Related Experiment Video
Updated: May 14, 2025
Eye Tracking Young Children with Autism
Published on: March 27, 2012
Assessment of Gaze Fixations and Shifts in Children with Cerebral Palsy: A Comparison of Computer- and Object-Based
Tom Griffiths1, Michael T Clarke2, John Swettenham3
1School of Computing, University of Dundee, Dundee DD1 4HN, UK.
Insights
Object presentation yielded higher success rates for gaze behaviors in children with cerebral palsy (CP) compared to eye tracking. However, both methods have limitations, highlighting the need for individualized assessment of gaze skills in CP.
Area of Science:
- Pediatric Rehabilitation
- Assistive Technology
- Clinical Ophthalmology
Background:
- Gaze behaviors (fixation, shifting) are crucial for communication and computer control in children with cerebral palsy (CP).
- Objective quantification of these behaviors is challenging for clinicians, lacking standardized methods.
- Current methods for assessing gaze behaviors in children with CP have limitations.
Purpose of the Study:
- To compare the utility of eye-tracking technology versus direct observation of object presentation for assessing gaze behaviors in children with CP.
- To evaluate the feasibility and effectiveness of two distinct methods for eliciting and observing gaze behaviors.
Main Methods:
- A comparative study involving 39 children with CP.
- Children performed single-target fixation (STF) and target-target fixation shift (TTFS) tasks using both eye-tracking and object presentation.
- Gaze behaviors were recorded and analyzed for each presentation method.
Main Results:
- Eye-tracking calibration failed for 6 children, indicating an accessibility barrier.
- Object presentation showed significantly higher success rates for both STF (81.3%) and TTFS (70.1%) compared to eye tracking (30.3% and 26.9%, respectively).
- Performance was not influenced by developmental age, CP severity, or strabismus.
Conclusions:
- Object presentation may overestimate fixation abilities, while eye tracking faces calibration challenges, especially for children with strabismus.
- Neither method is definitively superior; each has unique advantages and disadvantages for assessing gaze behaviors in children with CP.
- Individualized assessment of gaze behaviors is critical due to performance variability and the limitations of each method.
Abstract:
Background/Objectives: Gaze behaviours, such as fixation on single objects, and switching gaze between two objects are important for signaling messages, making choices or controlling a computer for children with cerebral palsy (CP) and similar movement disabilities. Observing these behaviours can be challenging for clinicians, with a lack of agreement on how they can be objectively quantified or rated. Methods: This study compares two methods of eliciting and observing gaze behaviours: a computer presentation using an eye tracker and an object presentation scored by two independent observers in order to explore the utility of each to clinicians working in this area. Children with CP (n = 39) attempted single-target fixation (STF) and target-target fixation shift (TTFS) tasks using both presentations and the results were compared. Results: Six children were unable to calibrate the eye tracker to the accuracy level required. Significantly higher scores for both STF (81.3% object presentation and 30.3% computer presentation, p < 0.01) and TTFS (70.1% and 26.9%, p < 0.01) were seen on the object presentation, with children's performance not predicted by developmental age, severity of CP or presence or absence of strabismus. It is not possible to definitively state which method gives the "correct" result; however, the difference in reported success rate merits further discussion. Conclusions: Whilst eye tracking may present an "entry barrier" for some children in terms of its accuracy and calibration requirements, object presentation carries with it the risk of over-interpreting children as having fixated. Conversely, eye tracking may be better at recording fixations in children with strabismus, where object-based paradigms may offer more flexible administration for clinicians. The variability in children's performance on both presentations underlines the risk of assuming these skills to be present and the importance of assessing gaze behaviours in individual children.
