Hopkins Hour event: 'New Technologies in Rehabilitation'
Hello, everybody. Hi everyone. And welcome, to the fourth hopkins, hour in a virtual environment. We hope you're all doing well given the current challenges, presented by koben19. We would like to extend, a warm welcome, to all our hopkins, center members, partners, collaborators. Who are joining us today. Before we begin, today's, proceedings, i would like to acknowledge, the traditional, owners of the land on which we meet virtually, today. My home office, is on the land, of. Turbo. People. And i would like to pay my respects, to elders, past, present, and emerging. We encourage. All attendees. To research, the traditional, owners of the land, on which they live and. Work. For those of you who don't know me i am dr david painter, a cognitive neuroscientist, by training. Research, fellow at the hopkins, center, and co-lead, of the beehive, flagship. Project. Exploring, virtual reality, software, and hardware. In the rehabilitation. Setting. I'm your host and research presenter, today. And i'm joined by my esteemed colleague. Dr camilla shiroto. Advanced, queensland, fellow, and research, lead at habitat. And invited, guest speaker, mr miker, ching. Business manager. G-tech, neurotechnology. Hong kong limited. As well as the event team, manning, our question. And answer. Function. Just a few housekeeping. Items, if you do experience, technical, difficulties. Please feel free to put a question in the q a function, which is on the right, toolbar. Macaula, kerr has put up a few. Reminders. For people in there and one of our team members, will try to assist you if possible. Internet connections, obviously, will, vary in this environment. There may be a delay. They say between. Seven and ten seconds or it could be more. For you on your end, it is seamless. So although, you may be on a delay. It should not buffer, to the current time, frame, so you won't miss anything. If you have any technical. Trouble, or, a need to leave the event we are recording. So, feel free to do so. You can also pause at any time, without losing your place, we're going to distribute, the full recording. In the coming days. So you will have access to that there will be an opportunity, to ask questions throughout the presentation. Using the q a function. Which should be on the right hand side of your screen. The questions will be moderated, by our team, and we will address, them, at the end of the presentation. To post your question, just click into, the q a function. Enter your name. Or you can also, stay anonymous, if you wish to do so. Your question, will be published. And viewable. To the audience. With presenters. Addressing them at the end of the presentation. The structure, of today's, proceedings. Is that we have 15 minutes each for david, uh. Mika, and, camilla. And we have. Hopefully, around 15, minutes of. Question, time. So without further ado i will present, our first speaker. Myself. And i'm going to be telling you a little bit about. Virtual, reality. In particular, i'm going to be talking about an ai, enabled, spatial, attention assessment, and training system. This, is a concept, or an invention. That was, submitted. To, a. Bionx queensland, 2020, challenge, as a competition. Entry, in the artificial, intelligence. Ai category. And actually we won that category, uh so i can tell you a little bit about that it's quite. Exciting. It's, based, around virtual, reality. Technology. And there isn't actually a very good definition, at the moment, uh but this is the best, that i could find so far. Uh virtual, reality. Is the computer-generated. Simulation. Of a three-dimensional. Image or environment. That can be interacted, with in a seemingly, real physical way by a person, using special, electronic, equipment. Such as a helmet with a screen inside, or gloves, fitted with, sensors. It's been a relatively, brief history. The, first sort of concept, of that, was the uh late. 1970s. And there were various. Uh iterations. Um but it took around. You know 30 years or so the early 90s, before the first vr systems, came. And today. Uh 2016. And beyond. We have a variety, of vr. Technology, that you can use.
In Your house. The. Various, systems. That you can have you can purchase. And some of these are within. A sensible, price point, some of them are self-contained. In that the computer, is on the inside. Some of them. Require, a computer. And some of them require, a computer, an external. Tracking, station, so there's a trade-off, between. Portability. And, fidelity. The more, cables. And objects, you have obviously, the higher. Fidelity, that the experience, can be. There's a variety. Of different. Must-have, applications. At the moment, one, is, for example, beatsaber. Another. Uh. Is, tetris, effect, it is the definitive. Version. Of, tetris. And there's a lot of work in the indie, scene. Developing. Prototypes. And various, small, scale, games, that are very easy to construct. In. Game engine environments. I, am co-leading. With professor, heidi zeeman, the beehive, initiative. Uh, which, really takes, as a starting, point brain injury population. Who may have cognitive. And other. Deficits. And the idea, is to enrich, their experiences. Thereby. Creating. Opportunities. For neuroplasticity. And recovery. Following, brain injury. And virtual reality. Is actually, one of the most promising, technologies. In that space, customizer, headset. Allows us to capture. Motion, allows us to track attention. We can play games, we can do experiments. We can. Tap. Basic. Um. Cognitive, abilities, like memory attention, learning cognition. At a level directly, relevant to everyday life and everyday, um, behavior. So this invention, that we proposed. Was part of a bionics queensland, challenge 2020, bonus queensland. A, local, not-for-profit. That. Tried to accelerate. Growth, of bionics, technologies. In queensland. The most, uh, famous, bionics technology, is the cochlear, implant. And there's various other things you can do in the domain of say, exoskeletons. Brain computer interfaces. And so on we submitted to the third category, ai, enabled, bionics. So we propose, to integrate, virtual reality, with artificial, intelligence. Machine, learning approaches, which is a. Way to understand. And, analyze, patterns of information. In data. We won fifty thousand dollars that we're about to spend. And we were featured. On the news. Um so it was pretty exciting. Uh, so the the invention, is all around the concept of spatial, attention, it's a fundamental, cognitive, ability. It allows, you to find objects, in cluttered, environments. For example. Imagine. You're looking for your keys on the desk. You can search various