Machine-Learning Algorithm To Detect Speech Patterns Can Detect Early Alzheimer’s Progression

A machine learning algorithm to detect unique speech patterns from everyday conversation can provide early detection for Alzheimer’s disease (AD). The machine-learning model (XGBoost) had a 100% sensitivity score (no false positives) in recognizing trial participants who did not have AD. XGBoost also had a 100% specificity score in recognizing trial participants who did have AD.

In comparison, a classification method based on the Japanese version of the Telephone Interview for Cognitive Status (TICS-J) had an 83.3% specificity score for recognizing individuals who did not have AD (16.7% of participants who were identified to be . . .