#7: AI and Health

Ian Bowie
Hello and welcome to AI unfiltered with me Ian Bowie and our resident expert Michael Stormbom, where we will be talking about everything to do with AI in our modern digital society and what the future holds for all of us.

Alright, so today we’re going to be talking about how AI can provide an enhanced healthcare experience for patients. Is it possible that AI will soon become the primary decision maker in prescribing Medication to patients?

Michael Stormbom
I don’t know if that’s a future to hope for, but I think there’s certainly room for significantly for a significant amount of automation. I don’t know if I would like to see a future where it’s completely automated, where there’s no human supervision whatsoever, but there’s certainly quite a lot that can be done with AI that doesn’t necessarily need a human hand.

Ian Bowie
But is your attitude simply there because we’re so used to having human interaction?

Michael Stormbom
Well I do think human interaction is what human beings need as well. So I would certainly say there’s definitely that aspect. So I mean, there there’s definitely great benefit to using natural language processing and analyzing for example patient journals there’s sort of a lot of these material that is very valuable from a research perspective.

Ian Bowie
So now that we’ve spoken about this before, but is this analysis all in English, or is it happening in other languages as well?

Michael Stormbom
Well, in many different languages, so for example, here in Finland, the primary language of healthcare is Finnish, of course, Swedish to some extent, but the vast majority is in Finnish. So there’s been a number of different initiatives around that topic.

Ian Bowie
If we go back to my original question about are human doctors actually going to be necessary in the future. It is not possible to have a hierarchy of healthcare, whereby at the very top of the hierarchy, you have AI algorithms, and then you have slightly lower down the hierarchy, perhaps nurses and healthcare workers, so the AI algorithms actually replace the doctors, but we keep the nurses and the health care workers, would that be possible? It’s not that I don’t like doctors. Don’t get me wrong, but they are the most expensive component in the public healthcare system.

Michael Stormbom
And we, I mean, we certainly have a lack of other healthcare workers as well. So I think, I mean, we should certainly save the resources we have so that they are available when they are really needed.

Ian Bowie
Could we then take the savings that we could make by not having human doctors and pay nurses and other healthcare workers a higher salary, which would then attract more people into the, into the vocation and that would solve the lack of healthcare workers problem, wouldn’t it?

Michael Stormbom
That sounds like a good goal, at least. Particularly administrative tasks are those that are ripe for automation.

Ian Bowie
So do you think that’s one of the first areas of healt care that’s going to start benefiting from AI?

Michael Stormbom
I would say so. I think

Ian Bowie
In a big way.

Michael Stormbom
I would definitely say, yeah, yeah. But I mean, there are certainly already treatment of diabetes. So maybe you’re already these devices that you can basically put in your arm and carry around. which can keep a constant eye on your blood sugar for example. So that allows you verbally to eat and drink at the at the right times. And I’m in the next step is done on artificial pancreas in a way that will you take the data and then based on that it can sort of adjust the insulin levels. So I think that’s on on the experimental stage as far I know.

Ian Bowie
I was about to say we’re starting to diverge into the realm of science fiction again, I think…

Michael Stormbom
But that is definitely not…

Ian Bowie
It’s coming. I know, I actually watched it was a fictional television program on the television last night. And it was exactly what you’ve just described. It was a diabetic who had their system and it was operated via his mobile phone. And what the baddies had done is they hacked the mobile phone and stopped him getting his insulin and nearly killed him. So whether that’s, I mean, I don’t know I’m just wondering here out loud, that there are also pitfalls to all this technology.

Michael Stormbom
That there most certainly is and, well imagine them hacking pacemakers and whatnot and all that good stuff. So I mean, yeah.

Ian Bowie
But it’s very interesting what you said about the diabetes thing, because, again, you know, one of the questions I’ve got is that AI is also now being used as an intelligent symptom checker. But is there a danger that people will start to self assess together with an algorithm and not bother going to see a doctor at all?

Michael Stormbom
Yeah, I mean, there have been those WebMD and whatnot around for a long time. So yeah, I think definitely. There’s that risk. I mean, certainly, whenever I have some symptom, and then I Google it, and inevitably, inevitably, the diagnosis is cancer. Doesn’t matter what the symptom is.

Ian Bowie
It’s always cancer.

Michael Stormbom
It’s always cancer. Yeah. Yeah. So no, I think if you want to have this like intelligent diagnosis system, it needs to be actually intelligent, for one thing, and then I mean, you would reallyt need to explain to people as well that, even if, even if a symptom can be indicative of a particular disease, I mean has to still be put in proper context. For example, the likelihood of it actually being that thing.

Ian Bowie
How do we actually go about educating the masses?

