Purchase this article with an account.
Dorion Liston, Leland Stone; Signal-Detection Analysis of Neural Impairment using Oculomotor Assessment. Journal of Vision 2015;15(12):215. doi: 10.1167/15.12.215.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Various types of brain disorders affect visuomotor processing and generate characteristic patterns of deficits in oculomotor behavior (e.g., Leigh & Zee, 2006). Whereas current clinical tests that assess static foveal acuity, perimetry, and contrast sensitivity provide early indicators of visual pathology, the oculomotor system may also provide sensitive and reliable signals that can be used to detect disease onset and track the severity of functional impairment. Methods. We developed a 15-minute test that derives a ten-element vector of metrics from the oculomotor responses to direction-and-speed randomized radially moving stimuli, including pursuit initiation and steady-state tracking, as well as direction-tuning and speed-tuning metrics (Liston & Stone, JOV, in press). Using this task, we collected data from a baseline population of normal observers (N=41) and a few observers with retinal pathology (glaucoma N=2; retinitis pigmentosa N=1, sampled twice over an interval of 17 months). For each metric, we normalized the raw measurements into a standard normal distribution. For a given pathology, we defined the oculomotor "disease vector" as the distance between the average vector for a patient population and the mean of the normal population. The inner product between a disease vector and the vector from any individual patient yields a linear detection metric of the severity of their impairment or their “impairment index”. Results. For two glaucoma patients tested, we observed significant impairment in one observer (index: 2.27; p< 0.05), but not in the other (index: 0.70;p>0.05). For our single retinitis pigmentosa patient, we observed significant impairment at both time points (indices: 2.28 and 3.86, p< 0.05 and p< 0.01) with the index changing significantly over time (p< 0.01, bootstrap test). Conclusions. Using a multi-dimensional performance vector computed using our oculomotor assessment test, “disease vectors” and “impairment indices” can be constructed to characterize, detect, and track the progression of visuomotor pathology.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only