Abstract

The presence of untreated visual disorders in early childhood can result in abnormal visual cortex development (amblyopia). However, accurate clinical assessment of visual function in young children is highly challenging. Reflexive eye movements may allow for precise measurement of visual functions such as resolution acuity in young children if age appropriate, clinically acceptable, quantitative eye tracking techniques can be developed. Children do not tolerate chinrests or head mounted eye-tracking equipment, therefore we have developed a method to measure and compensate for unrestrained head motion that may facilitate detection of eye movements. We implemented an automatic feature-based algorithm to track features on the face in pre-recorded videos. These data were used to “lock” the head to its initial position. Secondly, we implemented a single un-calibrated camera method to estimate the 3D movements of the head. The method was tested using video footage from five children who observed visual stimuli designed to induce horizontal optokinetic nystagmus (a reflexive sawtooth motion of the eye consisting of pursuit and saccadic eye movements). The children’s heads were unrestrained, thereby exhibiting natural movement within the video. Markers placed on participants’ faces were manually segmented to yield ground truth data. The standard deviation of head movement improved from (18.6676, 8.9088) to (1.8828, 1.4282) pixels after stabilization. The average mean square error (MSE) between the manual and automatic stabilization methods was 7.7494 pixels. The percentage error for 3D pose estimation was 0.2428 %. Stabilization of the eyes (relative to the head) was achieved. In conclusion, our initial results suggest that head movement stabilization is possible as a post processing step which could significantly facilitate the monitoring of eye movements in children. Furthermore automated methods could improve the monitoring of neuro-developmental disorders that manifest through head movement.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.