locations, on the desk, in turn. And spatial, attention, allows, you to perform those detailed, searches, in those locations. There's of course unless you have an injury to the brain, you can have, a, spatial, ninja, neglect. Sorry. And with spatial neglect. This is an, uh. Difficulty. Paying attention to objects on the left side of space, so in that case your search for the keys may look something like this. You can only see objects on the right hand side or more accurately, only pay attention to objects in the right hand side. It's associated, with a series of right hemisphere, brain networks, that. We know are important, in attention, particularly in the parietal, lobe which is important for representation. Of, three-dimensional. Space. Here is a person, with neglect. And as you can see they're looking towards, the right side. It's as though the left side. Of the world doesn't, exist. For them. In clinical, settings, rehabilitation. Units and hospitals, for example. Neglect, is assessed using pending paper tests. Here is an example, the person, may be asked, to, draw a clock. And if the person draws a clock like this. With numbers, on the right hand side. You can conclude that this person has neglect, it's just sort of the left side of the clock doesn't exist. There are, a variety of these different pen and paper tests, and they all have limitations. Particularly, they can't measure neglect severity, they cannot detect several cases. And they can be cheated by the patient, for example by rotating, the page. That's why we propose. A so-called, attention, atlas. It's a virtual reality. Concept. For measuring, attention, in three-dimensional. Space the person wears a vr headset.
They Use a controller. Uh they're looking for objects. In the visual field we present some letters here, the person's task for example, may be to find the t. Uh. Among. Else. And we can present those objects, systematically. Around the person, and measure how the person pays attention. To their surrounding, three-dimensional. Space. This is, an example. Of what it looks like from the first person, perspective. There's a queue. Indicating, the person. Is defined at t. And the person is using a, hand controller. To indicate, the selection. Uh, and they're looking around here until they find a t. Uh this is me, expert, performance, i'm doing this incredibly, fast but you can imagine, that, if you have a brain injury it may, unfold, much. More. Slowly. But while the person is doing that we can measure attention, using the headset. With the eyes we can measure where the eyes are looking we can measure where the hand is pointing. And if we do so that creates. Attention, maps. Here's an example of an intentional, map from a person. With. A hypoxic. Brain injury without neglect, and as you can see. These, dots. Reflecting. Episodes, of momentary. Attention. Are distributed. Over time across the virtual space. Thereby, indicating. Even attention, to left and right. Here is an alternative, representation. Of the data, we can analyze. The. Patterns, of attention. Separately, for the headset, the hand and eye gaze to give us a detailed. Picture. And we can. Precisely. Examine, those, using, statistical, methods. For example. In the case of neglect, it's relevant to ask if there's a. Left right. Symmetry, because you predict a left right symmetry, but you can also look for. Vertical, symmetries. For example, and the overall distribution. Across space. In addition, to the attention, atlas, we can combine, that with a cognitive, prosthetic. The attention atlas, measures. Cognitive, function. We can design, scenarios. Sorry my. Uh scroll. Scroll wheel is super. Sensitive, today. Uh we can. Combine that with scenarios, designed to improve the person's attention, here's an example. Of a person wearing a vr headset. And they're in a driving, game. That we can modify, the parameters. To. To enhance. The ability of the person, to pay attention. Uh in a dynamic. Real world. Driving, situation. Uh it's pretty fun here's an example, of the first person perspective. So highly simplified, environment. Um. But somehow. Compelling. In that just a few simple cues. Create. The. Essence, i think of the driving, experience. The idea. Of an ai enabled spatial attention assessment, and training system, would be that you are able to. Train, attention, repeatedly. Using the driving task and to assess. Attention, using attention, atlas for example, before and after the driving test. And we can use artificial, intelligence. In, fancy, ways. Uh to, predict. How the person is paying attention, in real time during driving, and to adjust the driving scenario. To enhance. Episodes, of attention, that are helpful, and to minimize, those that are, unhelpful, for the person. The challenge, of the competition. Was to. Propose, a business model. And do some market, research. Uh the relevant, people who may be interested, in this maybe, clinicians, and brain injury survivors. In queensland, there are 77. 000 stroke survivors. 26, 000 with disability. About one third of neglect, and 15 stroke units in public hospitals. There are different ways, to expand, the business. One would be to focus on big data acquisition. That would be something like sending. Ai. Vr systems, to different rehab, units, and using ai, to analyze, the, emerging, data patterns. Another thing might be a local strategy, and this was conducted, actually. At gold coast university, hospital. Uh, to focus on a small number of patients, for whom driving may be relevant, or, they may have extreme, attention problems. And to work with them. To try to bring them back up. To everyday, life. The vision of this type of device, would be a cost-effective. Recovery, solution accessible, to all. And we propose a video game model to allow free access to clinicians, and inpatients. Sell to end users. At scale and return to home. So that's the ai, enabled, attention. Assessment. And training. System. Is one example, of what we can do with vr, but vr is not the only technology, we also have. For example, brain. Computer, interfaces. And to tell us a little bit, more about that, i'm going to head over to mr, mika, ching. Business. Manager, of gtech, neurotechnology. Hong kong limited. And he is going to present on the topic, of recoveries. I believe. Thank you david. So, i'm going to talk about recovery. A whole new. Technology, they use computer, interface. To, help, help patients, with stroke recovery. So right now, we are working on upper dim but we'll talk more about how we're going to do it in the future, so i'm micah, from, hong kong. Uh maybe i introduced, a bit. For a company. We are based in austria. Headquartered, there. Then we got some branches.
