ArtEvoke: AI-powered platform to trigger memories in Alzheimer's patients using art imagery

(2025)

Files

Silva_21372300_Bejjani_22392300_2025.pdf
  • Open access
  • Adobe PDF
  • 12.12 MB

Details

Supervisors
Faculty
Degree label
Abstract
This thesis explores how artificial intelligence can support non-drug interventions for people living with Alzheimer’s disease through the use of visual stimuli and art. We present a prototype platform that combines large multimodal models (LMMs), large language models (LLMs), and embedding-based vector search to connect personal stories with curated artworks. Designed to assist memory recall and cognitive engagement, ArtEvoke generates personalized visual narratives or retrieves relevant paintings based on the user’s story. Art is used not only as a memory trigger but also as a tool to promote emotional connection and engagement. The prototype is intended for testing in future therapeutic and research settings, particularly as a scalable support tool for caregivers and clinical practitioners. We evaluate model performance, retrieval effectiveness using FAISS, and user interaction design with a focus on accessibility. These findings highlight the potential of AI tools to support scalable cognitive stimulation in dementia care.