Semi-Supervised Clustering for Identification of MCI and Dementia Cohorts with a Brief Digital Cognitive Assessment

Traditional neuropsychological assessment for diagnosis of mild cognitive impairment (MCI) or dementia requires a lengthy in-clinic evaluation by a specialist. This creates a substantial patient burden and prolonged diagnostic and treatment timelines. Digital cognitive assessments (DCA) offer a scalable solution to meet these challenges, but their validation is challenged by the scarcity of large, high-quality datasets with established ground truth. We applied a semi-supervised model-based clustering method to combine a large dataset (N=1189) of the Digital Assessment of Cognition (DAC), a brief, remote-capable DCA, with a smaller dataset pairing DAC assessments with ground-truth neuropsychological diagnoses (N=248). The resulting model identified cognitively unimpaired, MCI, and dementia groups with high accuracy on an external test dataset. Congruent validity was established through strong expected associations with traditional analog assessments. These results validate prior exploratory work and demonstrate the potential for more nuanced, holistic, and scalable cognitive assessments in non-specialist settings.