Of The road us, office, and hong kong office so, basically, covering, the whole globe. Or, we are not only working. On, stroke, rehab, we also work on different kind of technology. That combines. Bci. Into their life so for example. Or, coma, rehab. Or, even using, hospital, system, for stroke. And. Open brain surgery. So all these kind of projects are working, mostly, in europe because we are based there. So some example, how we use, brain computer, interface, for this example. Is an object. Or exhibition. We combine. A eeg, cap, measuring the signal from the subject brain. To control. A robotic. Or arm, to do some painting, so for artistic, expression. Another example, will be. What we call, cyberthon. Is an international. Competition. Where all participants. Use different kind of. Bci, machine. To move around, and do different kind of sports. Based on their brain function. So you can see bci, can be applied, in many different fields, as long as you can pick up the signal, and transform, the signal into some kind of output, through your machine. In our company, we do, are different kinds of combination. Uh you can see different, sorts of. Amplifier. For eeg. Or body sensor measurement. Then we also combine, with vr glasses. Eye tracker, to do, different kinds of function. Not only for research, but also, in. Medical. Use. So for those who, are not familiar, with bci. The full name is brain computer, interface. In simplicity. We require, patient, subject. We measure the eg. Or yi cod so some kind of bring signal from the subject. We feed the signal, through, a computer. To pick up the control, signal, it could be, a hand motor function. When the subject contract, their arm muscle. There will be a spike from the motor cortex. We pick up those kind of signal, compared with a baseline. So, after transforming, those kind of special, signal, through the computer. We usually, send a feedback. Back to the subject, itself. For recovering. We are focusing. On upper limb rehabilitation. So the feedback, will be electric. Stimulation. On the forearm. Of the subject. So. It helps them to contract, their arm. And. When there are some kind of stimulation. The brain signal will change accordingly. So you can see it forms a closed loop system, keep updating, the pickup signal. And the feedback, itself to the. Subject.
So How we use this kind of uh bci, technology. On recovery. It's all based, on the audi, rugby core, so when you do a, left hand imagination. There will be a right side brain. Synchronization. And same for the right hand movement, so you see a. Signal, activation. From the left brain and vice versa. So what we monitor. Here will be the brain motor, cortex. Or basically. C3, and c4, region, of your brain. In traditional, approach, or many rehab centers, use. Multi-imagination. For their stroke rehabilitation. So it helps them to stimulate, the brain barrier, cortex. And. What we want to do is to combine, traditional. Method, into a bci. Methodology. So what we do is adding more, imagination. With functional, electric, stimulation. And visual feedback, so somehow, be our system, here but it's not those are headsets, that uh david use, or we use just a monitor, with two virtual, hand, so it forms some kind of video therapy, for the subject itself. The principle, behind, would be. We ask the subject, to, do some kind of, left, and right hand movement, imagination. And, the subject wears a eg, cap, you measure, eg signal, and pick up those special, kind of motor, stimulation. So the special. System detects, the tracker. And, if the signal. Reaches. A, certain threshold. It will trigger, some kind of electric, stimulation, on, their left and right hand. And at the same time the vr. Monitor, will display. A, virtual, hand. To form some kind of mirror therapy. So the movement of the hand, plus the stimulation, on the arm will trigger, more, uh sensory. Neuron. In your cortex. Which means more activation, your brain, and more activation. In their neuron, will cause this um, more, brain plasticity. Reformation. Which helps the subject, to. Bring back their motor function. If we build it visually. We monitor, the brain motor, cortex. So, the upper limb will be around this region, lower limb or in the central. We monitor, c3c4. Region. So the subject, wears a eeg, cap, mostly, on the motor cortex, we measure eg signal, and feed the signal in real-time, to a computer. The computer. Already. Has some kind of classifier. Based on the subject. So if the signal. Of imagination. Of the left or right hand mutant, matches, the, classifier. You give them two kind of feedback, the visual, virtual, hand. And functional, electric stimulation, on their forefront. So this kind of stimulation. Support, the plasticity. Of the brain. And, it generates, different kind of eeg, signal, again in real time, so it continues. For, about an hour. Uh which helps them to build up the brain motor memory. The real system will look, something, like this so. There's a monitor, with two virtual hand here, vl. Setup. Then the subject where is a eg, cap which is wireless, so the signal will be transferred, wirelessly. To the control, computer. You notice, on the forefront, of the subject, there'll be two, uh fast path, for theatric, stimulation.
