A framework for automated speech and language analysis in Spanish‐speaking Latinos

Our initial findings attest to the usefulness of hypothesis-driven ASLA features to identify persons with AD, MCI, and bvFTD among the Latino population. The framework will now be leveraged in a large multicentric effort across the ReDLat consortium to establish the most robust markers in large, heterogeneous samples. This work will help us streamline ASLA research as an avenue for more equitable clinical testing of dementia in Latin America and beyond.