Health in a Handbasket - Episode 5: Are robots the future of rehabilitation?

Peter Snow is a lecturer working in rehabilitation engineering using virtual reality and robotics.

Peter develops systems to help people with phantom limb pain and other conditions. He is also using gaming software to create training tools for surgeons, allowing them to practice complicated procedures in a safe environment. 

We chat to Peter about how he started his career (from winning RoboCup to surviving the recession), to his future ambition to work within the NHS. 

About Peter Snow

Listen now:

• Apple Podcasts
• Spotify

View the episode transcript

Ferdouse Akhter  0:05  
Hello and welcome to health in a handbasket, your podcast about the sexy world of Healthcare Engineering. I'm Ferdouse Akhter and I'll be your host. I'm the Marketing and Community Manager at UCL's Institute of Healthcare Engineering. And although I don't always understand what's written in the research papers published by academics, I know that what we're doing in the world of healthcare engineering is important and impactful. And I want to share that with you by speaking to those who know a bit more about him than me.

So today, we're picking out the topic of virtual reality and robots from our handbasket. 20 years ago would never have heard of the word virtual reality, and thought robots were bad things out to kill Sarah Connor. But here we are in the future. And technically the Terminator travels back to 2029. So we're not that far off. Nowadays, we can buy virtual reality handsets from Curry's, and robots are helping with surgeries, and in Amazon warehouses. VR isn't just for enhanced gaming experience, VR or virtual reality can be used to help treat us - robots aren't taking over the world! They're also here to make our lives better. And that's where Peter comes in. Peter Snow is a lecturer at UCL in the Institute of Orthopaedics and Muscular Skeletal Science. He has a background in computer science. And then he went on to do artificial intelligence before completing a PhD in rehabilitation robotics, the kinds of robotics that helps you recover from injuries. Currently, he's working on robots that can alleviate phantom limb pain in Upper Limb amputees. And Peter's a busy boy, because he also developed surgical simulations using VR and robotics. Let's start off with what is phantom limb pain.

Peter Snow  1:51  
So phantom limb pain is a painful sensation experienced by those who generally have had a traumatic amputation, you get the Phantom Limb sensation, which is the fact that you feel that your limb is there, even though it's not. But for those who have had a traumatic amputation, they generally have a feeling painful sensations associated with that limb.

Ferdouse Akhter  2:13  
So you can still feel like your fingers and stuff, even if it's not there.

Peter Snow  2:17  
Yep. So you can feel your limb in like weird positions, you can feel your hands could be like crumpled up, it could feel like your hands being frozen or on fire or being crushed. But what's interesting is the fact that those who are born without a limb, they have the sensation that their limits there, but they they rarely have the actual painful sensations that actually come with that with that phantom limb.

Ferdouse Akhter  2:43  
Oh, that's interesting. So even if they've never had, like, I don't know, half their arm, they can still feel it. 

Peter Snow  2:49  
Yep. 

Ferdouse Akhter  2:50  
Is that like a social thing? Or is that a psychological,

Peter Snow  2:54  
um, it's probably a mixture of the both. It's difficult to tell. So I think when it comes to things like limbs, it's all connected to the brain and the sensations within the brain. So those those haven't changed, even though you don't have to have a limb, that for those people who obviously born without a limb, they obviously still have those sensations in their brain. But they obviously don't physically have the limb, which is why they probably get those feeling that their hand is there or the limb is there. But they don't have the pain. But it's a very difficult area. There's lots of back and forth between different theories. And there's a lot of arguing within the community. But that is changing over the course of the years. Neuro imaging is helping a lot.

Ferdouse Akhter  3:39  
What is Neuroimaging? 

Peter Snow  3:40  
Things like MRI scanning to look at what's happening in the brain.

Ferdouse Akhter  3:45  
Okay, cool. So, are there any treatments for phantom limb pain at the moment?

Peter Snow  3:50  
There's no effective treatments. One of the the two most common ones are drugs. And the other one is more surgical revisions. So that's if you've had an amputation, if there's other issues, they might need to give that patient further surgery. 

Ferdouse Akhter  4:08  
Do you mean like infection and stuff? 

