Scientists at King’s College London have developed a cognitive test that enables clinicians to assess the functional cognition of patients by enabling them to complete complex everyday tasks in virtual reality. The so-called VStore recreates a virtual shopping world that accurately mirrors the in-store shopping experience.
Current methods for the assessment of cognitive impairments such as Alzheimer’s and schizophrenia are time-consuming, costly and do not accurately reflect real-life scenarios (low ecological validity). Virtual reality may help overcome this limitation.
Virtual Reality appears to offer us significant advantages over more traditional pen-and-paper methods. The simple act of going to a shop to collect and pay for a list of items is something that we are all familiar with, but also actively engages multiple parts of the brain. Our study suggests that VStore may be suitable for evaluating functional cognition in the future. However, more work needs to be done before we can confirm this.
Prof. Sukhi Shergill, Director, King’s Centre for CNS Therapeutics
A Brief Introduction to Cognitive Testing
Cognitive tests are often given to patients who show signs of cognitive impairment. These tests are designed to test for mental abilities such as short and long-term memory, language ability, communication ability, concentration and spatial orientation. These tests range from the brief and simple to the protracted and complex.
Most cognitive tests are pen-and-paper based, using a points-based scoring system. They rely on patients completing a series of tasks such as remembering a list of words or following instructions.
The Mini-Cog and MMSE (Mini-Mental State Exam) are two frequently used tests.
During the Mini-Cog, the patient is asked to remember and repeat the names of common objects. They are also asked to draw a clock face displaying a specific time with all the digits in the right place.
During the MMSE, clinicians ask patients a series of questions. A maximum score of 30 points is achievable. A score between 20 and 24 points could point to mild dementia; A score between 13 and 20 may indicate moderate dementia; Severe dementia occurs below 13 points.
Other cognitive tests include Test Your Memory (TYM), the 10-point Cognitive Screener (10-CS) and the 6-item Cognitive Impairment Test (6CIT).
It is recognized, however, that these tests carry certain limitations. The patient’s education level may, for example, influence the outcome of these tests.
For this reason, computer-based testing is gaining in popularity. One advantage of such testing is that it can be administered in the same way for every patient, thereby eliminating subjectivity.
Combining these tests with traditional paper-based testing can provide clinicians with a more rounded assessment of patients’ conditions. The U.S. Food and Drug Administration, for example, has approved several computer-based testing devices such as CognICA, Cognision and Cognigram.
Using Virtual Reality in Cognitive Testing
The research team was led by Sukhwinder Shergill, Professor of Psychiatry & Systems Neuroscience at the Institute of Psychiatry, Psychology & Neuroscience (IoPPN) at King’s College London. The team recruited 142 healthy volunteers aged 20 between 79.
They wished to validate their VR-based testing system, Vstore, against Cogstate. Cogstate is a well-established computerized cognitive battery that has been designed to assess multiple cognitive functions.
To this end, they selected 8 Cogstate cognitive domains for benchmarking. Namely - attention, working memory, paired associate learning, verbal learning, executive functions, processing speed, technological familiarity and age.
More specifically, the team wished to map the contributions of Cogstate cognitive functions to variances in VStore performance. For this, they developed linear regression models to map Cogstate tasks to VStore outcomes.
VStore was developed in collaboration with Vitae VR. It incorporates a 30-minute evolution, including instructions, orientation, practice and assessment.
At the start of the exercise, the volunteers were read out 12 items from a shopping list by an avatar standing at the entrance of a virtual minimart. The minimart was stocked with drinks, snacks, food and toiletries.
To make it appear even more realistic, virtual fridges were filled with drinks and sandwiches and freezers with frozen meals. Moreover, there were checkout and self-checkout counters and a coffee shop. In all, there were 66 items organized into nine categories.
The initial task for the volunteers was to memorize and recall the 12 items from the shopping list presented by the avatar. They were then given the shopping list and instructed to move through the minimart and collect as many of the 12 items as they could recall.
They were also instructed to pay for the items and order a drink from the coffee shop. Once this was done, the task was completed.
The study proved that functional tasks that more closely resemble real-world scenarios engage a broader range of cognitive functions when compared to tasks performed in standard cognitive testing.
These are promising findings adding to a growing body of evidence showing that virtual reality can be used to measure cognition and related everyday functioning effectively and accurately. The next steps will be to confirm these results and expand research into conditions characterised by cognitive complaints and functional difficulties such as psychosis and Alzheimer’s Disease.
Lilla Porffy, Research Student, IoPPN
References and Further Reading
Shergill, S., et. al., (2022) A novel virtual reality assessment of functional cognition: validation study. Journal of Medical and Internet Research, [online] Vol. 24 No 1. Available at: https://www.jmir.org/2022/1/e27641
King’s College London. (2022). A virtual reality ‘Shopping Task’ could help test for cognitive decline in adults. [online] Available at: https://www.kcl.ac.uk/news/a-virtual-reality-shopping-task-could-help-test-for-cognitive-decline-in-adults
Tsoi, K. et. al., (2015) Cognitive tests to detect dementia: A systematic review and meta-analysis. JAMA Internal Medicine. [online] Available at: https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2301149