Connected Pathology in the UK: Part 1

The elegant city of Leeds in the north of England plays home to a digital pathology revolution. The pathology department at the Leeds Teaching Hospitals (Leeds TH) National Health Service (NHS) Trust was the first in the UK to own digital slide scanners, back in the early 2000s. Since then, the team, along with the University of Leeds, has built up expertise and renown in delivering world-leading digital pathology innovation and research.

In 2018, the UK Government injected money into the NHS to stimulate digitalization as a platform for artificial intelligence. Pathologist Darren Treanor and his team saw this as the ideal opportunity to fully digitize the histopathology service at Leeds TH and connect it to other pathology departments in the region. The National Pathology Imaging Cooperative (NPIC) was born.

The initial grant was for a network of six connected hospitals in the West Yorkshire region of England. Around two years later, thanks to additional government funding from NHS England and the UK Office for Life Sciences, the program expanded to build a system for a wider region. But, piggybacking on the general improvements in information technology, computing power, screen resolution, and data storage technologies, Treanor and his team had a much larger vision – a digital pathology network for the whole of the UK’s NHS.

Fast forward to today, and NPIC has rolled out its program to six hospitals with 100 percent digitization. A further 30 have signed up to join – creating the world’s largest digital pathology network – and enquiries are rolling in from all corners of the UK.

Keen to understand what goes into executing such an ambitious and groundbreaking project, we sat down with five of NPIC’s leaders to learn what was involved and what a difference it is making. What unfolded was a story of extraordinary vision, drive, collaboration, and teamwork – all on a huge scale.

In the first of our NPIC story series, we asked Treanor to introduce us to the project and its aims.

What is your background and what is your role on the NPIC team?
 

I went to medical school in Dublin, Ireland, and also started my pathology training there. But I was keen to study abroad, to broaden my training, so I came to Leeds. That was mainly because it was a world famous center for academic pathology, but also because it allowed me to do a computing degree in the evenings at Leeds Metropolitan University. That training led quite naturally to a PhD in digital pathology, which was just kicking off at Leeds. 

I actually have a number of jobs today. My NHS job, which is my clinical commitment, is diagnosing liver disease. I also run research projects with PhD students at the University of Leeds doing artificial intelligence development, for example. 

On the NPIC program, which is probably my biggest job by far, I'm Director. I develop the program – with the help of Dal Bansal and Bash Hussain – and oversee its day-to-day running. 

In addition, I’m Professor of Pathology at Linköping University in Sweden. They were one of the first pathology departments to go fully digital in around 2010. Recognizing that the UK was lagging behind the Swedish use of digital technology, I took a post there, as a part-time professor of digital pathology to grow my own knowledge and experience. 

And then the final job I have is as professional lead for digital pathology and AI for the Royal College of Pathologists (RCPath) – our professional body in the UK, where I coordinate the development of guidelines and education.

It sounds like I wear a lot of hats, but there's a lot of synergy between the roles: things I learned here, I've shared with Sweden; my observations in Sweden feed into NPIC; and the RCPath work is informed by NPIC’s practical deployments.

What goals and values drive the team?
 

The official aims of the program are to deploy digital pathology in the NHS, and then use the infrastructure firstly for diagnosis, but also to assist the development, evaluation, and use of AI, and to be a platform for further research and innovation. 

Perhaps the core value that we place on this amazing technology is the many benefits to patients. Just today, we will scan samples from between 300 and 500 patients across the NPIC network; and those patients will benefit from it. 

At a micro level, we make sure we manage that service well so that, day to day, everything is running smoothly. But, at the macro level, we recognize that the system will only be optimized if we are highly strategic. We need to think big – about centralizing the data and the artificial intelligence on a national level. So it's a combination of doing things well day to day, but also thinking forwards. 

If I was trying to summarize my driving values, I would say it’s trying to combine the big and the small to make sure we do the best we possibly can. 

How do patients respond to the NPIC project?
 

