People Power and Machine Learning
How collaboration could pave the way for super-efficient cancer testing
The success of the National Health Service (NHS) is all about people working together in partnership. We’ve just celebrated the great British institution’s 70th birthday this summer, and it’s hard to imagine how we could have gotten here without everyone pulling together, from the doctors and nurses on the front line to the managers and administrators making everything work. And, of course, the pathologists and laboratory medical professionals diagnosing illnesses.
Seven decades on from the founding of the NHS, the technology available to us has moved on significantly – but the principle of working together still holds true. And so I am delighted that Greg Clark (UK Secretary of State for Business, Energy, and Industrial Strategy) has announced a new collaborative effort, the Northern Pathology Imaging Cooperative (NPIC). The initiative will bring together the NHS and academia to pioneer the use of artificial intelligence (AI) in digital pathology, and will be supported by £10 million of investment from UK Research and Innovation as part of the Industrial Strategy Challenge Fund, as well as a further £7 million from industry partners.
It’s no secret that the NHS is facing unprecedented challenges: an aging population, long-term medical conditions, and significant funding issues. Leaders in the healthcare space have to find innovative ways to do more with less, all without compromising patient care. New technology, such as AI, presents an opportunity to help alleviate some of the pressure while also offering the potential for revolutions in what healthcare can achieve.
In pathology, AI can be trained to recognize the patterns of disease; for example, searching for small areas of cancer in a large sample. In short, pathologists will be able to use AI to diagnose cancer faster, better, and at lower cost – obviously an attractive prospect for the NHS.
But to work, AI needs to “learn” the patterns by looking at large numbers of images and becoming familiar with them, just like any human pathologist would. But there are no large collections of digital images that can be used for this – and computers can’t learn from glass slides like we can. The NPIC brings together the NHS, industry, and scientists to solve this problem. To begin with, the collaboration will cover a network of hospitals serving a population of around 15 million people. It is led by the University of Leeds, working with another six universities and 10 industry partners. Together, we will work to make new AI systems that can analyze our images and make better diagnoses. Crucially, everyone involved in the project will ensure that AI systems can be used safely, and that doctors and the NHS are in control of how they are used. However, it’s no less important for us to ensure that the public understands what we are doing and trusts that we are using their health data appropriately and securely.
Our ambition is to develop world-leading methods for gathering and using data to make AI systems. It’s the perfect balance – a collaborative partnership that uses “people power” to make new technology work for us. The project, which officially began on 1 December 2018, will continue for three years, so no time is wasted in taking the next step towards improving diagnostic capabilities for the NHS. Adopting cutting-edge digital technology will drive continued improvement in our ability to diagnose disease accurately and target treatment by pinpointing biomarkers in individual patients.
The future of healthcare is incredibly exciting, with potential for revolutionary digital pathways to support exceptional clinical decision-making. Earlier and more accurate diagnoses save lives, and AI has the potential to take us to the next level in this regard – good news for healthcare systems, for pathologists, and, most importantly, for patients.
Geoff Twist is Managing Director at Roche Diagnostics UK & Ireland, Burgess Hill, United Kingdom.