Rehabilitation in Virtual Reality

Meera -   This week, I'm at the University of East London which is located in Stratford.  I've come along to their psychology department's virtual reality lab and with me is Dr. Paul Penn, lead researcher of the virtual reality research group here.  Now Paul, some interesting work you've been doing here is using things like virtual reality in order to rehabilitate brain injured patients and stroke patients...

Paul -   Yes, that's right.  We're really interested in looking at VR as a means to improve the lot, in terms of rehabilitation and assessment of people that have suffered a brain injury.  A lot of brain injured patients, as part of their recovery, will spend a great deal of their time completely understimulated and not officially in therapy and things like that.  That's really a problem for patients from a neuropsychological perspective because the way the brain responds to injury is determined by the environment in which you're trying to recover.  So in other words, if you have a very stimulated environment, the chances of you getting better functional outcomes and better recovery from the injury will be that much better.  So what we're really looking at is using VR as a means to enrich environments post brain injury.

Meera -   What aspects of brain function do you focus on with brain injured patients?

Paul -   It tends to be things like memory.  So we'll look at how well people are going to remember route information for example.  We look at how well people can remember objects they see in an environment.  We also look at the way people can remember to do things at some point in the future, what we call prospective memory.  So that's things like remembering to close the front door after you've opened it, remembering to attend appointments at certain times.  Very simple things that we take for granted, but are actually very pervasive in everyday life.  And these are the kind of problems that, because they're so intertwined with everyday life, you can't really assess properly in the lab, and that's what VR gives you.  It gives you that chance to put a real world scenario into the lab.

Meera -   The key point of the virtual environments that you create here are that they can be used simply on laptops and desktops which must therefore make them much more accessible?

Paul -   Yes, absolutely.  This is a really important part of our remit.  So really, one of the fundamental criteria that we use when looking at VR is, will this run on just an average, modest-spec PC.  What we tend to do now is what we call a windows-on-world system where the virtual world just appears on a computer screen.  So, it probably looks like a computer game, like you or I would play on a PC game or a Playstation 3 game.  But the difference is that our environments have a purpose other than entertainment.

Meera -   So now, we've got a laptop setup in front of us and it's got a particular virtual environment on it which is a virtual bungalow...

Paul -   So what you can see here is a simple, four-room bungalow with a hall.  The patient's task here is to help the person who owns this virtual bungalow move to a larger bungalow.  The person's task is to go through these rooms; first, we've got a hall followed by the lounge, and then allocate furniture to these new rooms.  So they're engaged.  They're searching around the existing bungalow and they're looking for items of furniture that, for example, belong in the hall.  While they're doing this, what we actually have is a series of three memory tasks.  You can think about the removal task as a kind of distraction task, essentially because that's the way memory works in the real world.  It will be very easy if all we ever had to remember is just what we have to remember, but the problem is we have all these distractions around us all the time, and it's how well we can filter out those distractions that actually gets to the crux of the matter.  That's really what this environment is assessing, with the removal task as the distraction task.  What we're actually looking for is how well they can remember to do three different things.  

Now, you might remember that I told you that remembering to do things at some point in the future is what we call prospective memory in psychology.  Broadly speaking, there are three types of prospective memory.  First of all, we have what we call "event-based prospective memory" which is a kind of memory that's precipitated by seeing something in the environment generally.  So, in the example of the virtual bungalow we have here, when the participant has been strolling around the house, they have to look for glass items, and when they see a glass item, they have to remember to put a fragile notice on it.  For the very simple reason that we don't want the removal men  manhandling and breaking it.  The second type of prospective memory is what we call "activity-based".  This kind of thing occurs when you see something in the environment that itself serves as a cue for the memory.  So for example, if you turn an oven on, that's also your cue to turn the oven off again because you performed an action which should then prompt another memory to perform the reverse of that action.  The task we have in the virtual bungalow here is that they have to remember to close the kitchen door after they've opened it.  Simply because we've got a cat, a virtual cat if you like, in there and the cat escapes if they don't.  The third type of prospective memory is what we call time-based.  This is the category of memory whereby you have to remember to do something at a certain time.

Meera -   So if somebody's got a meeting or an appointment, they need to remember to do that...

Paul -   Yes, exactly.  Or remembering to tune in to a radio show for example.  The idea here is the person opens the front door every five minutes to let the removal men in.

Meera -   Having these three tasks in action then when someone's in this environment, what are you specifically looking for?

Paul -   The extent to which people have actually recalled or remembered to perform the tasks.  So; how many times out of the possible three have they remembered to open the front door for the removal men?  How many of the  - I think it's about eight  - items that are fragile do they remember to put fragile notices on?  And by looking at that kind of data, that gives us an idea of their memory profile, and from that you can extrapolate what kind of problems they might have in everyday functioning.

Computer generated voice -   "...Look for items and furniture to be moved into the hall..."

Meera -   I'm just having a go on this now and I'm walking through the hall; it's reasonably easy to move around.  I'm just using the cursor, the arrow keys on a keyboard to move around, and then just mouse buttons in order to open doors, and pick items.  So, I guess that's quite crucial, making it quite easy to use.

Paul -   Yes, the interface is really important obviously because what we're looking at here is potentially using this environment with people who may not just have memory problems.  They might also have problems with their physical mobility or their dexterity.

Meera -   Having actually tried this environment out on various stroke patients or brain injured patients, what have you found to be the improvement?

Paul -   What we tend to find with this task, we used it on stroke patients for example, is as you would expect really. They're impaired in all three types of memory, they're particularly impaired on the time-based task.  The time-based task tends to involve what we call self-initiated retrieval in terms of - there are no prompts in the environment.  You have to remember to provide your own prompt which is to look at the clock and we find that people who have had a stroke can often suffer with this kind of behaviour.  It's very, very difficult for them to self-initiate.

Meera -   What can you then do with this information to improve their condition or improve their memory?

Paul -   What this kind of information can do is allow this rehabilitation professional to actually orientate the rehabilitation very precisely, to address the problems that person has.  So for example, if they just have a problem with time-based retrieval, there's a technology you can use like personal organizers, maybe iPods, to actually provide prompts at certain points in the day for critical activities.  What you can do, having interacted with this environment, is get an indication of the kind of prompts that someone will need to offset their memory problems.

Ben -   That was Dr. Paul Penn from the University of East London, taking Meera Senthilingam on a virtual experience, to show how simulated versions of real environments can be used to monitor and rehabilitate patients that have suffered strokes or brain injuries.