A Systematic Review of NLP for Dementia -- Tasks, Datasets and Opportunities

TL;DR

This systematic review surveys 242 papers across medical, NLP, and speech communities to map how natural language processing has been applied to dementia, spanning dementia detection, linguistic biomarkers, caregiver support, and patient assistance. It finds dementia detection as the dominant focus, but highlights critical gaps in data diversity, methodological rigor, explainability, and real-world applicability. The authors advocate for personalized language models, synthetic data augmentation, multilingual multimodal datasets, and cross-disciplinary benchmarks to translate NLP advances into clinical impact. Overall, the review emphasizes ethical considerations and sustained collaboration between NLP and medicine to realize practical benefits for patients and caregivers.

Abstract

The close link between cognitive decline and language has fostered long-standing collaboration between the NLP and medical communities in dementia research. To examine this, we reviewed over 240 papers applying NLP to dementia-related efforts, drawing from medical, technological, and NLP-focused literature. We identify key research areas, including dementia detection, linguistic biomarker extraction, caregiver support, and patient assistance, showing that half of all papers focus solely on dementia detection using clinical data. Yet, many directions remain unexplored -- artificially degraded language models, synthetic data, digital twins, and more. We highlight gaps and opportunities around trust, scientific rigor, applicability and cross-community collaboration. We raise ethical dilemmas in the field, and highlight the diverse datasets encountered throughout our review -- recorded, written, structured, spontaneous, synthetic, clinical, social media-based, and more. This review aims to inspire more creative, impactful, and rigorous research on NLP for dementia.

Figures (6)

• Figure 1: Number of papers per period by task family.
• Figure 3: Distribution of venues across each of the task families.
• Figure 4: Statistical significance per publication type.
• Figure 5: Key research directions we view as foundational, actionable, and yet-to-be fully explored.
• Figure 6: PRISMA flowchart displaying study screening and selection process.