Gamified Rehabilitation, Anytime, Anywhere
Empowering invisible and paralyzed limbs
Image you lost a limb, or lost control of a limb due to paralysis. We’ve been thinking about that a lot in our project. Up to this moment, we’ve been focused on the difficulty of expensive bionic limbs, but it turns out that the rehabilitation process is quite costly as well. Our ambassador Bodo Hoenen mentioned more than 5000$ a month on physical and occupational therapy.
The literature backs up such large numbers; in the US alone 185.000 people lose a limb per year, require 6–12 months of rehabilitation, leading to a 5.4 billion dollar industry. This is comparable to 2500-5000$ a month per person. Next to that, Bodo mentioned paralysis from stroke is very common, almost 800.000 people suffer from stroke in the US, creating a 30 billion dollar industry for healthcare and rehabilitation.
So what to do?
Given the problem above, an affordable training method to supplement or replace expensive therapy is very valuable, especially for people whose insurance does not cover such costs. Therefore, we envision a simple solution that supports people in their rehabilitation;
- use available cheap (yet) reliable sensors to measure muscle activity,
- couple that into a user-friendly learning environment where algorithms do the heavy lifting,
- provide the amputees with feedback and fun activities,
- leading to well-trained minds and muscles.
- (oh, and did we note training all day might reduce phantom pain?)
Will it work?
We’ve been working with a myoelectric armband from OYMotion; the gForce Pro. Using my data science expertise, I focused on the data processing and algorithms. I also supervised Paul, who performed a magnificent intership with us, showing over 90% gesture classification accuracy. He was kind enough to place his report here for other data junkies.
However, with our new focus on gamified rehabilitation, I wanted to show it’s is feasible to control games with the armband. With some hacking I could get it to work and I am quite happy with the result:
Now what?
Several things are needed for further succes:
- Create a userfriendly-ish product or service, MVP-style of course.
- Create (or join) a sustainable institution to support said product.
Ah, can’t be that hard?
I am pursuing several avenues. Preferably everything is open source (serving mankind), yet also capable of motivating talented people (and myself). The most successful institutions these days to motivate talented people are either a non-profit (donation based) or a company (transaction based).
As a dry test for a company I joined a local startup competition. I had to write a good business plan to join the finals. While it’s useful to define a clear vision, writing a business plan turns out to be extremely boring. Evidently this is a hurdle for me, since I failed to join the finals of the competition, where it got more interesting with a pitching session.
So for now we continue with a few volunteers (time and skill donators). And we’re going to create a pitch deck and demonstrator anyway to gather more feedback and inspire like-minded individuals.
Further development
The current armband has some limitations, since it only fits adult forearms. For example, Bodo’s daughter still has a paralyzed shoulder. To help her recover we’d need to place sensors above the shoulder, or find another control method.
OYMotion sells the sensors separately, but that still requires us to build a control board and software. An alternative I am researching is openBCI; a startup building open source brain computer interfaces. They have an impressive GitHub repository and a control board that can be connected to EMG and EEG sensors.
Through OpenBCI I found out about NeuroTechX, a non-profit with a generalized vision for supporting open source human machine interfaces. I am curious to see if they have an active community.
Other gamified rehabilitation attempts
To conclude, let’s keep track of everyone else who’s trying similar things as us. We can either join forces or learn from each other.
Limbitless: A non-profit that already had the same idea for gamified EMG control. Last I knew they were working only with academia, so I do not know if something will be available soon.
ADAPT-MP: This design firm had the exact same idea. Seems there is no follow-up unfortunately.
MindMotion: MindMaze’s solution seems the most advanced out there, targeting stroke rehabilitation using a Kinect. Quite inspiring, though I don’t know the costs. And probably not suitable for amputees or full paralysis like Bodo’s daughter.
Rehability: another a Kinect system for rehabilitation.
Armeo Senso: a product aimed at self-directed arm therapy.
Ben Ryan from Ambionics is working together with Newcastle university on innovative EMG training. You can follow his work on SubscribeStar.
Wow, found this EMG controller project after publishing, including a nice video. It’s a large student team called Neuro Technology Exploration.
Another development is Ctrl-Labs developing another EMG armband, like the Myo and the OYMotion GForce.
Windtales is developing a similar approach as I suggest, but for breathing excercises. Very inspiring!
If you are working in this field and would like to reach out, feel free to contact me!
Disclaimer: we don’t know what we’re doing, we are not experts in this field and have no medical training! All information, thought, and code described here is intended for informational and educational purposes only. Use at your own risk, and do not use the information or code to make medical decisions.