So This is the controller, for electric stimulation. And everything, will be run, or, actually the setup is pretty small you just need a small table. And. This, whole, control, be controlled, usually, by. Physical, therapists. Uh, this alcoholic, in austria. And. In some countries, they require, doctors, on their so depending, on the region, itself. In simplicity. The subject. Sit here, with left and right hand connect, to the electric, stimulator. The patient monitor, in front of him. Then we try to separate the patient, from the, controller, itself, so, the physical, therapist, will control the operation, pc. Which the subject cannot, see the data itself because we want the subject to, perform. Or just focus, on the patient monitor itself. And, throughout the whole one hour section, he can be really, dedicated, to his own therapy. Um. Let me see, so there's a video. In, from our clinic in austria. It shows how we, apply, the. Electric stimulation, path. So usually, we clean the hand first, or especially. For hands which is dry, or. Bad for conductivity. We make sure. The patient is comfortable. For the therapy. Then we help him to wear the cap. For those with glasses, we ask them to remove, it, first. Just to peel on the cap, so notice the cap, the electro itself, is focusing. On the motor cortex so the metal, branding, of the brain. There are total, 16, electrodes. Which is sufficient, for us to pick up the motor signal. And, the therapist, just measured, the. 10 20 system. Uh the positioning. What we call montage. Of eg. Making sure the position. Is correct for the subject. Then we need to insert. Some. Conductive, gel. All the electrode, are active electrode. So there are no. Pre-processing. On the hair, just insert conductive, gel. From the electrode, itself, and to the scalp of the patient. You can see the. Preparation. Is pretty, fast. Or for 16 channels, around one to two minutes. Then you're done. In the past when we use or some older technology. Passive electrode. You always need to scrub the skin, and do some skin prepping. Before, applying the gel and it takes, 40, to one hour. So it it's quite time consuming, but the technology, advance.
That's Why we can use it more quickly. So this is the reference, electrode. Making sure we get the baseline. From the, subject itself. In order to have a good eeg, signal quality. So after setting up the ddg, site. Uh the therapist, helped the patient, to. Place, two, best path. So, he asked the subject to do hand dose reflection. Make sure the muscle. Location, is correct. It tries to. Pinpoint, the location. Sometimes, we add um. Gel to the pad itself. In case the subject hand is too dry. To make good connection, between, the skin, and. Fast controller. So after placing the pad. We can actually. Tune the. Electric stimulation. Based on the subject performance. So right now the therapist, is increasing, the current. To see if the electric, stimulation, is strong enough for handles, reflection. Of course we do not want to, over, excite, the subject itself. So making sure the subject, is comfortable. But at the same time having some kind of, hand construction. So that their neural, cortex, can be at the face. The whole paradigm, works like this so. Uh the whole, um, section will be, around. I mean one session will be around eight seconds. After two seconds there'll be a beep sound from a computer. As, kind of like attention, seeking, for the subject. In computer, ask the subject to perform, left or right hand, motor. Dot hand dorsiflexion. We try to ask the subject to perform. Or for around 8 seconds of imagination. If this kind of eeg, signal, matches, our classifier. We give them the visual feedback. And the muscle stimulation. From the fastballs. The whole therapy, runs, around, an hour. Or, which divides into three session, first session, is for calibration. For some of you know eeg, signal. Is. Um, depending, on the subject, itself, so that's why every time the subject comes in, we need to calibrate, the, eeg, signal for his day, so after calibrating. For 80 trials. Half of them will be for left hand imagination. Half of them will be right hand, we get base 9 of that subject, then we can perform. Second and fourth trial which is the real training, session. Again, our actually, trials, for, left hand aka trials for the right hand, so the whole therapy. For one session will be 240. Trials. Or. Including. A setup, time, that i showed you before. The whole session will be around, an hour. So, in, official. Training. Each patient, will performs, around 25, sessions. So around. Two to three. Months. Around, one to three times per week. So it's pretty, intense. I would say, but it really helps them to recovery. The plasticity. Of your hand. For interface. Uh when we enter software. Uh there will be page for, patient. Setup, so you can see there are different patients, you can select here. Just let the patient install. Then this is the drd. Map, that i showed you before, it shows the brain performance. Of the subject. So for example. For right hand imagination. That you can see some kind of activation. On the left, voltage, protest. But for left-hand imagination. You can see the activation. Is actually spread out through the brain, the green color part. Is because, the stroke patient is usually. Still calibrating, the brain so the activation, of the brain is not really. Precise. Usually, all the damaged, part of the brain will be. Helped, by other prophet brains so you can see the spread out function.