Peter Snow  4:10  
Could be infection, it could be if they have something called a stomping neuroma. And that's when if you've had an amputation, your nerves stop bundling, because they get tangled up that can cause pain. And that would mean that they'd have to have a operation to actually remove that nerve. So yeah, it's a pretty serious condition. And then other than surgical interventions, you can have things like obviously, pharmacological interventions, but you can have side effects such as, you know, issues with mood, sleep pain, it could cause more longer term issues. So in terms of, of effective treatment, that doesn't tend to be any number one way of treating phantom limb pain, but one of the treatments which is non invasive which is what the research is roughly based on is mirror box theory. So that's using what we call a visual surrogate, what you do is you'd have a big mirror, you'd put your intact limb and your amputated limb. And this works for both lower and upper limb.

Ferdouse Akhter  5:16  
So on each side, you'd put one hand kind of thing. 

Peter Snow  5:19  
Yeah, And then what you're seeing in the mirror image is what the mirrored image of your intact limb. So we know that has very success. Sometimes it works, sometimes it doesn't. And what you're seeing is you're seeing that the mirrored limb, and then when you move, you're seeing obviously the movements. So it looks like the your limb is there. And for some people, it works. Some people it doesn't. But you know, it's non invasive. So if it works, then it's good for that participants. But again, similar to the drugs, it can be quite varied in the effectiveness,

Ferdouse Akhter  5:55  
Can't really do it all the time. To come around in the middle of a restaurant kind of thing. And

Peter Snow  6:00  
Absolutely, yeah, it's quite quite a big thing. But that's what initial technology. So what happened was, when VR started to get cheaper and cheaper, one of the ways that people started to translate this is to use VR headsets directly. So you would have sensors on your intact limb.

Ferdouse Akhter  6:21  
What kind of sensors that I guess, like sticky things on your hand.

Peter Snow  6:24  
Things are like a like a like a glove with some sensors on. And then when you move your intact limb, so exactly the same as what you would do with the mirror, but then you'd see it virtually. So you have

Ferdouse Akhter  6:35  
a headset on, you put stuff on your hand. Yeah. And then that kind of like helps you.

Peter Snow  6:41  
Yeah, so what you'd see is the movement that you made in your intact limb, exactly like the traditional mirror box therapy, you'd see on the on the amputated side, in virtual reality.

Ferdouse Akhter  6:52  
And that's kind of what you're working on at the moment. 

Peter Snow  6:55  
Similar, the one or two big issues of the fact that number one, you're only visually seeing it, you're not getting any what we call tactile feedback. So obviously, vision is a sense but so is force feedback. So the the sense of touch. And that's what's most one of the things that's missing. The other thing that's missing is that your brain clearly knows that it's not the affected side, which is making those movements. Your brain knows that Okay, well, I'm actually moving my intact limb and that I can see it on the on the affected limb. So what we did is we use the idea of virtuality. But by using a robot, which has previously been used for stroke rehab, we can actually make you because you're the limb that you're moving around, is the effective limb.

Ferdouse Akhter  7:45  
You get like something attached to you.

Peter Snow  7:48  
So there's like a, when we say a robot is not as like fancy or doesn't look as sexy as like a normal robot. It's basically a big black box with like a big black cylinder arm - it's very much designed for rehab and for purpose. And then you have like a plastic what we call gimbal. So a gimbal takes the the orientation. So it's basically like a plastic cuff. So you place your arm on the on the cuff. And then what that allows you to do is it allows you to get the movement and the rotation, and we take that rotation in the movement, and we put it on the virtual avatar. So it looks like so for example, we're sitting in front of a desk, you, if you were a participant on a system, you'd have your arm on the robot, and as you're moving around, you'd see your affected arm moving around as if nothing was wrong.

Ferdouse Akhter  8:48  
So you put like sensors on like, whatever's left of the limb. And then you track the muscle spasm activity. Yeah. And then that kind of like is, is on a virtual stimulation, like it's in their VR headset, you can see themselves so you don't actually put anything, you don't attach anything onto what's left of the limb. Like a robot, you know, like how you see like prosthetics and stuff, you don't do that

Peter Snow  9:13  
We don't and that's that's a that's an important point. Because having a prosthesis fitted takes time, you need to have the wound to heal. There's lots of issues with you know, different variations of a prosthesis that can take time. The benefit of using the system that we developed, is the fact that you can use it fairly fairly soon after an amputation. And that's important because it could potentially stop the painful manifestations happening. So we have that we have the robot, we have the VR we have the sensors for the muscles, but the robot does something else. It provides what we call force feedback or haptic feedback. And we've demonstrated haptics for 10 years 15 years. And it's always one of those things where it's really easy to show them and actually to tell. But basically, a haptic robot allows you to feel virtual objects physically. So if if you're on the system, and you have like a like a jug of water, if you bang into the jug of water, the virtual jug of water, we can send commands to the robot to stop you from going through the water. If you pick up using the virtual opening and closing of the hand, from the muscles, you can pick up that that jug of water and you can feel the weight that's being applied on the actual robot. So that is one of the key elements of the system is to allow you to actually feel the virtual objects physically on your affected arm.

