Toolkit to Examine Lifelike Language v.2.0: Optimizing Speech Biomarkers of Neurodegeneration
Adolfo M. García, F. Ferrante, Gonzalo Pérez, Joaquín Ponferrada, Alejandro Sosa Welford, Nicolás Pelella, Matías Caccia, Laouen Belloli, Cecilia Calcaterra, Catalina González Santibáñez, Raúl Echegoyen, Mariano Javier Cerrutti, Fernando Johann, Eugenia Hesse, Facundo Carrillo
Abstract
INTRODUCTION: The Toolkit to Examine Lifelike Language (TELL) is a web-based application providing speech biomarkers of neurodegeneration. After deployment of TELL v.1.0 in over 20 sites, we now introduce TELL v.2.0. METHODS: First, we describe the app's usability features, including functions for collecting and processing data onsite, offline, and via videoconference. Second, we summarize its clinical survey, tapping on relevant habits (e.g., smoking, sleep) alongside linguistic predictors of performance (language history, use, proficiency, and difficulties). Third, we detail TELL's speech-based assessments, each combining strategic tasks and features capturing diagnostically relevant domains (motor function, semantic memory, episodic memory, and emotional processing). Fourth, we specify the app's new data analysis, visualization, and download options. Finally, we list core challenges and opportunities for development. RESULTS: Overall, TELL v.2.0 offers scalable, objective, and multidimensional insights for the field. CONCLUSION: Through its technical and scientific breakthroughs, this tool can enhance disease detection, phenotyping, and monitoring.