What We need to do is set up the current, of the fast. On their forefront, so the current, is. Around, 10 to 20. Uh, microampere. Or for those strong, athletes. Um. Usually you need more, um, electric. Current. But you can always tune it based on the function, or based on performance, of the subject. After that you can enter. The calibration. Mode which. Determines. The baseline, of the subject, so you can see there are 16, channels of electro. You can see data in real time. And the brain, of this, are there any. Interference. Of that signal. And here will be the left, or right hand imagination. Target. It's randomized. That's why all the subject cannot anticipate. Mixed, imagination. Itself, it helps them to really focus, on the task. After the whole session we'll see the eld, map again. Which the therapist. Can. Determine, whether, the performance. Of the subject, matches. The expectation. Of, this. Usually, the case, where the red color, is the activation, of the brain. You can see, for, right hand imagination. We, are anticipating. Left brain activation. More so more red color here. So for, right imagination. Or the subject is performing. Far right, but for left hand imagination. You can still see. Left side of the brain got more activation, than the right side. Which is not ideal, and not normal for, a healthy subject, but for stroke patient, as i mentioned before the brain is still calibrating. So after. Around, five to ten sessions you'll see more. Activation. On the right side so the brain is actually, shifting. The activation. In the motor cortex. After each session, we usually give them a training, report. Or in traditional, approach, or, all the performance. Will be assessed. By fuku maya assessment. Rock industrial, different kind of indexing. Based, on the devaluation. Of the therapies, itself. But in recovery. We try to give them a numeric. Uh performance. So what we call accuracy, score, also this report is in german, so that's why, it's in deutsch. Um. And, this, you can see, uh for 25, session the first few sessions, start from around 60, to 80. Or accuracy, score, because the subject, is not really used to this, system, itself, and it takes time for them to get. Used to the whole run-up. Or when to anticipate, the imagination. And giving them a score they will, have the aim to get 100. Accuracy, every time, and they can even compare, this goal with other subjects to motivate, themselves to get higher. So the trend is usually, up and down, some fluctuation. Based on an eeg, performance. I mean the performance, of the subject, your cdg, is based on the. Psychological. Background. And the motor, functions, so there are many factors, about an eeg. Sometimes, when the subject, comes in, and they said they didn't sleep well, yesterday, night, and we are anticipating. Lower concentration. For the therapy, on that day, so this kind of psychological. Event, will, cause, some drop, in. The performance. One or two days. Then. It usually, comes up again, because they know what to anticipate. Next time. So right now, as we are based in europe. Uh most of the recovery, is cleaning. Or what we call recovery, exchange, because we are using the brain power of the subject, to do all those workouts. The recovery, extreme, is mostly, located, in europe. We got better support there. Then, starting, more, in united, states as we are getting the fda license. There are field unit in hong kong for research, purpose. Also we got research partner in japan, to. Improve, the outcome from behind, to calculate, the brain signal. And. Working, on the lower limb also. So if you want to see more, like uh. Patient. Sharing. Or, for the technology, behind, just feel free to access the webpage. And, yeah thank you for listening, i will hand back the time to. David. My. Excuse, me david, sorry, you just have to unmute. My apologies. Thanks, for the presentation. Rico. Uh. Now we're going to hand over, to. Dr camilla shirota. Advanced, queensland. Fellow and research, lead at, habitat. She's going to present, on the topic of. Exoskeletons. All right, um can you guys hear me. Thanks, great um, hi uh everyone, good afternoon, and uh thanks for joining us today as david said my name is camilla. I'm an engineer, by training, um and i have a background, working with prosthetics.
And Exoskeleton. So, i'm very. Happy to be able to talk about exoskeletons. With you today. Um, so before, we. Delve into the the, topic, we're going to start by defining, what is an exoskeleton. Um so exoskeleton. Is a wearable, device, that um, we're supposed to use on the outside, and on top of the body and it attaches, to our body. Um it's, meant to work in parallel, to the person that's using it, and the um goal of the exoskeleton. Is to augment, support. Or restore. Movement. Um, even though we've only, recently, started seeing, lots of exoskeletons. The idea is not new. Um the first patent, for an exoskeleton. You can see up on the top right corner, is from 1890. So, we've been wanting, um to make devices, like this that can support, movement. And one of the reasons, why, you'll see that it's a growing, field. Um so, um, just a caveat, this isn't for just exoskeletons. It's about, robotics. And um. Vr, or, other technologies, being applied to rehabilitation. But kind of the. Um, message, is the same. That a lot of this research, or a lot of this work has been pushed by developments. In technology. And we're now able to do a lot of things that we weren't able to do before. Um what does that look like. In exoskeleton. Space, so this is, i'm going to focus mostly on commercial, devices, so things that you can go out and buy. Um. There's a range of applications. Probably the very earliest, ones were military, devices, so these were made to help, people, carry. Big loads, for long amounts of time. We've seen recently. This, um. Coming out uh consumer, grade devices, so an example, is that um the ski one right there in the middle um, robotics, there are quite a few. Um that are made to support, people that want to to go skiing. And try to ski for longer. One. Area that's been growing, a lot in the past few years, are industrial. Devices, so you can see for example the one on the um the very first one there on the left is from lockheed, martin and the idea of this device is that it's taking.
The Load that's in the arms or high, of over the head and it's transmitting, that all the way to the ground. So this is to help decrease, fatigue. And another one another, big application. In industry. Is the one that you see in the middle octivo, so a lot of devices, are looking at, i'm trying to prevent. Injuries, in workers, so that they can work for. Longer amounts of time and also prolong, the number of years that people can continue, working without having to, um. You know uh, get injured or have to go um. Rely on um health um. Sorry, injury related. Issues. Um. Medical, is one of the biggest ones traditionally. Um, we're going to focus on those today, you can see some examples, there on the on the far right so i'm not going to talk too much about it because we're going to, i'm gonna talk about that later, um and the last, kind of big application. Area is in research, i'm really not gonna, touch too much on that one today, but, um a lot of what we do in research, with these robots, is trying to, um, use them as ways to, learn, more about, how we move our bodies. Um be it motor control. How we learn how to move, um but also in neuroscience. And then of course education. Is another great. Application. I'm in medical, there are also a huge, range of devices, and again here we're focusing, on rehab, there are lots of other types of robots surgical, robots, um, all sorts of stuff. But you'll see that they have lots of different, shapes. They attach, to different, body parts, and all of this is related, to, the end intended, use of the robot, is, so if you have the device. In the middle, the locomat, i think almost anyone in rehab knows that device, it's the, was the first um, rehab robotic, device. And it's big and it's heavy, and it's fixed. And the reason, why is because it's meant to help people that have very poor walking, function. Um to be able to get some therapy to start walking. So to be able to deliver that type of function, it, kind of was designed in that way, that's very different for example. To the honda device that's right next to it you can see it's very small, it only attaches, to the hip and the only thing it's trying to do is support, hip movement.
