Eye-tracking project

As a joint project with KU-Leuven, we were given access to a dataset containing eye-tracking experimental data. The data comes from the Kinarm software. The setup allows for the study on bimanual control by means of the KINARM Exoskeleton and Gaze-tracker EyeLink Remote.


Our goal was to output an analysis of the gaze and hands tracking positional data.
First, from the raw CSV delivered by the software, we extracted dataframes containing the kinematics data. Four different type of tasks (and files) are available. From the files we then studied what knowledge is accessible in terms of statistics and graphical representation.
The statistics on the measures can be further used for the analysis of evaluation parameters, each parameter being linked in different applied research questions.

Visualisation example

Visually guided reaching task. The timed gaze and hands positions were recovered from the CSV files.

Research questions

  • Cerebral Palsy (CP) group compared to Typically Developed (TD): quantify the difference in eye-movements between the two groups?
  • Unilateral CP children: Is there a difference between eye-movements when the child performs a movement with the affected vs non-affected hand?
  • Is age affecting the movements results ?

For each of these research questions we have different sets of metrics combination meant to evaluate a performance on a given task.