In part two of our round table discussion, our panel digs into the changing face of pathology as AI unlocks more diagnostic power in precision medicine.
The panel
How do we ensure AI supports rather than constrains pathology expertise?
Derek Welch: I'm quite bullish about this. I don't believe AI will create additional barriers or reduce professional autonomy. Instead, I see it becoming a collection of tools that make the pathologist's work easier and better.
I'd draw a parallel with technologies such as the electronic medical record. At one time, moving from paper records to digital systems was met with considerable resistance, yet few people today would want to return to chasing paper charts. The same was true of electronic signatures. Even relatively simple digital innovations encountered skepticism when they were introduced, but they're now an accepted part of everyday practice.
I think AI will follow a similar trajectory. Whether we're talking about workflow tools or diagnostic applications, these technologies are designed to be used by pathologists to improve the way they work – not to create additional steps or add unnecessary burden.
Most importantly, I don't see AI taking decision-making away from the pathologist. The pathologist remains responsible for interpreting the case and making the final diagnosis. AI should support that judgment, not replace it.
David Gibbs: For me, it all comes down to the use case. Before implementing any AI tool, you need to be clear about the problem you're trying to solve, the outcomes you're aiming to achieve, and you need to engage clinicians throughout that process.
There are two priorities that drive my thinking. The first is improving the quality of the service we provide. The second is ensuring we can cope with the growing demand facing pathology services.
Across the NHS, we're dealing with chronic workforce shortages, too few pathologists, limited laboratory space, aging equipment, and constrained investment. At the same time, workloads continue to rise. If we simply carry on working in the same way while demand increases by another 20 percent over the next few years, we'll be operating in an extremely challenging environment.
That's why I see AI as part of the solution. Used appropriately, it can help laboratories manage increasing demand without compromising quality. The objective isn't to reduce professional autonomy; it's to give pathology services the capacity to continue delivering high-quality care in the face of growing pressures.
Syed T. Hoda: If you look at what's happened in radiology, it's a useful model for what could happen in pathology.
Radiologists are handling far greater volumes of work than they did in the past, aided by digital technologies and increasing efficiency. Initially, that translated into greater productivity and better reimbursement. But over time, as technology became more capable, and workflows more streamlined, the economics changed. Efficiency alone doesn't necessarily protect the profession in the long term.
That means we have to ask a bigger question: what new value will pathologists create as AI takes over more routine tasks?
I think the answer lies in expanding the profession rather than simply accelerating what we already do. One area is companion diagnostics. I was speaking with a large pharmaceutical company this morning about the possibility that pathologists will increasingly move beyond diagnosis and become integral to treatment planning alongside oncologists.
In that future, AI helps to make today's diagnostic workload more efficient, while pathologists take on new scientific and clinical responsibilities. Rather than being defined solely as diagnostic specialists, they could become treatment pathologists, using their expertise to guide therapeutic decisions based on companion diagnostics and other advanced biomarkers.
To me, that's the opportunity.
David West: I completely agree. Pharmaceutical companies are developing increasingly sophisticated, targeted therapies that depend on accurate pathology and diagnostic testing to identify the patients most likely to benefit. In fact, there are billion-dollar drug development programs that struggle in clinical trials because they can't consistently generate the diagnostic information needed to enroll the right patients.
To me, that's a tremendous opportunity for pathologists. As precision medicine advances, their expertise will become even more central to the development and delivery of new therapies.
The challenge, of course, is that diagnostic complexity is increasing at the same time. Every new targeted therapy brings new biomarkers, testing requirements, and clinical guidelines. Left unmanaged, this complexity could become overwhelming.
Here we see why AI has a dual role to play. It can automate the routine administrative and technical tasks that burden the pathologist, allowing them to work at the top of their license. At the same time, it creates capacity for pathologists to order and interpret more sophisticated tests, contribute more directly to treatment decisions, and engage more deeply with the scientific aspects of precision medicine.
