaze Tracking Analyses in Toddlers |
| Contribution code |
D3.003 |
| Authors |
-
Chui Mae Wong
KK Women's and Children's Hospital
Presenter
-
Shui Er Han
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
-
Laura Sophie Arden-Gardner
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
-
Chuanchu Wang
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
-
Yongwei Zhu
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
-
Yuen Siang Ang
Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR)
-
Kang Hao Cheong
Nanyang Technological University
-
Pratibha Keshav Agarwal
KK Women's and Children's Hospital
-
Hai Hong Zhang
A*STAR Institute for Infocomm Research
|
| Form of presentation |
Poster |
| Topic |
-
T01 - AI and digital health
|
|
Abstract
|
Aims: Eye-gaze tracking (EGT) has been used in autism research and holds promise as a biomarker in autism screening and diagnostic tools. However, EGT results can vary depending on child’s age and video stimuli being presented. Fixation on video stimuli can be as low as 30-40% if not sufficiently engaging, making data interpretation challenging. We aimed to develop a new suite of visual tasks with varied facial features and background complexities to improve task engagement. Methods: Gamified visual tasks were developed to elicit natural responses to social stimuli, including Response To Name (RTN), Response to Gaze/Point (RGP), and Face Preference (FP). Avatars included cartoon versus human faces. Backgrounds were either a non-cluttered still scene or a cluttered animated scene. Total visual task suite duration was 5 min. Neurotypical children were prospectively recruited for EGT responses. Results and Conclusions: Data were collected on 9 children (3 boys, 6 girls; aged 36–51 months), with data from 2 excluded due to technical issues. Mean ± SD percentage fixation time was 83.9±6.8%. RTN and RGP were demonstrated 73.8±27.0% and 61.9±24.5% of the time respectively, regardless of face type or the level of non-social background distraction. During FP, fixation on human faces was higher than non-faces (32±5.7% vs. 20±7.0%), while cartoon face and non-face preferences were comparable (29.4±6.1% vs. 28.6±7.0%). In all, our task suite demonstrated the ability to engage young children in naturalistic responses to social stimuli, and data collection has now started on children with autism.
|