A bold title, possibly sensational, however, a growing case for providing real-world results. The short answer today is simply, Yes!
Enter Project Fizzyo
Back in 2017, I joined an adventurous project being run by Great Ormond Street, The Cystic Fibrosis Trust, University College London - Great Ormond Street Institute of Child Health (UCL GOSICH) and Microsoft, creating the largest collaboration of Technical, Clinical, and Research staff together in a new research project looking into physical treatment for Cystic Fibrosis.
The aim of the project was to simply ask the following question:
What physiotherapy treatments do children and young people with CF actually do and is the amount you do related to health outcomes?
It might sound odd, but it’s not actually known if the prescribed treatment is wholly beneficial or if regular exercise and medication are a better path. It “Should” based on current evidence but it’s not been scientifically proved.
What exactly is Cystic Fibrosis?
I’ve mentioned it a few times, so it’s poignant to mention what Cystic Fibrosis (CF) is.
In short, Cystic Fibrosis causes a build-up of thick sticky mucus in the lungs, digestive system and other organs, causing a wide range of challenging symptoms affecting the entire body.
The Fizzyo Project’s goals
The Fizzyo project was created to focus on the breathing exercise treatment a patient undergoes to clear the mucus from a patients lungs, where a patient has to blow long and hard into an airway clearance device, in repeated sessions for upwards of 20 to 30 minutes at a time, several times a day. As you can imagine, for any individual this would be very boring. For a child, however, this would seem like an eternity.
For the research, the project wants to capture data and feedback to validate the effectiveness of the treatment, finally proving or disproving the effects, or better yet, identify ways to improve the exercises and tailor them to each patient for the best effectiveness.
As a secondary effect, we wanted to make these time consuming and sometimes painful exercises more fun and encourage kids to do them through gaming.
The Fizzyo Framework
The framework takes input from a custom input module attached to an airway clearance device that vibrates the air as a patient blows through it (attempting to loosen any mucus in their lungs). This module has two inputs:
- A pressure sensor to capture how strong/long the patient is blowing
- A single button to drive interaction
Paired with the sensor is a FitBit, to also track the patient's pulse and other health data but this isn’t available to the games at this time.
For the games, this is where the real challenge began. How do you build a game with only 2 inputs and some very strict usage guidelines, those being:
- Pressure input with only positive values
- Button input with only click actions (no hold )
- You want the player to blow hard, but not too hard
- You want the player to blow in cycles
- You need to give the player pause between sets (to cough up mucus, nice)
- No cheating, can’t let the player cheat the game with fake breaths
The sole aim of the games is to both make the activity of blowing into the airway clearance devices fun and ease the collection of all the necessary data.
The Analytical backend
Behind the games and all the data flooding into the Azure cloud, is an array of services that receive all the data and allow a set of trusty clinical scientists crunch the data, married up with the treatment plans and information coming from the patient.
One thing to point out is that EVERYTHING is anonymous, no patient details exist within the entire framework, players/patients are simply assigned a kid-friendly name (one theme is animals).
The story so far
It has certainly been an interesting ride to date, from the initial hardware investigations to create a brand new sensor to attach to the airway clearance device (designed to meet medical standards no less) resulting in the sensor you see below:
Like the airway clearance devices themselves, the sensor and connecting parts had to be easy to clean (all CF hardware has to be sterilized after each use to prevent bacterial contamination as CF patients are very susceptible to infection) and rugged. They also had to be as “cheap as possible” to keep down costs should they need to be replaced.
The development framework is built in the Unity Game Engine, gives the games that are being created a common platform to build upon, each stage, hack-a-thon, research stage, and trial opened up new challenges for the framework to handle.
The Framework and especially the calibration of the framework took most of the focus for development, trying to balance the settings for games based each patients breath volume, capacity and ability proved the most challenging of all. We eventually settled on an easy medium for the time being. It’s certainly one area that will still need further attention in the future as the research evolves.
