Testing behaviour using virtual reality

C-PLACID stands for the computational platform for the assessment of cognition in dementia. So really it’s a big project that’s saying, can we harness the power of computer science to help us with what’s, up to date, been a kind of paper and pencil task of assessing people’s thinking and memory skills in dementia? Today we’re in the digital cave, which is a facility within the heart of UCL computer science. So it’s a part of UCL campus, which I don’t usually venture into. We’re over on the clinical site over in Queen’s Square. But we’re planning to bring patients with frontotemporal dementia and other conditions here to participate in a series of virtual scenario settings with some of our computer science colleagues.

One section of the project is saying, can we use virtual reality scenarios to try and help us to assess better so-called social cognition changes, or changes in behaviour and personality, which are particularly prevalent in people with behavioural variant frontotemporal dementia? So we’re constantly in the clinic hearing, particularly from carers, about really quite distressing changes in people’s behaviour. So, for example, people being less empathetic or behaving in a way which is completely out of character for them, perhaps being socially inappropriate, swearing more, making comments they would never have previously made. But as the clinical team, we don’t see that directly manifested. Often the person with the dementia is just sitting there, and you’re hearing about the second hand.

So what we’re trying to do is to say can we use virtual reality to create some social scenarios where we can be a bit more in control of the social setting and see how a whole set of different patients with frontotemporal dementia react in certain situations? It’s basically, it’s a walk-in room. And there are four surfaces that have projectors behind them. And we can project images onto these surfaces, anything we want from the computer. And all the projectors are synced in time. So we can project images of a virtual world. And what happens is, a person walks into the room with these shutter glasses in.

And the projectors will fire the left hand side eye image, while the right hand shutter glass is closed, and then alternate. And it does this quite a few times a second. So you get a 3D image, basically. You walk into this 3D world. And it’s quite immersive because you’re surrounded by these images. The system tracks your head on these balls when you wear the glasses. So wherever you look, the system will readjust the image. So the perspective is always correct, or nearly correct, to your eye line. There’s also spatial audio in there. So we’ve got, I think, it’s about eight speakers, which project audio at different points.

So we can create a very immersive environment and really make somebody believe that they’re somewhere else. That’s the idea really. So one of the ideas with this scenario– this is our pilot work. We’re just at the outset of this project. And the idea is to try and create a cafe or bar scenario where people come in and they experience– they try and make friends with an avatar, a waitress character in this particular situation. They see her experiencing, either directly or indirectly, other people’s emotions, so a happy or an angry customer, for example. And also witness some slightly odd characters moving across the scene.

Because we’re interested to see whether we can capture different aspects of behavioural reactions within certain social situations. So, perhaps there are people with behavioural variant frontotemporal dementia more likely to react in a odd or overly inappropriate way to certain scenarios– to look at certain people more than those of us without dementia might, to fail to demonstrate care or empathy toward someone who might, for example, trip over and be hurt. And also to see if they perhaps miss things which for the rest of us are very socially salient. Say someone else’s distress or upset, for example. At the moment, virtual reality is quite an expensive technology. So the idea of this being rolled out to every memory clinic it’s unlikely.

But the technology is getting cheaper. And so one output I can really see for it is certainly for clinical trials as we begin to get disease modifying treatments for frontotemporal dementia. Because we know social change is one of the earliest features, or earliest symptoms, I can definitely imagine us wanting to use this as an outcome measure for trials. So having some standardised social situation you put people into and then see if measures of the normality, if you like, or reasonableness of the social behaviour can be used to track whether the treatment is being effective or not.