Gallery Game: Smartphone-based assessment of long-term memory in adults at risk of Alzheimer’s disease

ABSTRACT

Introduction: Gallery Game, deployed within the Mezurio smartphone app, targets the processes of episodic memory hypothesized to be first vulnerable to neurofibrillary tau-related degeneration in Alzheimer’s Disease, prioritizing both perirhinal and entorhinal cortex/hippocampal demands.

Methods: Thirty-five healthy adults (aged 40–59 years), biased toward those at elevated familial risk of dementia, completed daily Gallery Game tasks for a month. Assessments consisted of cross-modal paired-associate learning, with subsequent tests of recognition and free recall following delays ranging from one to 13 days.

Results: Retention intervals of at least three days were needed to evidence significant forgetting at both recognition and paired-associate recall test. The association between Gallery Game outcomes and established in-clinic memory assessments were small but numerically in the anticipated direction. In addition, there was preliminary support for utilizing the perirhinal-dependent pattern of semantic false alarms during object recognition as a marker of early impairment.

Conclusions: These results support the need for tests of longer-term memory to sensitively record behavioral differences in adults with no diagnosis of cognitive impairment. Aggregate behavioral outcomes promote Gallery Game’s utility as a digital assessment of episodic memory, aligning with established theoretical models of object memory and showing small yet uniform associations with existing in-clinic tests. Initial support for the discriminatory value of perirhinal-targeted outcomes justifies ongoing large-sample validation against traditional biomarkers of Alzheimer’s disease.

KEYWORDS:

• Alzheimer’s disease
• digital technology
• episodic memory
• cognitive assessment
• perirhinal cortex

Author contribution statement

C.H. supervised the project, designed the Gallery Game task, and led development of the Mezurio smartphone platform and data processing pipelines. C.H., C.L, K.T, and C.C. contributed to the experimental paradigm used in Gallery Game. C.L and C.C. contributed to the statistical processing of results. C.L. prepared this manuscript and reviewed related literature. I.K, A.C and J.B. recruited and supported study participants. I.K. led the PREVENT study within Oxford, wrote the ethics application for this extension, and acquired supplementary funding to provide handsets for participants without. All authors provided feedback on the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Claire Lancaster is jointly funded by Eli-Lilly and F. Hoffmann-La Roche Ltd. Chris Chatham and Kirsten Taylor are full-time employees of F. Hoffmann-La Roche Ltd.

Notes

1. Subsequent to this first study, a second learning criterion has been implemented within Gallery Game to limit the number of trials available if a participant consistently falls below 100% accuracy on learning iterations. Participants may attempt each size of learning iteration (up to 6 images) 10 times, with a maximum of 128 stimuli presentations per learning task.

2. Subsequent to this first study, image category, orientation and swipe direction are now balanced within each individual learning task.

3. Hereafter we use mathematical interval notation for RIs, where brackets represent an inclusive bound and parenthesis an exclusive bound. For example, the [0, 1.5) RI refer to any retention intervals greater than or equal to zero and less than 1.5 days.

Additional information

Funding

Development of the Mezurio app was supported by funding from the Robertson Foundation, NIHR Oxford Health Biomedical Research Centre, Eli-Lilly and F. Hoffmann-La Roche Ltd. Additional support for this study is provided by the Oxford Clinical Academic Graduate School Clinical Lecturer Support Scheme, the Academy of Medical Sciences Clinical Lecturer Starter Grant [SGL016\1079] and the Medical Research Council Deep and Frequent Phenotyping study grant [MR/N029941/1]. The PREVENT research programme was developed with a grant from Alzheimer’s Society [grant numbers 178 and 264].