With the way the framework is architected however, the impact on the games is almost nil, ensuring we can simply update the framework without affecting the runtime of each game.
In the backend, we also provide a whole set of Achievements, High score services along with the critical analytics services for capturing the breath data captured from the gameplay, all to give better feedback to the players as they progress in the games. We all love achievements and unlocking “stuff”.
At the time of writing, we now have 3 games published with 2 being actively played with tons of feedback and a handful of more awaiting the final polishing.
Our first game, which was used as our primary research app to gather data on how the framework should work and look for ways to improve the overall functioning of the research, was a simple game called Qubi.
Qubi is a simple endless runner with coins on platforms (which have more coins the higher you go) the player has to collect by jump up as they speed through each procedurally created level.
For the Fizzyo exercise, we did a load of tweaking to ensure we could get a good breathing exercise:
- As the player blows, the outer circle of the player cube fills up, if they manage to blow for long enough, it qualifies as a “Good Breath”.
- For each Good breath in the breathing cycle, the inner circle fills up, based on the number of breaths required for this session.
- When the game detects the player has stopped blowing (to cough or stop for any reason) the game paused until the player is ready to begin again where they left off, or quit.
- The Button on the Fizzyo sensor causes the player to jump.
- As you finish a “Set”, (normally between 3–10 breaths per set), you get an “end of level screen”, where the player can pause to cough, and also get their current score.
- Once they have finished enough for the session, they get a “Game Over” screen showing their final results.
Players loved the game, enjoying gathering coins to boost their score. There are also a bunch of achievements they can collect while they play to add to the gameplay. One player reportedly stacked up over 3000 coins (which was amazing as I only got several hundred in testing, and I wasn’t even using the breathing device)
My goodness, when players want to beat a game they really find ways that you never imagined or thought possible, these included:
- Not blowing and just jumping (which is also why we added the stopped breathing pause) to grab endless coins.
- Just blowing slightly to beat the stop breathing pause and jump again to grab coins. (which we put in a method to reduce the number of jumps you can do for every bad breath. Eventually it hits 0)
- Blowing slow in Calibration and then hard in the game to cheat the good breaths. (Sorry kids, analytics don’t lie :D)
All in all the kids are loving playing and grabbing those coins
Archipelayo was created by a talented set of students from Abertay University in Dundee Scotland. The crew, whilst being new to game development have truly excelled themselves, creating not just one but a trio of games and a little hub world to bind them together.
I am really impressed with how professional this project has been handled their project from beginning to end. Straight from the get-go, you get AAA Studio level production on starting the game and the fit and polish is better than most. The team should be really proud of what they have accomplished so far, especially with the many presentation sessions they have had to deliver as part of their university work with some very high up people involved in the project.
Game 1 — Blowdart challenge
Simple little game, whereby the patient has to fill up a pressure bar to then use the button to launch the dart and pop balloons, with the level advancing once the balloons are popped.
Game 2 — Coconut Shy
An all-time favourite fairground game, powering up the players throw to knock down the coconuts from the stands for fantastic prizes. As an added level of difficulty, the game includes a little character mask that tries to knock back those stones.
Game 3— Cannon Fire
Have you ever wanted to blast a Unicorn out of a canon just to see how many times you can make it skim across the ocean, well your prayers have been answered, that and more. Certainly one of the craziest games in this bundle but very kid friendly and fun.
All in all, this is a well round set of mini-games which kids have found immensely fun
Conclusion and the Future
There are many more games coming to the project, some almost ready, some needing a little more work, but I’m doing my best to make them all fit for purpose for the research project, and of course to ensure the kids will find them fun.
Feedback is always welcome and some of the comments we’ve received are:
“Mouse really enjoyed the game! And wanted to continue! He’s already gone over his physio daily treatments! I had to tell him to stop before he passed out :smile:
Obviously didn’t moan once’s about doing physio… hopefully will be still enthusiastic through the 8 months”
“Really enjoying this game (Qubi) now, got to the top and gathered loads of points. It shows by playing the games he is working harder than yesterday when we didn’t, Speak soon
It is all worth it guys!!”
