Technology isn’t taking over – we have the power to shape it
Pathologists and hospital administrators can shape the future of our pathology services, because we are in control of how we implement digital pathology and artificial intelligence. Together, we can transform cancer diagnostics and improve patient care.
In my last Blog, I introduced one of the hot topics of the Future of Pathology project: digital pathology (DP) and artificial intelligence (AI). I spoke to a number of healthcare leaders and pathologists at the start of the project in order to understand what they see as the challenges and priorities for cancer diagnostics. They agreed that DP is increasingly being used instead of traditional microscopes for primary diagnosis in pathology labs around the world, and that it is paving the way for the future use of AI. They also agreed that whatever challenges these new technologies may bring, they are not an optional extra, and they have the potential to bring real benefits to patients.
As a great advocate of DP myself, I can see how it is transforming our pathology service in the Leeds Teaching Hospitals. However, I want to dispel the fear that DP and AI are moving so fast that we won’t be able to control the changes they bring, and I also want to address the suggestion that AI will make the pathologist redundant. These are not only the concerns of pathologists but of hospital administrators too, because both groups are responsible for making the best use of resources to optimise diagnosis and monitoring. Dr Darren Treanor, a consultant histopathologist and lead for digital pathology at the Leeds Teaching Hospitals NHS Trust in the UK noted that, while digitisation of pathology images is already happening, adapting our services to these developments is “not going to happen overnight.”
So how do we ensure the steady integration of DP and AI into our workspaces and introduce positive change that benefits patients, pathologists and their clinical and hospital administrator colleagues? The answer lies, I believe, in working together at every step to manage the change. A survey in 2017 suggests that DP has been introduced by around 60% of pathology labs in clinical and academic institutions in the UK.1 In the US, adoption of DP has likely been on the increase since the FDA approval of whole slide imaging tools.2 I would recommend support of and investment in DP to hospital administrators. I would ask pathologists to think about how DP systems might be built into existing workflows in order to increase efficiency and diagnostic accuracy. In the Future of Pathology report, I look in more detail at the power of images to further our aims as pathologists. But for now, just consider the difference it can make to have a microscopic image on a computer screen rather than on a glass slide: how much easier it is to share with others for their opinion on a case or for teaching purposes, and how much less likely it is that such an image, unlike the slide, can get damaged or lost. We have the opportunity now to establish IT systems and invest in scanning equipment, to adapt our workflows if required and shape our environment to optimise these benefits for patients.
Opening the door to DP enables us to open a further door to AI. AI apps are beginning to come into use in pathology, mostly in supportive tasks including tumour detection and grading. With the advent of deep learning, AI apps could, with time, be able to carry out more complex tasks. How will pathologists work with colleagues, other healthcare professionals and hospital administrators to harness these benefits when they arrive in the not-so-distant future? For now, I would suggest that pathologists share their evolving experience of AI across the care setting and beyond. In this way we as a profession can develop a vision for how we integrate human and technological capability as the latter changes and develops. We need to consider how we use DP and AI as tools to execute specific tasks, enabling pathologists to focus on tasks that benefit from the human insight and creative thought we can provide. As pathologists, we are better able to answer challenging questions and diagnoses than DP and AI can, and can focus on drawing conclusions from data, rather than filing slides or producing data.
We pathologists need to work with hospital management and other clinicians to identify an efficient and effective diagnostic and monitoring pathway which makes the most of human, technological and financial resources to give the best outcomes for our patients. Good communication is central to this kind of collaboration, and you will be able to read more about how to achieve that in the report by my Future of Pathology project partner, Dr Jerad Gardner.
There will doubtless be hurdles to overcome as we move forward: questions of timing and cost of app development, regulatory issues and ethics approval associated with patient data, for example. There is no better time than the present to look around us and take stock. Pathologists, administrators and colleagues can work as a team to pool their ideas, experience and expertise. Please read my chapter in our Future of Pathology report to join the discussion.
Follow #TheFutureOfPathology on Twitter, Facebook or LinkedIn.
1. Williams BJ, Lee J, Oien KA, Treanor D. Digital pathology access and usage in the UK: results from a national survey on behalf of the National Cancer Research Institute’s CM-Path initiative. J Clin Pathol 2018;71:463–466.
2. Food and Drug Administration. FDA allows marketing of first whole slide imaging system for digital pathology. Press release 2017. www.fda.gov/news-events/press-announcements/fda-allows-marketing-first-whole-slide-imaging-system-digital-pathology Accessed 28 January 2020
This editorial is part of the Future of Pathology series sponsored by Leica Biosystems; it reflects the views of the authors, in their individual capacities.