Michael Bird:
Hello, hello, hello. And a very big welcome to Technology Now, a weekly show from Hewlett Packard Enterprise where we take what's happening in the world around us and explore how it's changing the way organizations are using technology. We're your hosts Michael Bird, coming to you from sunny-ish, UK.

Aubrey Lovell:
And from Florida Aubrey Lovell. And in this episode we're exploring healthcare and how emergent technology can help us prepare for the future ahead. We'll be looking at ways in which AI is making healthcare more efficient and overcoming roadblocks. We'll also be examining the ethics of letting algorithms influence human outcomes. And of course we'll be looking at the books that are changing the way you and some of our previous guests see the world. And once again, if you like what you hear, subscribe on your podcast app of choice so you can stay connected. Michael, are you ready to go?

Michael Bird:
Oh, I'm ready to go.

Aubrey Lovell:
Let's get on with the show.

Michael Bird:
So it's been a pretty harsh winter by any standard here in the Northern Hemisphere. Shocking news I know. And that has meant a world of painful healthcare services. Every year, flu kills up to half a million people. And in the US alone around that number end up in hospital. Now with the additional strain of Covid, as well as injuries and delays caused by poor weather, that has put an enormous strain on medical facilities and stuff. In the UK, the winter rush took up as many as 3,800 beds a day. That's a five full hospital's worth of extra patients, which put national hospital capacity at nearly 98% and caused knock-on delays. Things like non-essential surgery on diagnostic clinics. It's so bad that in some parts of the world, including the UK, specialist data intelligence centers have been set up to try and spread the load on hospitals over winter. So we have to ask ourselves, what on earth can we do? And of course, why other than having staff off sick, does it really matter to other organizations?

Aubrey Lovell:
Well, this week we're joined by Andy Cachaldora, north European general manager for digital services at GE HealthCare. Andy, thanks so much for joining us. So we know that GE is really heavily invested in AI. What part does AI play in the digital transformation efforts of healthcare and why does it matter?

Andy Cachaldora:
AI is a massive part, especially when I think about where we will go in the future. It plays in all elements whether you are looking at a particular image and that image needs to use AI in terms of manipulation to make it clearer so that physicians can actually make a better diagnosis, right through to the workflow and pre-populating reports with data that's been captured from the image so that the physicians can report quicker to help with the backlog of Covid, right down to clinical decision support in patients that may have suspected cancer and trying to identify areas that a physician may not necessarily see quite clearly enough, to even predicting whether the patients are going to turn up for appointments. So it's in every single aspect of healthcare now. In the future it'll be even to predict whether patients are even going to be sick, but I think we're still a little bit away off from there yet.

Aubrey Lovell:
From your experience, what would you say right now is the biggest challenges facing AI in healthcare?

Andy Cachaldora:
I think there's two. There's adoption and then there's also the commercial models as such. The adoption, if you think about clinical staff, they're almost stuck in the weeds in terms of trying to do their day to day process that they don't have the headspace to really evaluate what is the best AI, and why do they need it in the first place. The other aspect is you have the markets small to medium sized businesses that are trying to develop AI and trying to bring that to market very quickly. It's very clinically led trying to prove the outcomes rather than the commercial models, which tend to stifle how AI's adopted because there's a reason why there's AI, but nobody can actually justify the cost for it and they need to go hand in hand.

Aubrey Lovell:
Speaking of outcomes, what specific outcomes has GE HealthCare seen using AI to improve that path to health and patient experience? What are you seeing from your side?

Andy Cachaldora:
Sure. So we've seen in the workflow specifically about reducing Did Not Attend. So patient adoption in terms of their appointments, so that clinicians can maximize their scanning time and use all their slots in scanning. Other aspects of AI is in clinical decision support. So for example, thorax and identifying lung nodules. And we do some fantastic work with NCIMI, Oxford University in developing AI and being able to structure AI and create quality data for AI to run in. There's more in progress. I think there's too many to list for this call.

Aubrey Lovell:
So a big question on AI, right? And we see this across industries, but specifically within healthcare is around the ethics of AI, right? And you talked about that piece around adoption. Where does AI become responsible for errors versus humans and how do we access or assess the decision making for this?