Michael Stormbom
I think that’s a, I think that’s a difficult one. I mean, I think the if there were to be such an AI system, I think you need to be very carefully designed in how it communicates with the patient, so to speak. So I think that would be and well there, of course, would be room for of course, dialogue with an actual human being can put it in context. But yeah.

Ian Bowie
Of course. Another question I’ve got is are we also raising the expectations, that the medical profession will have on the patients themselves? I mean, basically, what we’re talking about is automating healthcare to an extent which has a demand on the patient that they can actually use the new system. So if you if you say to them, Well, you’ve got to now wear a wearable and know how to operate it. If you’re talking to somebody who’s maybe in their late 70s, early 80s, who has never really used technology at all. Isn’t that a bit of an unfair ask?

Michael Stormbom
Yeah, and I mean, I think in a future episode, we will definitely be talking also about this inclusivity of this new digital world, but I certainly think there’s a generational divide that the younger people and who have sort of grown grown up in the in the digital world are more keen to adopt these things than the other people, well…

Ian Bowie
You’re going to end up with with a percentage of the population who are going to be marginalized and excluded through no fault of their own.

Michael Stormbom
Absolutely. So I’m in digital heaven have not basically.

Ian Bowie
Yeah, but it’s not just about the digital haves and have nots, I mean, this whole healthcare thing you know, you take an area the size of Southwest Finland, which is actually rather large. And then think about centralizing healthcare, in the bigger city, which is Turku. So then if you live you know, 60, 70, 80 kilometers away from Turku, but you’re still in the southwest Finland region. And you need medical assistance. You’re absolutely stuffed.

Michael Stormbom
Yeah, or if you live in the archipelago, yeah.

Ian Bowie
So it really is starting to create a two tier society, or I think it’s, it’s starting to make people have to consider taking very hard decisions. Is it worth taking the health risk of living in these outlying areas?

Michael Stormbom
No, indeed. So in a way it will probably accelerate the urbanization process, possibly.

Ian Bowie
Yes. But at the same time, everybody wants to live where there are services, people who are living in areas where the services being reduced all the time, are going to have problems because actually nobody else will want to move there, which means that any property that they may own is worth basically nothing. So they’ve got no asset to sell in order to afford the expensive luxury of living in the city.

Michael Stormbom
Yeah, not no indeed and I mean, that is already a problem. I think. This was discussed quite a bit under the previous government. We were talking about this, getting people back to work, but if you are stuck with then with a house that you can’t sell. So I mean, you are where you are.

Ian Bowie
So all these people who moved out of the city, when they were told that they can work remotely and they bought houses, in outlying villages and towns are probably going to live to regret that decision within the next maybe five years.

Michael Stormbom
Possibly so, definitely. And I mean, it’s not just about healthcare. I mean, the postal service is not working at all into these remote areas and so forth and even Internet access isn’t a given, like high quality internet access.

Ian Bowie
Yeah. But basically everybody’s going to be living in a city. And those that don’t have just got to take the chances. There isn’t a stable internet connection even AI can’t help. No, no, because it relies on it.

Michael Stormbom
Very much so. And I mean, the trend has been to use ever more bandwidth so it’s not even that there would go into this sort of like lightweight that you can manage with dial up kind of thing.

Ian Bowie
All right, well, we’ve dealt with symptom checkers and diagnoses and possibly administering drugs. But AI is also being used to develop new medicines. Now in light of the COVID pandemic, do you think AI will make it possible to develop medicines and vaccines more quickly in the future?

Michael Stormbom
I think there’s certainly a possibility for that because it is all about processing data quickly and finding correlation between different concepts and different different things in the data. So I certainly think that data, that AI and is or AI methodologies can play a big role there.

Ian Bowie
So do you think that in five years time ever another pandemic comes along? We might even have a vaccine within a couple of months, just using AI algorithms?

Michael Stormbom
Well, I don’t know. I’m not a medical researcher. I don’t know how that goes. I’m sure AI can help by we think there still requires human…

Ian Bowie
You love your human interaction, don’t you.

Michael Stormbom
I do. Yeah. It’s been a long two years.

Ian Bowie
Stuck at home on my own on the sofa. Yeah, me too. But I don’t, actually it’s quite interesting that because the longer I’ve been stuck at home less, I’ve needed any human interaction. And I’m not saying it’s turned me into a hermit, but I quite like the idea of embracing technology that reduces my need to actually have human interaction. And maybe it’s come from the whole idea during the pandemic that you have to keep a safe distance from other people, that if we reduce our human interactions, we possibly also reduce the chances of catching anything nasty, and having to come into contact with our healthcare services at all.

