To develop and validate a gamified, tablet-based task that allows assessment of short-term memory both remotely and in clinical settings.

Global demographic changes have led to an increasing aging population and patients presenting with memory complaints. Their aetiologies can range from mood disorders to early symptoms of cognitive decline in Alzheimer’s disease (AD), thus diagnosing becomes a great challenge. Despite a great interest in early detection of AD, tests commonly used in the clinic often lack sensitivity to detect early symptoms. Furthermore, most assessments require patients to come to the clinic and see trained personnel. There is an obvious need for a portable task that can be run at scale outside of clinical settings, does not require specially trained personnel, yet is able to detect early signs of cognitive decline in AD.

We developed a new, engaging game based on a previously developed lab task that has been validated in 1800 healthy participants across age groups, and shown to be sensitive to memory deficits in asymptomatic individuals with familial AD due to autosomal dominant genetic mutations. Participants viewed a number of randomly located star constellations presented briefly on a tablet screen. After a short delay, they were presented with one of the constellations they had previously seen together with a distractor (previously unseen) constellation. They were asked to touch the constellation they thought they had seen in the memory array and drag it to its remembered location.

This test allows assessment of participants’ ability to remember objects and their locations, thereby providing a quantitative measure of memory precision. Embedded in an established app for cognitive assessment, large longitudinal datasets can be collected with ease.

Datasets collected with this task can be examined for further disease specific patterns of short-term memory impairments, such as found for familial AD in previous work.