Identifying neuropathologic disease in primary progressive aphasia using narrative speech - PubMed
7 hours ago
- #Primary Progressive Aphasia
- #Neuropathology
- #Artificial Intelligence
- Artificial intelligence is applied to narrative speech to predict neuropathologic disease in primary progressive aphasia (PPA).
- Natural language processing extracts features from transcribed narratives of the Cinderella story.
- Machine learning models classify participants as normal controls (NC) or PPA with Alzheimer's disease (AD) or frontotemporal lobar degeneration (FTLD-4Rtau).
- Models successfully distinguish AD from FTLD-4Rtau and NC from PPA narratives.
- Feature permutation reveals different patterns of language domain and disease pathology contributions to classification.
- AI in naturalistic language tasks may aid in differential diagnosis and disease-specific interventions for PPA.