Um So that people can more freely, move outside. Um, similarly. To that the idea of, being able to give people the ability to move outside, is the re-walk, on the far left. Um, you can see there that. The person, can walk outside but the trade-off, there is that the. Person has to share, balance control, with the device, so that's why they have to use crutches. And that was, um. Something, that, for example, the wrecks, on the, far right, um, they kind of decided, to do it in a different way, um it doesn't, it's big, it's clunky, it's not meant to be used, to. To um. Kind of walk in in your environment, but rather. Kind of restricted. Mobility, inside of the house but the advantage, is it balances, by itself. Um, and. At the same time it helps, free the hands, of the pa of the person that's using it, um so there are lots of trade-offs. And the fun thing is um, you know there are lots of different exoskeletons. To fit lots of different purposes. And then finally, up on the top, corner you'll see there, of course, upper limb and hand devices, i won't talk about them too much. Um, but it's not for, any reason, in particular. Other than um. All the topics i'm going to talk, about today, kind of apply equally for both. So. Um, within medical devices, we kind of exoskeleton, devices we divide them in two big groups, one is focused, on clinical, use and we're talking about therapeutic. Devices. And the other one is for personal, use where the devices, are really focused on just providing, assistance. And one of the big reasons we divide it like that is related, to safety. So clinical, devices, for example. You can. You can predict what type of environment, you can be in or you can expect, a very controlled, environment, where the. Robot is going to operate. You always have therapists, that have to go through formal, training so they know how to, use the device they know how to adjust the device, and they know what to do in case something happens. And in case something, does happen you have access to a third person there to be able to help.
The User. And, um, ultimately. It's all about. Remembering, that it only takes, one. Bad event, for, the user, to to, lose confidence. On the device, the clinician, as well and then also, the companies. They become liable, for. All sorts of injuries, that could happen. So to address that of course we have medical device. Certification. These are examples, i just literally just copy pasted them from the websites, of the companies, that um commercialize, these devices. And you'll see that they're very very specific, about, who. Um the intended, user is. What, the um. How they're supposed to use it or who's supposed to be there with them and it's also very explicit, about what it's not intended, for so if you look at the, um top paragraph, the last sentence says, the endigo, is not intended for support, for stair climbing. Medical device certification. Is also, very complex, so depending, on what part of the world you're in you have different regulations, you have different regulatory. Bodies. So. For example. The other two snippets, are from the exo, there's different, uses, um. For the us, and for the eu, and that also helps explain, sometimes, why. Some people have access to some devices, and some parts of the world don't have. Because you have to be able to get this regulation. Wherever you want to operate. Related, to that explosion, that we saw in the. Other slide, one of the big. Technological. Challenges, for exoskeletons. Is being able to. Um. Move the devices. And that's a trade-off, that we have in engineering, so usually, the more. Power, or the stronger. Your. Motor or actuator, is, the heavier, it is and this is really really critical, for hand devices, and that's why um i put those pictures, there, so you can see the hand of pope it's very big it's very clunky, but it's very strong. Versus something, like the, hand exo, from the relab. It's. Much smaller. Um the interface, is much nicer. But at the same time the amount of force that you can get out of there um, is very, but it's much smaller. The graph on the side kind of shows you that um. We. Or. Don't have a huge. Range, of devices, that we can rely on um but there are lots of research, um in that direction, of trying to develop. A more lighter, and more powerful. Devices. That, also, don't consume, too much energy. Um, one, of the biggest, challenges, in exoskeletons. Is sharing the control, between, the robot, and the user so if you imagine, here for example, you have a spectrum, of different.
Users, From somebody, that needs a lot of support. On the left all the way to an independent, walker, that just needs a little bit of support, at specific, points of the gait cycle. How to, manage, this different. Being able to tailor, to each person. How much support they need and, also being able to. Have the robot. Provide, the support, and figure out by the by itself, is a challenge. Um. And there are lots of people doing research in that as well so for example, here again we go back to the locomat. On the top there, you see, um. Two. Virtual legs moving, the blue trajectory, is the desired, trajectory. And the orange, is the one that the patient is doing together with the robot. And it's running what's called an assisted, needed algorithm. So if we look at. The amount of support, that the robot, is providing, over time, so, we can, make these graphs so you see the robotic, assistance. At different, points of the gait cycle. And. As the number of steps, change, and what you can see is that in the very beginning, you have a lot of support that's provided, in the entire gait cycle and as time goes by. You can decrease, the amount of support that's provided, but that is still depending, on what point of the gait cycle. The patient, is in. And lastly. The locomat, in particular, can also provide different amounts of body weight support so you can also see how that changes, over time. And just for reference. If you compare it to a non-ambulatory. Spinal cord injury patient you can see the assistance. Also goes down, but the shape is different depending on where you are in the gait cycle, and also the amount of weight support that needed to be provided, was relatively, high. That's, only one example, of of um, how the robot, and the user have to interact, it's not only for, target movements, but it's also for balance control. Um for tasks, so deciding. Does the do you want to get up from the chair do you want to kick a ball or do you just want to walk around. How to negotiate, different terrain, so we're talking about overground, walking, stairs, ramps or walking over gravel or grass. How to negotiate. Obstacles. When the um. The exoskeleton. Can't see for example. Um and also how to react to perturbation. So the challenge, here is balancing. Between. Um what the user does and what the exoskeleton.