Ferdouse Akhter  10:47  
That's so interesting. So what stage is this work currently at?

Peter Snow  10:51  
So the work with MBTs was part of my PhD. So that's carrying on. So we've done about in total 18 to 20 participants,

Ferdouse Akhter  11:02  
You're not doing your PhD at the moment, what stage in your career, are you at now

Peter Snow  11:07  
Postdoc lecturer. So like, I guess you could say early career?

Ferdouse Akhter  11:10  
Okay, so that comes like a few years after PhD, right? Okay.

Peter Snow  11:14  
So with the initial work was with amputees, and we've now carried on to the pain populations, including upper limb nerve damage, who are potentially at risk of having an amputation, who still have pain,

Ferdouse Akhter  11:30  
What's upper limb nerve damage?

Peter Snow  11:34  
So for example, we had one participants who have unfortunately had an accident, and it basically crushed their hand. And they had no movement of their other fingers. And they had pain as a result. So putting them on the on the system and the protocol for the potential treatment. And we've also had people with spinal cord injury who have had phantom limbs, not so much with pain, the more troubling sensations. We've had a lower limb, nerve damage participants. And we've recently started UK study looking at shoulder pain, which is obviously a massive issue as well.

Ferdouse Akhter  12:13  
Okay, so what's the end goal of all of this? Is it currently being used at the moment? How can people access this?

Peter Snow  12:19  
So at the moment, it's purely research, but we're in the process of applying for grants to create innovation clinics and hubs to help technology with people with pain,

Ferdouse Akhter  12:30  
So, like, people would go to those hubs and use it? 

Peter Snow  12:34  
the main goal of the actual work is to include it in to the NHS as part of their service delivery plan. And also to have a clinical system and maybe a more low cost version, or a VR only or VR and small low cost robot version that people can take home.

Ferdouse Akhter  12:55  
VR has come down a lot in pricing nowadays, like, like I said earlier, you can buy VR headsets, off Currys, and so on. So how much are we talking about these low cost systems?

Peter Snow  13:05  
Well, I mean, you can even use a mobile phone, like a VR headset, there's that the price of headsets have come down massively, since the Oculus was released. So back in the 90s, and even the early 2000s, it would have cost 1000s of pounds for a headset. But with companies like Oculus, developing the Oculus Rift, and others, it's massively come down in price. So you're talking even a couple of 100 pounds for a fairly good headset.

Ferdouse Akhter  13:41  
So how did you get into this field?

Peter Snow  13:42  
Well, initially, my background was in computer science. So I first had a computer when I was four years old, my parents, my parents use my own savings for my grandparents, because they thought it'd be a good investment. And well, it turns out to be right. So I've always been interested in computing, and I did Computer Science. And then I took modules in artificial intelligence, some quite interesting modules on social intelligence in animals and robots and quite interesting

Ferdouse Akhter  14:12  
stuff. Tell me something interesting. It sounds very interesting.

Peter Snow  14:15  
So I mean, one ant is fairly useless. But 1000s of ants are very, very intelligent, and they can create pathways they can use, guess what we call social intelligence in order to solve problems. So taking more bio inspired ideas, and then putting them into computers and artificial intelligence. And as a result of that, I decided, okay, maybe I want to do a Master's in this and I did a Master's in Artificial Intelligence with robotics. And then during that we did the Robocop which was getting robots to play football.

Ferdouse Akhter  14:49  
So you pit two robots against each other, or two robot families against each other.

Peter Snow  14:55  
Robot families - teams of robots - family as well if you if you're talking about football, even though I'm not a big football fan. And then I got involved in that because my my final project for my for my masters was looking at stability controls in how we make humanoid robots meet how you basically make them more stable and walk faster, and not fall over. And that was applied to the football team, the Robocop Football Team and we did very well, we won the German open in 2018, which is like the European Cup. And we came joint second joint first sorry, in 2009. And we came second in the world in 2000. Oh,

Ferdouse Akhter  15:40  
second in the world. Pretty impressive.