Getting such feedback still brings a tear to my eye, we work hard, we try and make the games as fun as possible to encourage exercise. But to get this from patients and their parents is truly humbling.
The journey doesn’t stop here and everyone who can, should get involved in whatever way they can, whether your a developer, designer, sound engineer or just love games.
Building these projects is already having a tremendous impact in helping patients get through their exercises, but the journey is long and the data we are getting back from patients completing their treatment, helps to improve the overall practice and helps to find new ways to improve the breathing techniques. Are more breaths better? Longer breaths? More sessions? Tailoring breaths to patient capabilities? In short, we need to know more.
The Development site for the Fizzyo Framework can be found here:
It contains a load of information, including the sample Qubi game for reference, to show you how to build games on the framework, including getting started guides.
What kind of games are we looking for?
First off check out the conditions required for each game in the Fizzyo Framework section earlier. Apart from that, we simply need ways to encourage and improve the patient's mental health while doing their exercises.
Some of the things we’ve heard back from patients are things like:
- I want to be able to power/charge up something while exercising that I can then use in a bigger game when I’ve finished.
- I’d like a game that is different in the morning to the night, so it feels like it’s more tuned to me.
- I wonder what it would be like to venture underwater and see things that would be hard for me to experience normally.
- I want to breathe fire like a dragon!
If you are building, I’ll drop one last little requirement. The laptops the patients have hooked up to their sensors aren’t all that powerful (mainly due to the funds needed to support over 100 patients). So watch that CPU use. Basically, treat it like a mobile game but running on Windows.
I’ve been proud and honoured to be a part of this project, but it’s not a one-man band and everyone has been really working hard over the past few years to finally reach the live research milestone. My heartfelt thanks to all the new friends I’ve made on this journey:
- Lee Stott — My go-to guy at Microsoft who has dragged me along (all too willingly) to more events / hackathons and speaking engagements that I care to remember. He truly inspires me with his absolute dedication to his work.
- Professor Eleanor Main — The lead for Project Fizzyo and by far one of the nicest people I’ve met, always wanting to be as involved as she can and guiding everyone.
- Emma Raywood — a PhD Fellow working at Great Ormond Street directly with CF patients who is always quick with a smile and a friendly word. I can truly say her patients are the luckiest around as she really cares for them and always brings us critical information back from them to improve the project. We’ve had many a late night over slack testing / retesting / breaking and fixing soo many things with the framework. Without her, I would have been working blindly.
- Greg Saul — Former Senior Design Technologist at Microsoft Research, Greg is the brains behind the analytics framework for project Fizzyo, including the portals/sites and other associated components needed to collate the masses of data we’re working with. Had so many late nights with Greg, discussing practices, testing ideas and working out the best way (technologically speaking) to make the most out of the data we are getting and the best ways to get it from games safely, constructively and ensuring we are benefiting both the patients and the project.
Who am I?
I’m Simon Jackson, Dual Microsoft MVP (Xbox & Windows), Unity book author and all round Game Development Advocate. By day, I’m a technical architect at DXC Technology in healthcare, working to provide technical solutions to improve healthcare in the UK. By night, I don my Game Development Cape supporting teams like Project Fizzyo as well as Microsofts new Mixed Reality Toolkit (MRTK) and many more. I’m always out there educating and helping developers with their projects in any way I can, using my skills to teach others. I try to learn something new every day, then break it down, take it apart and show others the best way to put it back together again.
I’m supported by my Lovely wife Caroline and my ever so challenging family team, Alexander, Caitlin, Jessica & Nathan.
You can read more about me here: https://www.linkedin.com/in/simonjacksonsdj/
Or find me practically anywhere as SimonDarksideJ.