Researchers at Texas A&M, led by Dr. Mark Benden, are developing an AI "digital human" that asks screening questions while tracking facial expressions, response time and other biometrics to flag early signs of apathy, a possible precursor to dementia.
The project, funded by the Texas A&M Health Dementia and Alzheimer’s Research Initiative (DARI), aims to replace or augment subjective self‑report scales with objective measures collected during a computer‑based interaction. Benden said the team plans to combine behavioral, cognitive and emotional data into a single score they call a "Digital Apathy Signature."
"Because apathy frequently precedes more visible impairment, identifying it early offers a crucial chance for proactive intervention," Benden said. He noted current clinical tools rely heavily on self‑report questionnaires and can miss subtle or early changes because of recall and wording biases.
The team cites limits in newer computerized tasks as well. For example, the Philadelphia Apathy Computerized Task (PACT) provides behavioral measures but lacks validated individual cut‑offs, which reduces its diagnostic utility, the researchers say.
During a session the virtual interviewer will guide patients through apathy‑related questions while the system records facial cues, timing and other objective signals. The researchers say those signals could allow comparisons across individuals and over time that subjective tests cannot provide.
The project’s stated goal is to validate the Digital Apathy Signature so it can distinguish nuanced levels of apathy severity and support clinical decisions such as referrals, monitoring and treatment evaluation. Texas A&M presents the work as part of DARI’s broader effort to improve detection and care for Alzheimer’s disease and related dementias.
For more about the funding program, Texas A&M links to the initiative at health.tamu.edu/dari.
Photo credit: stories.tamu.edu
Tags: apathy, dementia screening, digital human, facial expression analysis, biomarkers
Topics: Mental health technology, Neuroscience & neuroplasticity, Wearable neurotech