Peter Snow  15:43  
And then since then, the recession hits and I needed to get a job. So I ended up working as a as an IT technician for a few years. But then I met my PhD supervisor who is now my line manager, Rui,  and we had a few projects that he'd apply for. And one of them was the project on phantom limb pain, which was funded by the Ministry of Defence,

Ferdouse Akhter  16:07  
Ministry of Defence, what because of the wars and stuff. 

Peter Snow  16:09  
Yep, so this was 2012, which was the height of Iraq and Afghanistan. And obviously, you were having a lot of people come back with missing limbs and phantom limb pain. So that was one of the the calls for funding that we applied. And we we got, thankfully, and the rest is history.

Ferdouse Akhter  16:27  
I think that's super interesting that you, it wasn't like a linear trajectory for you. It wasn't, you know, master's, PhD postdoc, it was, you had a time in in IT, right? It support support. So you, you left the world of academia, and you did something else. And that's so interesting, because I think a lot of people think that you stay in academia, you just live in academia. And that's all there is. So how has your own personal experiences shaped your journey like all of those things happening?

Peter Snow  16:58  
Well, interestingly, about a year ago, I had to talk to undergraduates about Healthcare Engineering and my journey. And one of the things it actually occurred to me was the fact that because I was born three months, too early, premature, I ended up being put in a incubator. So the only reason why I survived, and I'm here today is because of Healthcare Engineering. Because if that incubator hadn't been developed, then I might, I wouldn't be here. 

Ferdouse Akhter  17:25  
So you're born to do Healthcare Engineering? Pretty much.

Peter Snow  17:29  
Yeah. Yeah. And then subconsciously, I kind of gone down that path. And I think, yeah, you kind of like fall into the cracks, and you kind of make your make your own way there. Yeah, there's no, there's no path to working in healthcare engineering, or, or research, you know, you kind of just fall into it. So for my master's work with the robots, making them walk faster, I initially wanted to actually do that. And obviously, the recession hits, but then I met Rui,  and then we looked at different types of limbs, so instead of the lower limb, and that's how he kind of got involved in that project.

Ferdouse Akhter  18:04  
I liked the stuff you said about the fact that you have more like bumps on the road and stuff like that.

Peter Snow  18:09  
So yeah, there was no kind of like, no set trajectory into healthcare engineering. So when the recession hits back in 2009, and then it really bit bit in 2010. That was right when I was finishing my, my, my masters, you know, I wanted to go into research to carry on, and all the research budgets around the world just, you know, tightened up, and you know, you did feel kind of feel like a, ya know, you've got delay again, and it is frustrating, but you you do get there and you know, I'm a big believer, and you know, what is meant to be will be, basically, you know, so, yeah, it's no simple path. But yeah, you get there in the end. And during the PhD, I was diagnosed with testicular cancer, right, when we got the ethics to actually start, so we had to delay that. So yeah, nothing is straightforward. If you're straightforward, it will be it will be boring. So but you kind of get there in the end, and yeah, things happen for a reason.

Ferdouse Akhter  19:07  
Yeah, definitely. I mean, yeah, kudos to you, for overcoming all of that stuff.

Peter Snow  19:12  
And actually, you could probably say, health care engineering helps again, because obviously, you know, the CT, MRI, the surgery and everything involved in the cancer as well, again, got me through that as well. So you can kind of book in the, you know, being born and getting through a part of my life using Healthcare Engineering.

Ferdouse Akhter  19:32  
I think that's the crazy thing about healthcare engineering is people don't realise how it impacts their regular lives on a nearly everyday basis. And like you said, like, things like CT scans and an incubator and stuff like that. You don't think of it as Healthcare Engineering, but is someone thought about it, someone created it. Someone saw a solution to a problem, and that's why you're able to live and you have these medicinal technologies to help you live.

Peter Snow  20:01  
Absolutely. And actually a lot of the the health engineering is actually developed here in the UK and my wife who actually works for a healthcare company, for MR. You know, I get to talk to the people involved in their company as well. And I see the similarities between utterly the work that we we kind of do and the work that she does, because that's the commercial. And obviously, we're not commercial. But we do have commercial aspirations. So it's seeing those difficulties and seeing what's happening in healthcare technology in business versus research as well as is quite interesting.