Rather than diminishing the profession, I think AI has the potential to expand the pathologist's role at a time when their expertise is becoming more valuable than ever.
Could AI actually increase the value and influence of pathology within healthcare?
Syed T. Hoda: Yes, I think it could. Whether it happens depends on the profession itself.
Pathologists need stronger leadership, more visible voices, and a greater willingness to engage with emerging technologies rather than simply responding to changes imposed from outside. If we take an active role in shaping AI, I think pathology's influence could grow significantly.
One idea I'm exploring is the concept of a patient-centered pathology clinic. Imagine combining AI-assisted documentation, computational pathology, and companion diagnostics within a clinical service where pathologists play a much more direct role in patient care and treatment planning. That's a very different vision of pathology from the one we have today, and I think AI could help make it possible.
The real question is whether the culture of pathology is ready to embrace that kind of change.
History suggests that if a specialty doesn't seize new opportunities, another specialty often will. Radiology is a good example: in the 1980s, radiologists were primarily interpreting imaging studies. Today, interventional radiology is a major clinical discipline, with image-guided biopsies and minimally invasive procedures transforming patient care. Had radiologists not expanded their role, many of those procedures would likely have remained within surgery.
I think pathology is at a similar moment. We can already see new opportunities emerging around precision medicine, companion diagnostics, and AI-enabled clinical decision support. The important thing is to recognize that shift early, and to help shape it, rather than waiting until those responsibilities have been claimed by someone else.
David Gibbs: I agree with much of what's already been said, but I'd make one distinction. I don't think digital pathology or AI, in themselves, are the primary drivers of change.
The bigger force is the move toward precision medicine and increasingly personalized treatments. As therapies become more targeted, diagnostics become more central to clinical decision-making. That inevitably shifts greater responsibility towards the diagnostician.
I think that creates an opportunity for pathologists to play a much more active role in patient care. Rather than working solely behind the scenes, they could become more directly involved in diagnosis, treatment planning, and clinical decision-making.
That opportunity extends well beyond histopathology. We can imagine pathologists leading specialist clinics, seeing patients directly, and, in some cases, taking ownership of elements of the patient's diagnostic and treatment pathway. The same principle applies across laboratory medicine, including disciplines such as clinical biochemistry and microbiology.
Derek Welch: Digitizing pathology fundamentally changes what a glass slide represents.
Once a slide becomes a high-resolution digital image, it's no longer just something a pathologist looks at. It becomes data – a rich source of information that AI can analyze in ways that simply aren't possible with conventional microscopy. In that sense, digital pathology provides the ideal substrate for AI.
The question then becomes: what can that data tell us beyond what we already see?
Increasingly, it can help us identify clinically actionable information more quickly, whether that's biomarkers, companion diagnostics, gene mutations or other molecular features that influence treatment decisions. As AI matures, it has the potential to extract those insights directly from routine pathology images, bringing pathology much closer to the center of clinical decision-making.
Speed matters here. The faster we can generate clinically relevant information, the sooner patients can begin the most appropriate treatment. Shorter diagnostic pathways don't just improve efficiency. They also reduce the emotional burden of waiting for answers, minimize delays to care, and decrease the risk of patients being lost to follow-up.
David West: From what we're seeing, pharmaceutical companies are eager to engage more closely with diagnostic labs because they recognize how central diagnostics have become to the success of modern therapies.
The real opportunity for pathology to grow its influence is here now. As David and Derek have both said, the driving force is the remarkable pace of therapeutic innovation, not AI on its own. We’re living through a period in which diseases that were often considered fatal even a couple of decades ago can now be treated, and in some cases, cured with highly targeted therapies.
Targeted therapies are reaching the market faster, and there will be many more of them. Identifying the right patient for the right treatment increasingly matters just as much as developing the therapy itself. This is elevating the role of pathology and diagnostics to the center of precision medicine and makes pathologists’ expertise even more impactful.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
