Artificial intelligence-assisted oculo-gait measurements for cognitive impairment in cerebral small vessel disease.

INTRODUCTION Oculomotor and gait dysfunctions are closely associated with cognition. However, oculo-gait patterns and their correlation with cognition in cerebral small vessel disease (CSVD) remain unclear. METHODS Patients with CSVD from a hospital-based cohort (n = 194) and individuals with presumed early CSVD from a community-based cohort (n = 319) were included. Oculo-gait patterns were measured using the artificial intelligence (AI) -assisted 'EyeKnow' eye-tracking and 'ReadyGo' motor evaluation systems. Multivariable linear and logistic regression models were employed to investigate the association between the oculo-gait parameters and cognition. RESULTS Anti-saccade accuracy, stride velocity, and swing velocity were significantly associated with cognition in both patients and community dwellers with CSVD, and could identify cognitive impairment in CSVD with moderate accuracy (area under the curve [AUC]: hospital cohort, 0.787; community cohort, 0.810) after adjusting for age and education. DISCUSSION The evaluation of oculo-gait features (anti-saccade accuracy, stride velocity, and swing velocity) may help screen cognitive impairment in CSVD. HIGHLIGHTS Oculo-gait features (lower anti-saccade accuracy, stride velocity, and swing velocity) were associated with cognitive impairment in cerebral small vessel disease (CSVD). Logistic model integrating the oculo-gait features, age, and education level moderately distinguished cognitive status in CSVD. Artificial intelligence-assisted oculomotor and gait measurements provide quick and accurate evaluation in hospital and community settings.