Effect of aging on ocular fixation and microsaccades during optic flow

Abstract

Visual fixations correspond to moments of "useful vision", unlike saccades during which we are blind. Optic flow can influence the statistics of fixational eye movements, and therefore affect the quality of visual information accessible to the brain. While aging is known to alter the processing of dynamic visual cues, the influence of optic flow on fixational eye movements remains poorly characterized, partly due to the difficulty of microsaccade analysis in aged subjects, caused by the presence of pupil detection noise. This study assessed ocular fixation statistics and microsaccade characteristics under no-flow (control) as well as radial and tangential optic flow conditions in 51 subjects (21 – 80 y/o). All subjects were ascertained to be healthy by visual, vestibular, sensorimotor, and cognitive clinical examination. Eye movements were recorded at 1000 Hz in head-fixed conditions. Microsaccades were analyzed using a novel unsupervised clustering method that permits reliable detection in the presence of high-frequency pupil detection noise. The algorithm's performance was compared against state-of-the-art methods on both noiseless and noisy data using a ROC analysis. We found that the fixation area was larger in aged compared to young subjects. All optic flow conditions reduced the fixation area to a similar extent in all age groups. Moreover, tangential optic flow significantly affected the ocular fixation drift slope, and it amplified the extent of drift significantly more in aged compared to young subjects. Our microsaccade analysis extends previous data, showing that healthy aging significantly increased microsaccade frequency, amplitude, and peak velocity. Moreover, optic flow influenced all microsaccade characteristics, by reducing frequency, amplitude, and peak velocity. Tangential optic flow significantly triggered microsaccades in the opposite direction to the flow. Importantly, this directional bias tended to be stronger in old adults compared to young subjects.

Meeting abstract presented at VSS 2017