Director of the AUT Centre for eHealth and former Clinical Psychologist, Dr Duncan Babbage explains the purpose of his face-name recognition app ‘Intro’. He gives insight on what cognitive prosthetics are and what features the holy grail of one would have—as well as why learning from your mistakes isn’t how you should learn.
Highlighting future aims of the industry from personalized medicine—which is already being implemented locally through kiwi cardiologists who have genome sequencing machines on their desk—to clinician efficiency having the potential to be improved through health technology. Possibilities are endless in this field you should have on your radar. Dr Babbage shares his thoughts.
Q: What did you visualize Intro would do for people going through neurorehabilitation?
Dr Duncan Babbage: The Intro app (http://intro.app) came out of work I was doing with people with significant brain injuries, and partly out of my own experiences. Personally, there would be times I would see someone I should have known relatively well and have no idea who they were, or sometimes I would know exactly who they were but still wouldn’t be able to retrieve their name. And so for me, it’s been a real challenge for years.
Alongside this, I was doing work with people with significant brain injuries, which developed into an interest in the face-name recognition difficulties experienced by some people with brain injuries also. Dr Phil Howard, a clinical psychologist who did his doctoral studies with me, focussed his research on how to support people with brain injuries who had these difficulties to be more effective at these kinds of tasks. He and I developed a prototype app as part of that research. While that app was separate from Intro, and took a somewhat different focus, it was the the start of a journey.
Along the way, I heard from lots of people in the general population who said, “I really need an app like that!” Intro is the result. The purpose of Intro is to give people a tool to help them be more effective at learning names. Intro is starting from some of the same problem space as that earlier research, but explicitly focuses firstly on the general population—with the likelihood that it will be applicable with some people with brain injury as well and that’s actually the ideal cognitive prosthetic.
Babbage breaks down face-name learning in the Intro app through two key principles of Errorless Learning and Spaced Retrieval.
Babbage: Errorless learning research demonstrated that for people who have any difficulty remembering something, the last thing you ever want to do is make a mistake. We say people learn from their mistakes, but actually mistakes very powerfully compete with your ability to correctly recall the right answer.
The right way to learn something new, if you’re going to have any difficulty remembering it, is to always get it right every time. Use as many supports as you need to make sure you keep getting it right and then gradually remove those training wheels, so you continue to get it right without ever making a mistake. That’s ‘errorless learning’—don’t let yourself make any errors.
The tricky thing is it’s hard to give yourself an errorless learning experience. In trying to learn someone’s name, the real world might only give me three or four opportunities to hear it before it’s assumed I should know this person’s name—and maybe I need to hear that name thirteen times. It’s socially difficult to get the extra ten, so if you have a tool like Intro that can help you get that extra practice, that’s great.
The other principle underlying Intro’s name learning tool is spaced retrieval. The way your brain decides what you still need to store is do you use that piece of information or not? The things that are still going to be retrievable later are the things you’ve continued to access over time. Your brain says, ‘this is clearly important—I will make sure to maintain and improve the connection of this information’—and so you can recall it again. Again, activating this function is powerful. You can make efforts to do this yourself, but it’s fairly impractical to do this systematically without a tool that ensures you’re getting this carefully graded experience.
Q: What are cognitive prosthetics and how do you think they will be playing a part in future healthcare?
Babbage: A cognitive prosthetic is any tool that helps your brain processes be more effective than they otherwise would be. They should be something you want to integrate with your day to day life, be as easy to use as possible, and be something that will continue to support your functioning over time, rather than being a training tool that just goes away.
In the context of people who have a neurological impairment of some kind there are a few things that make a cognitive prosthetic better. One of them is that it’s something that is desirable to use—something that looks like something your peers would use. So a smartphone is great because it’s normal… there’s nothing abnormal about it. The third thing that makes a great cognitive prosthetic is if you can pick it up and use it correctly the first time, without any training. That’s the holy grail for a cognitive prosthetic, because it means that even if you forget any training you have received, you’ll still use it correctly. And that, of course, contributes to an errorless learning process.
Usability testing and thinking about how to simplify as much as possible the tools that we’re creating is really important. At the recent /dev/world software development conference in Melbourne (26-28 August), I spoke about practical approaches to cognitive accessibility in software design. One of the starting points of that talk was suggesting to software designers and devlopers that thinking about designing software for people with brain injuries is a really good idea, because we often overcomplicate our solutions. This therefore leads to better tools for everyone. I also talked at that conference about mechanisms for providing users with choice about the level of complexity they want to have in a tool like a mobile app—would they prefer the standard interface, a simpler interface that provides only the most central one or two features of the app, or do they want expert mode, with maximum features and controls? Giving greater flexibility again provides better experiences for everyone.
Q: What do you imagine healthcare to look like in 10 years with reference to technological advancements?
Babbage: I think the future of technology in healthcare is in most cases not nearly as sci-fi as people imagine—but hopefully it's much more important in terms of actually creating some useful change. A big part of that is going to be solving how we close up the gaps in health information storage in ways that give end users both more control over their information and also more access to it—enabling them to get a better insight into what’s going on in their own health.
Personalized medicine is often the term which is used, referring to analysing your genetics and then deciding what treatment you should receive on the basis of your genome. There’s a cardiologist at Auckland Hospital that has a genome sequencing machine on his desk. By sequencing the genome of patients he’s able to ensure that patients who would not benefit from the standard medication can go straight to one that will actually protect them.
Babbage carries on to explain hearing about a subgroup of Pacific Islanders that aren’t responsive to the first line of heart medication so through using this system they’re able to be put on the treatment that will actually work. Babbage says this is a great practical example of personalized medicine actually at work in New Zealand, stressing the importance of making sure information doesn’t get into the wrong hands and is there to better the people of the country.
Babbage: It’s about increasing population health, vitality and our autonomy as New Zealanders—rather than feeding into the disparities between rich, super healthy people and poor downtrodden people.
Q: What is an area in health that you feel has a lot of potential for positive change through introducing or increasing technological advancements?
Babbage: Clinician efficiency. I just think a lot of us, not just clinicians, are doing work for computers instead of the other way around. A focus on the workflow of clinicians would unlock significant benefits, through thinking about how can we help them do more of the parts of their work that are unique and spend less time on administration and data and all other things that create friction in their day. It would also lead to happier clinicians, and happier health service users. This is not about complex health technologies, it’s about doing the hard work of implementing solutions to make things work better.
Q: What opportunities can young professionals coming from data science and non-health related backgrounds expect in health technology, when it comes to skills that would be useful for the industry?
Babbage: In reference to health, it's really about drawing on the expertise from the people who have thought long and hard about these things—because if these were easy problems to solve they would’ve already been solved.
On the technology side, there’s an enormous amount that can be contributed, because people in the health system often aren’t technically savvy. But what’s really important to understand is what a technologist brings to a situation needs to still be focused on human behaviours as the primary question.
Build expertise in co-design, engaging with end users, and really understanding service-user perspectives and clinician perspectives. Understand why these end users might not be enthusiastic about your particular technology, and look at how it can actually make their life better, rather than harder. Thinking about what it would really mean for them to be using this in their practice as opposed to what do you think of my tool. If you can straddle those two worlds by bringing a technology background, you’ll be making a valuable contribution.
Reporter Anzel is interviewing Mr. Duncan Babbage
Author: Anzel Singh
Photographer: Victor Chateau
Editor: Christiane Zhao
Copyright By HTA.