Does, And the more, we look into, how this interaction, needs to work, the more complicated. The system becomes, because we need more sensors, so the exoskeleton. Can perceive the environment. Um the interfaces, with the user have to be complex, for example, you could use a bci, like, mika was talking about before. You have to have different actuators, and, a lot of people are also trying to embed some artificial, intelligence. So that the exoskeleton. Can. Make decisions, on its own. And understanding. Also, what the exoskeleton. Needs to do if you want to be able to provide, natural, movements, is something that we can do with exos, um other exoskeletons. So for example, this one is a research, device. That attaches to the leg and it applies perturbations, at different points of the gait cycle. And the idea of this device, is that um so you can see on the left, is the, the person walking, we have on the right the knee angle on the top and the amount of torque that the device is applying on the system, on the um. And. What it does is it blocks the knee at specific, points of the gait cycle. So you can see there, um, how the pattern, changes, and based on this pattern change, we can, understand. The underlying. Mechanics. Of the leg during a normal gait cycle, and then what you would want is to copy that in these devices. Um, finally, the clinical, implications. So unfortunately. Exoskeletons. Have not, fulfilled. Everything, that we've ever dreamed, in um, movement. Um and actually quite the contrary, a lot of what we've seen are improvements, in secondary. Things like bone density, and bowel function, and spasticity. Which are related, to. You know posture, and being upright, and. Promoting, some movement. And one of the, difficulties. Is really. Kind of. Shown, in this, tech, piece of text that i expected, from one of the most recent reviews, on, robotic, therapy, for stroke. That says that it might improve their activities, of daily living arm function, and arm muscle, strength but the issue is we have to take that. With a grain of salt because, there's huge variations. Between the trial, and intensity, duration. Amount of training, the type of treatment, the participant, characteristics. And the measurements, used, so just. Talking, to. Um, all of those parameters. That, i've. Kind of, um. Very. Superficially. Hit on, is that it's really hard for us to make hard conclusions. Just because, there are so many parameters, that we can change and that how, that really affects, the outcome, that you have. So hopefully, the take-home message, um for me is that there's, it this is a huge area, um it's really growing really fast there's lots of potentials. We have really high expectations. For these devices, and they're very complicated. But hopefully. What i would like people to take is that they're only tools, that enable different types of interventions. And what we should be doing is not judging, exoskeletons. As a whole. But rather looking at them as, ways that we can achieve. Um different, goals, and i leave you with a video of the cyber salon that mika was talking about earlier. This is the leg exoskeleton. Racing so you can see how people are managing, different, different obstacles. Thank. You. Thank you camilla, that was. Fantastic. Uh this leads to the last section, of our, time together today. Question, and answer. Uh so we have some time for that if you haven't put a, question into the q a function on the right hand side, we now welcome you to do that, uh, there's a slide up now.
And All of the presenters. Uh, will be ready, and unmuted. Um, to answer some, some questions. Now. I'll start, uh we have a question. Uh, from, robin, which is about vr, how. Soon, post uh, tbi. Can you use a vr headset so a person has a brain injury. How long do you need to wait after the injury for the person to be able to use the dr. Actually i think that depends, on the individual. Person. They need to have some. Awareness. About them. Tell you it's a case example of a gentleman, that we. Tested. Uh we had him, about. Two months. Uh, post. Uh, tbi, that was actually the data that i presented, in the in the talk. Uh, and he, had a lot of fun. And he, he said he wished. That he could have done it earlier, on, and the idea of earlier, on might be that there could be a critical, period, in which the brain is mass. Maximally, plastic, to recover. Faster. Um. But the nurses, there, uh had a slightly different opinion. And they suggested, that he would he wouldn't have been able to perform, the task, he didn't know where he was. Much, so it'd be hard for him to get. To do that kind of search task, but it's worth pointing out that in those situations, of the person, not, being completely, in the. Faculty. That it may be possible, to use more passive. Scenarios. And to measure attention, uh. Passively, in that case. We're going to take you to the beach. And while you're doing that we're going to be measuring your brain activity. Via. The. Virtual reality. Attention, mechanism. We have a question. He hope that answers your question. Now we have a question, from francis, this one, is from mikko. Uh. Bci. Use is that impacted, by people's ability, to visualize. I guess what you mean visualize. Their ability. To imagine, their hand movement. But it. It does, all, help if the people. Has a more, higher cognitive, function. So some of our inclusion. Exclusion, criteria. Is related to cognitive, function, assessment. So it's the best result, if the subject, has a higher cognitive, ability. To visualize. And concentrate, on the tasks, of course some people have difficulty. On that, and in such cases, we use different kind of motivation. For example. Uh putting a bottle of water in front of the subject. And ask them to try to grab, the bottle, in front of them. Which might help them to visualize. Such imagination.
