Brenda Strafford Centre on Aging

Reminiscence Therapy (RT) is a valuable non-pharmacological intervention for individuals living with dementia, particularly those experiencing responsive behaviours such as apathy, depression, and anxiety. Traditional RT uses a "Life Review" process, in which individuals recall past life events and people through multiple senses utilizing memory triggers such as photographs, videos, and music; it is a process that has been found to reduce responsive behaviours and depression, enhance communication and cognitive function, and improve the quality of life of people living with dementia. However, personalizing RT can be labor-intensive and often requires personal mementos that may not be accessible. Our project, the Autobiography Gallery, addresses these challenges by using Generative AI to create personalized photo albums that facilitate RT. This innovative approach reduces the effort needed for personalization, making RT more accessible and tailored to everyone’s life story.

To access the effectiveness of the Autobiographical Gallery, our research team—Payam Ranjbar (MSc student, Schulich School of Engineering, University of Calgary), Dr. Mohammad Moshirpour (Donald Bren School of Information and Computer Sciences, University of California, Irvine), and Dr. Linda Duffett-Leger (Faculty of Nursing, University of Calgary)—conducted focus groups with dementia care stakeholders, including healthcare aids and recreation therapists. We also performed a user study with individuals living with dementia and their family caregivers. Participants shared memories that were transformed into custom photo albums. Feedback indicated that AI-generated images effectively stimulated emotional and cognitive responses, enhancing memory recall and reducing agitation. 

Overall, our study highlights the potential of AI-generated imagery to deepen emotional connections between individuals living with dementia and their caregivers, reinforcing the therapeutic value of personalized reminiscence. By leveraging Generative AI, our project reduces the time and effort required to create personalized content, allowing for immediate and engaging experiences that were previously labor-intensive. This innovative approach makes personalized RT more accessible, less dependent on existing personal archives, and tailored to resonate with everyone's unique life story.