Automating the analysis of eye movement for different neurodegenerative disorders

Abstract

Eye tracking is one of the emerging techniques clinicians use to establish the presence and monitor the progression of neurodegenerative disorders (NDs). The clinical observation and assessment of extra-ocular movements is common practice to detect motoric and cognitive impairments but remains observer-dependent and subjective. In the present study, we propose an algorithm that can automatically identify saccades, fixation, smooth pursuit, and blinks using a non-invasive eye-tracker and, subsequently, extract response-to-stimuli-derived interpretable features that objectively assess patient behaviors. The cohort analysis encompasses control subjects (CTRL) and persons with Alzheimer′s disease (AD) or Mild Cognitive Impairment (MCI), Parkinson′s disease (PD), and Parkinson′s disease mimics (PDM). Overall, results suggested that the control group had significantly longer smooth pursuit distance (p < 0.05) in contrast to other cohorts. Additionally, the average hypometria prosaccade gain was significantly smaller (p < 0.05) for PD and AD/MCI relative to CTRL. The number of omitted saccades relative to a presented stimulus in the antisaccade task and latency were significantly greater (p < 0.05) for PD and AD/MCI compared to CTRL. These features, as oculographic biomarkers, can be potentially leveraged in distinguishing different types of neurodegenerative disorders in their early stages, yielding more objective and precise protocols to monitor disease progression.