Andy Cachaldora:
Okay, I probably want to cover the first step first in terms of adopting AI in terms of ethics. There are a lot of AI companies that are offering AI free of charge. There's no validation for AI at the moment because it's not clinically used, if that makes sense. It's just for decision support. The front end of the ethics is trying to understand what the AI is trying to do and what the business model is. So for example, by utilizing AI, is it more cost effective than a human intervention? That's the first aspect of it. We could talk about the ethics around the force positives of AI and the challenges around that. And if you think about the capacity and demand model within imaging and diagnostics at the moment, not having AI has massive pressure on departments. There's no way we're going to clear the backlog of Covid and cancer pathways because we just don't have enough manpower.
Being able to actually prioritize them is absolutely crucial, whether it's accurate or not, if it means that you can prioritize a case and you can stop a patient from going from stage one to stage two cancer, that means survivalship for patients is massively increased. If you think about the treatment path between stage one and stage two cancer, it's approximately a hundred thousand pounds. So without a question, AI needs to be adopted. I think it's being more adopted now than ever, purely for the fact that physicians realize that they can't cope with the demand that's coming. Compared to possibly two and a half, three years ago, there was this fear of AI is going to take over our jobs. I don't think that'll ever be the case where it'll be a clinical decision support. There are other aspects around ethics as well around to utilization of data, privacy, impact assessments and things like that. So there's a lot of governance that needs to be put in place before AI is utilized in the lively environment.

Aubrey Lovell:
So what does healthcare look like in five to 10 years? And is there any specific predictions you have within that timeframe?

Andy Cachaldora:
Yeah, sure. So in 10 years time we will look at probably a connected health ecosystem. We're not quite there yet for a couple of reasons. One, in how services are commissioned, GPS are not connected to primary care, and primary care is not connected to secondary care or acute trauma centers, et cetera. I think what we'll see is a consolidation of data between the system to be able to actually, A, get a single patient timeline of your patient records. So despite where your healthcare's been delivered, everybody has access to your data at any given point, so they can really tailor what your care might be for you as a particular patient.
But also we will no longer wait for patients to become ill to treat them that actually footprints of hospitals be reduced and they'll be reduced to certain extent that actually early intervention will already be active in the marketplace where we could predict whether a patient's going to have cancer or not and treat them early enough, so they don't have intrusive surgery. We could predict whether people are going to have a heart attack before they have it so that a stent could be put in place. That means that hospitals are going to be very proactive in terms of their healthcare delivery, but actually patients end up living longer and managing their condition in a more effective way.

Aubrey Lovell:
So a closing question for you Andy, is when we think about businesses across other sectors, from your perspective, what can businesses learn from the way AI is being implemented in healthcare?

Andy Cachaldora:
I've seen some AI around emotional intelligence and voice recognition. So this is around trying to understand whether patients would turn up for their appointment or whether they would follow their treatment path, for example, that's been recommended to them. And obviously some of the medication and some of the radiotherapy they might need to have is very intrusive and obviously patients are going to say yes, they're going to do it, but actually the emotional intelligence within AI in recognizing that whether they will stick to their treatment path adds a lot of value because then you could try and tailor that to particular patients.
That's still very relevant to how people buy services and understand the emotional intelligence with buying goods, irrespective, whether it's healthcare or not, or whether there is a good service in customer service, et cetera. We always talk about how bad service is, but we never talk about how good service is. And trying to link that to the emotional intelligence is very rewarding for retailers. So I think there's a lot of crossover that could be learned from both sectors. I don't see the sectors working closely together, but I think within healthcare that will come closer together, for example in wearables, the Apple Watch and other wearables, where actually collecting that data and showing a physician who'll add value later on.

Michael Bird:
Thank you Andy, that's given us all something to think about. We'll drop some useful links about some of the stuff that we've been discussing in the show notes. And we'll be back in a moment with our questions for Andy from the audience. So do not go anywhere.
Aubrey Lovell:
Okay. It's book club time. We are opening up the floor for you to give us your recommendations on books which have changed the way you look at the world, life, and business in the last 12 months.

Michael Bird:
And if you want to share your recommendations, there's a link in the podcast description. Just record a voice note on your phone and send it over. All right, let's roll the clip.

