Neguine Rezaii, Bonnie Wong, Paul Aisen, Laurel Beckett, Jeffrey L. Dage, Ani Eloyan, Tatiana M. Foroud, Kyle Womack, Marı́a C. Carrillo, Joel H. Kramer, Clifford R. Jack, Robert Koeppe, Walter A. Kukull, Kelly Nudelman, Angelina J. Polsinelli, Rema Raman, Malia Rumbaugh, Alexander Taurone, Maryanne Thangarajah, Arthur W. Toga, Prashanthi Vemuri, Alireza Atri, David Clark, Gregory S. Day, Ranjan Duara, Neill R. Graff‐Radford, Ian Grant, Lawrence S. Honig, Erik C. B. Johnson, David Jones, Joseph C. Masdeu, Mario F. Mendez, Erik S. Musiek, Chiadi U. Onyike, Meghan Riddle, Emily Rogalskı, Steven Salloway, Sharon Sha, Raymond Scott Turner, Thomas S. Wingo, S. J. Wolk, Dustin B. Hammers, Kala Kirby, Gil D. Rabinovici, Liana G. Apostolova, Bradford C. Dickerson, Renaud La Joie, Ganna Blazhenets, Alexandra Touroutoglou, Bernardino Ghetti, Lea T. Grinberg, Melissa E. Murray, Kathy L. Newell
Early detection of Alzheimer's disease (AD) is critical yet challenging, particularly in younger individuals. This study leverages artificial intelligence to analyze digital voice recordings from the Craft Story Recall task within the Longitudinal Early-onset AD Study (LEADS) to (1) detect cognitive impairment and (2) differentiate early-onset AD (EOAD) from early onset non-AD cognitive impairment (EOnonAD). Using speech samples from 120 patients and 68 cognitively unimpaired controls, we employed two classification approaches: feature-engineered machine learning and end-to-end deep learning incorporating a Large Language Model. To detect mild cognitive impairment, the feature-engineered model, using acoustic and linguistic features, achieved an AUC of 0.945 on the holdout test set, while the end-to-end model yielded an AUC of 0.988. For differentiating EOAD from EOnonAD, the feature-engineered model achieved an AUC of 0.804, and the end-to-end model yielded an AUC of 0.904 on the holdout set. Explainability analyses revealed reduced linguistic informativeness as a key AD indicator.