Michael Stormbom
Yeah, well, I mean, certainly. And of course, during the corona pandemic, there’s been less, less flu going around as well, because of course, people have been social distancing then. So I mean, certainly. And I think there is one place where AI can play a big role in the future of healthcare, which is to well, preventative healthcare that helping people take better care of themselves. So they will need less healthcare.

Ian Bowie
Can you expand on that? So how…

Michael Stormbom
Well I mean, think for example, how many AI can help for example keeping your diet in check, for example, or measuring your values continuously and so you can take appropriate action at the right time. So I mean, the diabetes example is a good case, a good example of that. So I mean, I think they’ve been doing that here in Turku that you can get this sort of device that you can plug into yourself and you can measure on your own and… I mean, the more precise data you as the patient have ,the more precise and timely action you can take and that will certainly reduce the risk of complications in the future. And I mean, of course, not just for like physical healthcare, but in terms of mental healthcare as well.

Ian Bowie
But of course, I mean that that’s the issue of the moment, isn’t it? Mental healthcare. That’s what everybody’s talking about.

Michael Stormbom
It’s a big one. Yeah.

Ian Bowie
But where’s it come from? Has has the need to address mental health come from the fact that there is actually far too much technology around?

Michael Stormbom
Well, I do you think that sort of like, cognitive overload that we are constantly exposed? Or do you think that that plays a big, big factor. Well, another thing is, of course, that there’s been success, successfully cutbacks to mental healthcare in Finland and certainly across the rest of the Western world for sure. So I think there’s also not enough emphasis on preventative health care there as well. While also in terms of for example at work, I mean, there are more people who are burned out than ever before. So I think also the increasing demands increasing productivity demands I think is a is certainly a contributing factor there.

Ian Bowie
So technology causes increased mental degeneration within the population, and then we use technology to try and fix it.

Michael Stormbom
Yes,

Ian Bowie
It just sounds like a vicious circle, doesn’t it?

Michael Stormbom
The technology is the gun and the bandaid we put on the wound.

Ian Bowie
Yes, so you’d be better off going back to the farm and planting potatoes.

Michael Stormbom
Well I mean, I think that’s I think it has literally been even shown in in research that if you just disconnect from all the social media and whatnot that your your mental well being will be far far far improved. So I don’t think that’s even a controversial finding in the first place.

Ian Bowie
But I think I don’t know what you think. But I think technology anxiety, actually worrying about technology and having to learn new technology is also quite a big contributing factor.

Michael Stormbom
I would say so.

Ian Bowie
So mental problems, mental health problems.

Michael Stormbom
I mean, it’s certainly a stress factor. I mean, how am I going to earn a living in the future if I don’t understand these things, for example? So I mean, I think that’s certainly a contributing factor and…

Ian Bowie
Or how am I just going to function as a normal member of society? How am I going to pay my bills?

Michael Stormbom
Yeah, indeed, I mean, if I can’t, can get access to healthcare, I need to figure out this, some device to be able to talk to someone and so for me, certainly. But I think the other thing when also when it comes to the anxiety is , we have spoken about social media, and media manipulation before, but I think also this constant fear mongeringor the clickbait like keeping people constantly afraid so as to get you to read their, read the articles and get the ad impressions, I think that’s certainly contributes to a sense of anxiety as well.

Ian Bowie
Now is that not an area where AI could actually be a huge help? I mean, could we not set the AI algorithms upon the scaremongers that these algorithms would be designed to identify fake news and fear news? And basically just wipe it off the planet?

Michael Stormbom
Yeah, and of course, AI is currently being used to push that sort of fake…

Ian Bowie
But not enough obviously. No. So when you say AI is being used to push the fake news, yeah.

Michael Stormbom
So we need to we need AI to fight AI.

Ian Bowie
Yeah, absolutely. So now we’re going into Terminator territory, again, machines fighting machines.

Michael Stormbom
Pretty much pretty much.

Ian Bowie
One of the other things that AI is increasingly being used for is actually to record health records, isn’t it? Is there a danger that some information might get misrecorded?

Michael Stormbom
Well, I have a personal experience of a human error, and nothing nothing severe, but, I was going to the doctor one day and he had had some other patient’s like patient journal open. So he then managed to put in my data into that patient’s… Well, luckily, he noticed that just as I was about to walk out the door, but I was like, okay, that’s not reassuring. And yeah, but that was an example of human error. So I mean, certainly I see that that will definitely be the place for AI to prevent that sort of thing from happening.

Ian Bowie
Absolutely. Yeah. And the final thing that I’ve got on my little list of questions is, AI is at the forefront of what is now termed telemedicine. So the remote treatment of patients who use body sensors or wearable devices, to be monitored for their own health. What do you think of the benefits and also the potential pitfalls?

