Re-test reliability and accuracy of the Dikablis eye-tracker when sitting, standing and walking Stuart S, Alcock L, Galna B, Lord S, and Rochester L Institute for Ageing and Health, Newcastle University http://research.ncl.ac.uk/hmst |
[email protected]
Introduction
Methods
• Visual sampling involves the ability to quickly redirect the eyes via saccades to
• A Dikablis mobile eye tracker (Figure 1) sampling at 50 Hz (Ergoneers, GmbH)
environmental stimuli relevant to task demands.
was used to track saccadic activity in 20 healthy older adults (≥50 years) with normal or corrected to normal vision.
• Mobile eye trackers capable of tracking saccades during real-world activities
• Accuracy and reliability of saccades were measured on 2 occasions
have recently been developed however no previous studies have investigated
approximately 1 week apart by observing two markers at set distances (5°, 10°,
their reliability or accuracy when walking [1].
15° [2]) (Figure 2) in time with a metronome (60BPM) during 3 tasks: sitting
AIM: To evaluate the accuracy and reliability of using a Dikablis mobile eye-
(with chin rest); standing; and walking on a treadmill (80% comfortable gait
tracker to measure saccadic activity in healthy older adults.
speed). Saccades were identified via a velocity (°/s) threshold based algorithm.
Figure 1 – Dikablis infra-red mobile eye-tracker
Figure 2 – Testing Board
Results Table 1 – Relative error (%) between measured saccades and known distances (5, 10 and 15 degrees) Sitting
Horizontal Vertical
5° 23.4 2.8
Session 1 10° 5.4 24.1
15° 14.3 35.4
5° 24.6 21.3
Standing Session 2 10° 5.0 15.2
15° 11.9 31.5
5° 31.5 11.0
Session 1 10° 4.6 25.2
15° 15.4 33.6
5° 31.2 14.0
Walking Session 2 10° 7.9 15.5
15° 14.2 31.1
5° 11.2 1.0
Session 1 10° 0.8 26.5
15° 15.4 34.3
5° 21.0 6.4
Session 2 10° 1.4 25.8
15° 20.9 31.4
relative
error
(Absolute
Horizontal
Vertical
Saccade Error (°) (Session 2 – Session 1)
Accuracy SITTING
STANDING
WALKING
• Average
error/known distance x 100) was 16.6% for 5° saccades, 13.1% for 10 ° saccades and 24.1% for 15° saccades. • Task error was 17.9% for sitting, 19.6% for standing, and 16.3% for walking. Reliability Mean Saccade Distance (°) (Session 1 & Session 2)
• Overall limits of agreement were 3.9° (4.6° in sitting, 4.1° in standing and
Figure 3 – Bland and Altman plots to show the between session mean saccade distance error (°).
3.0° in walking).
Conclusions •
•
The Dikablis eye-tracker is reliable for measurement of saccades in older
References [1] Stuart, S., Alcock, L., Galna, B., Lord, S. and Rochester, L. (2014) 'The measurement of visual
adults with ~4° error. Error and reliability was influenced by saccade
sampling during real-world activity in Parkinson's disease and healthy controls: A structured literature review', J Neurosci Methods, 222, pp. 175-88.
distance, direction, task familiarity and task demand.
[2] Bahill, A.T., Adler, D. and Stark, L. (1975) 'Most naturally occurring human saccades have magnitudes of 15 degrees or less', Investigative Ophthalmology, 14(6), pp. 468-9.
Observation during testing also suggests that reliability may also be influenced by error derived from calibration quality, eye vs. field camera positioning, head unit movement and the ability to track the pupil through glasses or contact lenses.
Acknowledgements The Research was supported/funded by the National Institute for Health Research (NIHR) Newcastle Biomedical Research Unit based at Newcastle Hospitals NHS Foundation Trust and Newcastle University. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.
Presented at the International Society of Posture and Gait Research World Congress, Vancouver, Canada, 2014