Erin Young:
My name is Erin Young, I'm a research fellow in the public policy program at the Alan Chewing Institute. And one of the books that has changed my year is called Programmed Inequality by Mar Hicks. It discusses how Britain lost its early dominance in computing by discriminating against women workers in the industry. During the advent of computing in World War II, women made up a vast number of the computing workforce in the UK, women were the original computers. But as money and power began to enter the industry, more men also began to enter the industry and women were sidelined into subspecialization. And we see this starting to happen in data science and AI again now. And so it's a very important book to read, which takes a historical perspective by which we see probably happening again in the current AI workforce.

Michael Bird:
And once again, do keep your book reviews coming in via the link in the description. Andy, have you read anything recently that you can recommend to us?

Andy Cachaldora:
To be honest with you, I haven't read a lot of books recently because I've been inundated, it is the end a financial year. But I am looking forward, I ordered a book recently, which is called Remote Presence - A Practical Guide To Communicating Efficiency In A Remote Environment, which is by a good friend of mine, Sarah Brummitt, which is actually looking at the new models of managing workforce in a remote environment. Obviously on the back of Covid, things are dramatically changed and trying to make sure your staff are well performing is key to an organization like GE.

Aubrey Lovell:
Fantastic. So this is actually where it gets a little more fun because it is time for questions from the audience. Our listeners have been sending in your questions to Andy on the topics of healthcare. So the first question, Andy comes from Laura in Manchester. She wants to know if you think we'll reach a stage where machine led diagnosis and triage will only need human sign off. For example, are we heading towards machine led rather than machine assisted medicine?

Andy Cachaldora:
Yes, I think we will. I think we are very, very far from it. If I think about the last two to three years, everything's about Covid recovery, et cetera. And there'll be a tipping point where we have captured or caught up with Covid and diagnostics. And then what we'll start to see is what we call vague symptoms where you may feel some pain or ache and go for diagnostics, not actually knowing what you might have. And that's where the machine learning will try and tease out suspected things that might be wrong with you, with the intention of capturing them really early and then being referred to a physician.

Michael Bird:
Meanwhile, Ben from Sydney would like to know what healthcare professionals or really anyone working with AI should be researching right now to prepare themselves to stay on top in the next five to 10 years.
Andy Cachaldora:
I think there's a lot of research around AI itself. I think what's really needed is organizations almost doing an exchange like you would do with students where you swap staff to share their experience. I think that's where the value will be in the years to come, bringing both industries closer, working collectively rather than partner customer, relationship.

Aubrey Lovell:
So it's like reverse mentoring essentially.
Andy Cachaldora:
Yeah, absolutely.

Aubrey Lovell:
Thanks Andy. And again, we'll drop a couple of links in the podcast description for more on these topics.

Michael Bird:
Right then. We are getting towards the end of the show, which means it is time for... You're ready, Aubrey? This weekend History.

Aubrey Lovell:
This weekend History.

Michael Bird:
Yeah, okay. Which is a look at monumental events in the world of business and technology, which has changed our lives. Over to you Aubrey.

Aubrey Lovell:
Well, the clue from last week was, you won't miss Melissa. So of course you might be thinking, "Hey, it's Courtney Cox." The love interest Melissa Robinson from Ace Ventura: Pet Detective. Great movie. But hang on, that is not what we're talking about. It's the 1999 appearance of the Melissa computer virus, the first major virus to be spread by email. Within just a few days of its release, it had infected over 200,000 computers and it actually didn't do any direct harm, but it did self-replicate and sent itself on that generated so much email traffic that it completely crashed the email systems of dozens of major organizations. Fun fact, next week we're heading to 1880 and the clue is, it's no longer Eerie in Indiana. So keep that thought to yourself. We'll discuss next week.

Michael Bird:
Nice. Well, that brings us to the end of Technology Now for this week. Next week we'll be discussing working under stress with Claire Blackett, senior research scientist and expert in nuclear safety and human center design at the Institute of MG Technology in Norway. I promise you, it's going to be a fascinating one.

Aubrey Lovell:
Absolutely. And until then, thank you to our guest, Andy Cachaldora. And thank you all so much for listening. Technology Now is hosted by myself, Aubrey Lovell and Michael Bird, and is produced by Sam [inaudible 00:16:25] and Zoe Anderson. With production support from Harry Morton, Alicia Kempson, Allison Paisley, Alex Podmore, and Ed Everton. Technology Now is a lower street production for Hewlett Packard Enterprise. We'll see you next week.