Abstract
BACKGROUND: Cognitive impairment (CI) poses a major global health challenge. In China, neuropsychological scales, regarded as the gold standard for cognitive diagnosis, are largely inaccessible in resource-limited communities. The Mobile Eye-Tracking Application (m-ETA), which captures and quantifies eye movement features, has emerged as a promising tool for CI screening. METHODS: We developed a tablet-based m-ETA using a two-step approach. First, a logistic regression (LR) model was trained to discriminate dementia based on six oculometric features in a hospital cohort (N = 204), and regression analyses were conducted to validate the biological relevance of these features with Alzheimer's Disease-related phenotypes in an exploratory dataset (N = 101). Second, the generalizability and accuracy of the LR model were externally validated in a community-based cohort (N = 433) and further evaluated in two real-world community populations (N = 2,685). Model performance was assessed using sensitivity, specificity, negative predictive value (NPV), and area under the ROC curve (AUC). RESULTS: m-ETA achieved high diagnostic accuracy for dementia (AUC = 0.99). Regression analyses confirmed that the m-ETA-derived oculometric features were significantly associated with cognitive performance, brain atrophy, and tau deposition in the exploratory dataset (all P < 0.05). m-ETA accurately detected CI (AUC = 0.80), with excellent negative predictive value for ruling out CI, and identified individuals with lower cognition performance across diverse communities. CONCLUSIONS: m-ETA offers a low-cost, non-invasive, and efficient tool for large-scale CI screening, particularly suited to underserved and low-literacy communities in China.