Ferdouse Akhter  20:37  
Definitely.

You're also working on another project, like I said, busy busy boy, and you're working with VR robotics again, but with surgeons.

Peter Snow  20:57  
That's correct. So the the technology we use VR and Unreal Engine, which is the software we use to develop the virtual Unreal Engine.

Ferdouse Akhter  21:06  
Unreal Engine?

Peter Snow  21:07  
So that's what they use to make video games, we use a same or similar version of the system we use for rehabilitation, but we use it for training surgeons.

Ferdouse Akhter  21:17  
So they play a game?

Peter Snow  21:19  
Kind of but the game is a is a slice of the actual certain procedure that they actually will be doing in real life. So we can take patient data from CT and MRI, we can put it into the simulator, and we can get them to practice on the on the simulator. And the benefit of that is the fact that traditionally, they would use a cadaver, or they'll use like a like a phantom. A physical phantom as in like a, an approximation, so be like a, like a, like a plastic version. And the benefit of using a VR system is the fact that it will robots like data, we can see what you're doing, how you're moving, if you're cutting too much, we can even see what you're looking at as well. So we use it for training as you get like feedback. And we get the same haptic feedback as we do for the rehabilitation. But then we can do different types of feedback. So bone or a tumour within like, for example, a liver, when you're cutting things. So we doing a series of procedures, and we're doing some usability studies with those as well.

Ferdouse Akhter  22:28  
So at the end of it, I guess the surgeon gets like a report saying you did this well, and you did that not so well.

Peter Snow  22:34  
pretty much. And then we can also look at the what we call the gold standard. So that is a consultant will perform the procedure on the simulator. And then we can see how trainees compare to the actual consultant. 

Ferdouse Akhter  22:48  
This is like before they get into an actual surgical ward?

Peter Snow  22:52  
pretty much yeah, we can practice them as many times and it's completely safe. And it's mainly the that the manual skills of cutting, and pressure and how much you cut, and then the actual procedures, the steps involved in the actual surgery that they're practising safely.

Ferdouse Akhter  23:09  
So fewer mistakes and all that stuff. Yeah. Okay, so what's next in your very enlightening career?

Peter Snow  23:15  
Well, hopefully, I can carry on working in this field. At the moment, I'm technically part time. So hopefully, we can try and move move myself into a onto a full time contract to be a permanent lecturer. But also I'd like to do more clinical work. So I'm in the process of applying to be a clinical scientists,

Ferdouse Akhter  23:34  
How's that different from the kind of scientist now? 

Peter Snow  23:37  
This is more heavily involved with the with the NHS or with healthcare providers. And what this involves is allowing me to work independently within a hospital, in say, for example, a innovations clinic, such as using technology like VR, or robotics for pain or patient populations. And then from that, it's going to be a long procedure. I'm hoping to do more training to hopefully become maybe a clinical academic or something along those lines. And so

Ferdouse Akhter  24:11  
currently, you're working at UCL. And so then you're going to go on to be working in a hospital.

Peter Snow  24:18  
So in theory, UCL offers clinical academic roles. So those are kind of split between hospital and UCL.

Ferdouse Akhter  24:27  
Okay, I didn't know that you could do that. I didn't know I thought you had to just work at university or work at a hospital.

Speaker 2  24:33  
Yeah, it's not not commonly a route that people go go down or that people are actually aware. But if like me, you've got a lot of experience in working in a hospital and working with intervention, which actually, on a day to day basis, you're working with patients, you know, you're technically eligible to actually go down that route.

Ferdouse Akhter  24:53  
Okay, cool. super interesting. Thank you for coming along. Peter. Let's go. I love loads of things. You doing some great work. And I think it's a reminder to everyone that someone's academic career or any career really isn't linear. There's there's loads of roads that you can take towards becoming a clinical academic.

Peter Snow  25:14  
Yep, one day, or different pathways which could support that role or Yeah, yeah. So thank you. Thank you for having me.

Ferdouse Akhter  25:27  
Health in a handbasket is produced by UCL Institute of Healthcare Engineering and edited by Cerys Bradley. The issue of Healthcare Engineering brings together leading researchers to develop the tools and devices that will make your life better. We're using this podcast to share all their amazing work taking place. You can learn more by searching UCL Health in a Handbasket or following the link in the show notes. So share with your friends and family if you found this interesting. We're available everywhere, especially where you just listen to us