Movement, On the brain. And another example, would be uh, ask them to imagine, not only hand dorsiflexion. But, they can try to, imagine. Uh, like handling, a basketball. Like patting the ball itself. Such kind of, upper limb, imagination. Will also trigger. Motor, function. Or, signal. So it also can be feedback, to the pci, system itself. Uh here's another question from mikko this is actually my. Metamorphosis. Of a couple of anonymous, questions. Uh. So. Can you you, can use bci, on the lower limbs. Uh. And. Is recoverics, being used simultaneously. On both the affected, limb and the unaffected. Limb. Okay, both of this question, are the most popular, ones during my talk so. We. We do use a bilateral. Method, for recovery, eggs, because. We did some research, and there are other paper, published, outside. Saying bilateral, training, for. Lymph. Improvement. Is actually. Has a battery effect, than a lateral. Method. It's because. Of the brain activation. Uh, need to be calibrated. Not only from a single, side, as i mentioned before, they are all eg, signal, from also, the healthy, side, so by doing bilateral. Training. You are helping the brain to stabilize. And reactivate. The whole brain region, instead, of a single-sided. Region. As for lower limbs we are already, doing, our research, on it, and developing. An algorithm. For separating, the dg signal from the lower limb, the challenge here is lower lymph, motor cortex, at the center, of our head, which is really small, area, for eeg, signal pickup. So we need really precise. Uh, algorithm. In order to, pick up the signal. For. A better. Bci, methodology. Okay here's a couple of questions. For camilla. First one from harvey. Does the excessive, use of exoskeletons. Lead to muscle, skeletal, weakness. That's a good, question, um i have not heard of that um this far i think a lot of the use in exoskeletons. So far has been, um for supporting, people with spinal cord injury to walk, and. That's actually, been the contrary. So, from no activity, at all to some activity. We see positive. Effects. Especially, related to spasticity. Uh here's a question, from uh, robin. Uh more of an engineering. Uh, field. Some of the systems. Have, uh split. Instrumented. Treadmills. Which provides, uh, better gate data. Could you elaborate, on that please. Um. Sure, i mean i guess uh the. One of the biggest benefits, of split belt that we. Know in rehab, is for stroke. So you can change the velocity, of the belt for each of the legs. Um and for sure i think there's lots of. Parameters, that we still haven't explored, and that's kind of the front of exoskeletons. Is that they've also been. Opening. More questions, of basic science, that uh we get to explore, as well. Here's another one for you while you've got the mic on. We've got two minutes left yeah, uh i think we can do this. Maybe not if you keep asking questions we could be here all day but, uh here's one from deleno. Uh, exoskeletons. Maybe, we should market them as exercise, devices. Rather than asset assistive, devices. I think that uh, clients, use them as uh, do you think they might be assisted, devices. Might be disappointed. Um, i think there's, uh, both. Um. Definitely. The therapeutic. Devices. Um. Are. Well you could see them as exercise, devices, as well. Assistive, is more. In the phase, where you're not doing therapy, anymore, and really, to. Try to. Replace, functions, that are completely, lost. Uh there's a couple of questions about. Vr. And i think we might. Just answer these and we'll. Leave it there and. Thank everybody, for their time. And to point out that if you do have further questions. Uh for any of our speakers today that you can contact. Contact, us directly, and we can provide you. With more information, should you require. Uh, so there was one. About. Uh, driving. Driving, from kathy. Uh driving program for training and attention and neglect, have you used this with patients. And you currently have medical clearance to drive. Uh. Do you have any issues, uh with this this is something that we're actually, in, uh. Conversation. Now with clinicians. Exactly, how to, go about this uh so far we've been using this as a sort of gaming.
Approach. Uh where we use it as a gaming scenario. To enhance the person's, ability, but you could also imagine that you would use it as a. Um. As a. Rehabilitation. Technology. If we pitch it as a gaming, and. Recreational. We can bypass, some of those. Uh. Official, approval. Uh regulatory. Pathways. Uh if we, propose it as a clinical trial then it requires, uh additional. A red tape, so those are some of the strategies, that we're, discussing, with clinicians, at the moment. Um yeah that's a good point. Um. This one's from anonymous. Uh. Hi just wondering, are there any, vr resources, videos, on youtube. That are available. Use. Off-the-shelf. Vr. Glasses. I recommend. Looking. Uh, up the unity, game engine. And the steam, vr, plugin. For unity. There's lots of tutorials. On youtube. That allow you to make your own. Vr, scenarios. And. They can be quite simple and you just follow them as a tutorial, series it's huge, support in the independent, developer, scene, which makes some of these vr, tasks, possible. To construct. Uh where you think macaula. Time for one more. Maybe. Make this one the last question. Uh this one, uh is from anonymous. And this one is for mikko. Uh hi just wondering, if there, are any formal assessments, to understand the sensory, murder and cognitive abilities, of, individuals, who may use bci. Okay, so we do have some official. Or scale. For cognitive, function, uh the most common is mokka. Montreal. Cognitive, assessment, moca. Which is widely used for, physical, therapies. And for motor, function, uh we usually use fuku, mayo test, i think it's from japan, but it's the most widely used test for stroke rehab. Thank you very much, uh thank you camilla. Mika. Mikala. Thank you for you, for your attendance. Everybody, i thank you for joining us, for today's hopkins, hour, we really appreciate, everyone making the time to join us today it's been such a pleasure to have you. Um, although very strange, not being able to see any of you uh we hope you found it uh, useful. Uh, and please feel free to reach out to us if you have any uh. Further. Questions. As mentioned, earlier, we will be distributing, the event recording, q a responses, and any further. Materials, in coming days. Again, thanks for your time. And hopefully, we'll see you. In real life. One day. Thank you. Thank. You. You.
2020-10-09 11:38
