Agentic Cognitive Profiling: Realigning Automated Alzheimer’s Disease Detection with Clinical Construct Validity

ACL ARR 2026 January Submission10808 Authors

Keywords: Alzheimer’s Disease, Dementia, Agent, Large Language Model, Construct Validity

Abstract: Automated Alzheimer's Disease (AD) screening has predominantly followed the Inductive paradigm of pattern recognition, which directly maps the input signal to the outcome label. This paradigm sacrifices construct validity of clinical protocol for statistical shortcuts. This paper proposes an Agentic Cognitive Assessment Framework that realigns automated screening with clinical protocol logic. Rather than learning opaque mappings from transcripts to labels, the framework decomposes standardized assessments into atomic cognitive tasks and orchestrates specialized LLM agents to extract verifiable scoring primitives. Central to our design is decoupling semantic understanding from deterministic measurement via function calling, thereby eliminating hallucination and restoring construct validity. On a Cantonese cognitive screening corpus, the framework achieves 90.5\% score match rate in task examination and 85.3\% accuracy in AD prediction, surpassing popular baselines while generating interpretable cognitive profiles grounded in behavioral evidence. This work demonstrates that construct validity and predictive performance need not be traded off, charting a path toward AD screening systems that explain rather than merely predict.

Paper Type: Long

Research Area: Clinical and Biomedical Applications

Research Area Keywords: Clinical and Biomedical Applications,AI / LLM Agents,NLP Applications

Contribution Types: Model analysis & interpretability, NLP engineering experiment, Data analysis

Languages Studied: Cantonese, Chinese

Submission Number: 10808