I’ve probably demonstrated digital pathology and NPIC to a couple of thousand people so far, including members of the public and patients, and they seem to assume that every lab is already digital. They think, “Well, why would you not look at biopsies in that way?” 

We have a patient advisory group of about 20 patients and members of the public, led by Graham Prestwich, our patient lead. We’ve learned from the group that patients actually expect their NHS hospital to be using digital technology to improve care. They think it's nonsensical and inefficient for the NHS not to be digital. 

In reality, the NHS often struggles to invest in digital technology because of the upfront cost. But patients expect digital – they don't go into a GP surgery and think it's okay for them to write notes on a piece of paper any more… 

From our group, we hear that patients want to know whether they’re getting the right diagnosis from the right person quickly enough. For example, if a patient’s cancer biopsy needs a second opinion, it’s important to them that digital technology is used to get that second opinion quickly, which improves the quality of the diagnosis. There’s a personal benefit to them.

The other thing that our patients care about is that NHS staff are able to provide the best care together, working in networks. They tell us they don’t just want this fancy technology in the big teaching hospitals, they want it in their small, local district general hospitals or community hospitals as well. They like our centralized approach, because all the centers have the same digital system installed. Those smaller hospitals that sometimes suffer in terms of recruitment of staff and ability to maintain services – they get the same standard of kit as the larger centers. And that's the sort of thing patients tell us that they want from digital pathology.

What would be your “elevator pitch” to reassure skeptical pathologists about digital pathology?
 

Well, one of the things we’ve achieved as a team is to publish the “Leeds Guide to Digital Pathology” which really serves as our elevator pitch.

When we first started running our workshops, a few years ago, pathologists would ask us, “Why should I go digital?” We'd answer by explaining the benefits, the use cases, the business case and so on. So, if you’d asked me that question three or four years ago, I’d be telling you about significant pushback and skepticism from a vocal minority. I heard and agreed with some of their concerns and we improved the systems accordingly. 

But as you ask that question today, it feels like we've gone past the tipping point. Now people say to us, “We have decided to go digital; how are we going to do it?” There has been a massive step change, and now the more cautious people are in a really small minority. And I'm proud to say that some of the people who were anti-digital are now avidly using it. Because it's useful, and it works.

Bethany Williams, who coordinates our training initiatives, has loosely defined three groups of people in terms of digital pathology perception. Firstly, there are the “cheerleaders” who really want to go digital. In the early stages of the program, they were brilliant because they were willing to try it out – even though it wasn't perfect. 

Secondly, there are the “skeptics,” who are concerned that digital pathology isn’t safe or will undermine the pathologist or is a way to outsource pathology or reduce quality. They are actually very useful, because they are good at picking up problems and pointing out ways to improve systems. They can be critical, but we take those criticisms and use them to improve things. 

And then there are around 40 percent in the middle: the “uncertains.” In some ways, this group is the hardest to work with because they don't tell you one way or the other whether they like or don’t like it, or will or won’t try it. 

There are still some people who like using the microscope for certain tasks or think it's safer. However, when I log on to my computer, I can see everyone else who’s logged on to our digital pathology system, with a little green light beside their name, which means they are available for second opinions or queries. And that will include some of the people who would have been most vocally opposed to digital a few years ago. For NPIC, that's a great success.

What has been your highlight so far?
 

The thing that really gives me great satisfaction is when I’m using the digital pathology system, and I can see that a lab that’s just gone live is also using it. About a year ago, I was preparing a talk and I wanted to show an image from another hospital that was new to the network. I logged on to the system, opened a test image and zoomed in and, spontaneously, a little box and an annotation appeared – a simple box drawn around a piece of tissue, with the word “Wow!” beside it. Someone at that other hospital must have been testing the system and practicing with the annotation tools on the same test image, and left a spontaneous message of their impressions of the NPIC system. That little moment was really gratifying and makes all the effort worthwhile, especially when we know, at the larger scale, that is just one of hundreds of users in over 30 hospitals that will benefit from our work.