Integrating Clinical Narratives and Structured Data in Alzheimer’s Research: A Multiple Factor Analysis Approach

This study introduces an integrated approach to monitoring Alzheimer's disease in Memory Clinics by combining standardized quantitative measures with narrative-derived information. Using Multiple Factor Analysis, the authors jointly analyze socio-demographic data, cognitive-behavioral tests, and semantic domains extracted from patient narratives through Non-negative Matrix Factorization. Rather than aiming at classification, the study explores how these data sources contribute to a shared factorial structure. Results show that narrative domains provide complementary insights, highlighting convergences and discrepancies between measured deficits and reported difficulties. Although cross-sectional, the framework offers a concise way to map multidimensional patient profiles and may enhance clinical interpretation and follow-up.