Michael Stormbom
Yeah, I mean, we talked about people who live in distant areas and don’t necessarily have access to the same or possibly in the future won’t have access to the same amount of health care that you would in while living in a city. Well, okay, that’s already the case in Finland. Let’s not pretend otherwise, but, but I mean, certainly, if you live in a remote area that sort of telepresence and telemedicine that might be the only only way of getting timely health care in a way. In terms of pitfalls, well, I mean, we’re talking about devices that are connected to the internet. So there’s certainly a risk with certainly a risk with that. And it can be of course, device malfunction might be hard to deal with remotely and so forth. Or devices just showing the wrong measurements could be, due to a device fault, could be… So there are certain risks there as well that one has to be mindful of for sure.

Ian Bowie
But then of course, you have to imagine I suppose device malfunction or something of the same of the sort. You have human error as well. So yeah, no, I mean, certainly, I mean, the machine the machines have only got to be better than the human equivalents.

Michael Stormbom
No, indeed, indeed. And I mean, there are other cases where AI is used in medical imagery, where it has been shown that it’s better than the human doctor at, for example, diagnosing particular types of tumors. So there are certainly those sort of use cases where, where the AI already outperforms the human doctor.

Ian Bowie
How far away do you think we are from AI in the healthcare? Yeah, how far away do you think we are with before AI is an intrinsic and integrated part of healthcare? It isn’t at the moment is on the periphery. I mean, it’s a little bit diagnosing this and diagnosing that, but it’s not a fully functioning integrated part of our healthcare system yet, but how far away are we from that?

Michael Stormbom
I think we’re quite many years away from AI really playing a very significant role in healthcare.

Ian Bowie
Do you think… again, I’m going to the realms of science fiction here because I’ve seen this on films but do you think that we might end up one day with a an AI enabled self care medical kit? We don’t feel very well with us literally hook ourselves up. Dum de dum de dum. Immediate diagnosis, this is what you need. Go and buy the drugs or order them or whatever. And off you go again.

Michael Stormbom
I would say that probably sort of like a universal diagnosis machine. I don’t see. I don’t necessarily see that. I mean there are certainly those individual things that they can, I mean, we can take those measurements already and draw some conclusions.

Ian Bowie
That’s now I’m talking about the future. I’m talking about, you know, an all knowing almost in your pocket doctor.

Michael Stormbom
That sounds a tad optimistic. I would say.

Ian Bowie
Maybe in our lifetime.

Michael Stormbom
Maybe in our lifetime. Yeah.

Ian Bowie
But then, you know, a couple of 100 years ago, people were saying to Cat optimistic I think that people would fly, a tad optimistic to think that you would have a telephone or a tad optimistic because then you go to the moon or whatever.

Michael Stormbom
Archaeologists will dig up this, this taping in 200 years time and then. Boy how little they knew.

Ian Bowie
What a clever guy. Exactly, indeed. Yeah. You know? Absolutely. When humans are part machine already. Yeah.

Michael Stormbom
Well, I mean, of course, in the future, if we’re all cyborgs, anyway, maybe diagnosis will be.

Ian Bowie
Maybe that that is a secret to immortality that you have got your own knowing in your pocket doctor. And it can keep you going forever.

Michael Stormbom
Yeah, that or you can sort of like grow new organs in a laboratory and then just…

Ian Bowie
Plug them in.

Michael Stormbom
Yeah. When one start failing just put in a new one.

Ian Bowie
So I think you know, I have no problem with the idea of growing organs in laboratory in isolation of anything else.

Michael Stormbom
Yeah. And I think that’s already in the experimental phase.

Ian Bowie
I’m not entirely sure about growing organs in other animals like basically pigs, that to me strikes have something else. But yeah, growing it from a Petri dish. Why not? Absolutely.

Michael Stormbom
Yeah. I mean, if you managed to grow an organ, that is genetically you, you would get around to problem for example, with of potential rejection and so on.

Ian Bowie
They can grow all the organs in the world, but they’ll never be able to grow a brain.

Michael Stormbom
Well, no.

Ian Bowie
And that’s gonna atrophy and die one day, so-

Michael Stormbom
That is true.

Ian Bowie
Yeah. Yeah. So we’re basically stuffed. We’re gonna die one day, but it might be.. the conclulsion from today is that we’re all gonna die one day, but AI can probably help us live longer.

Michael Stormbom
Yeah, and perhaps more importantly, we’re gonna live healthier while we are on the planet.

Ian Bowie
Of course, the healthier you are, the longer you should live, theoretically. Yeah. And of course, the best news of all, it’ll be a lot cheaper for the government.

Transcribed by https://otter.ai