Development and evaluation of an autobiographical memory training protocol for older adults using wearable cameras

Abstract

Autobiographical episodic memory is the cognitive ability most impaired by healthy aging and it has been a primary target of cognitive training studies in this population. We propose a memory training that uses images from the participants’ past days as stimuli thus training abilities closer to the way memory is used in the real world. Wearable cameras take hundreds of pictures per day from a first-person perspective without user intervention and can be used to collect personalized retrieval cues. This thesis is divided into two papers: the first describes the development of the structure of a memory training protocol and the process of adapting equipment and software for older adults. We developed a Memory Task program that integrates image processing software from Dublin City University and presents the task over the course of several weeks without the need for the researcher’s intervention. The second paper discusses preliminary results for the autobiographical memory training protocol. A sample of 15 older adults was divided into two groups that trained with temporal order tasks using either wearable camera pictures (AM group) or images from a database (LM group) as stimuli. The results showed a trend for participants in the AM group to improve more than participants in the LM group. Moreover, the AM group improved significantly in a fluid cognition measure after the training. This thesis presents an innovative method for training autobiographical memory using ecologically valid stimuli that are engaging and meaningful to the participants.