The rapid development of AI and deep learning tools has prompted many pathology specialists to grow increasingly optimistic about the role that digital pathology can play in assisting their work and enhancing their capabilities. Such innovations herald a new era of healthcare powered by tech-enabled diagnostic precision.
The transformation of pathology
Routine diagnosis in pathology involves the application of a stain called H&E (hematoxylin and eosin) to tissue samples on microscope slides. This highlights and distinguishes cellular structures, helping pathologists spot any abnormalities. While H&E staining is a crucial diagnostic tool, the technique has its limitations. Not only is it time-consuming, there is a high probability of varying interpretation and diagnosis from even just a single slide.
In recent years, parts of this process have benefited from digitalisation, like automated whole-slide imaging (WSI). This method, pioneered in the late 1990s, uses an automated microscope that scans a tissue section to produce a composite high-resolution image file (similar magnification as optical microscopes) that can be easily stored and shared.
The widespread adoption of WSI in pathology took several years, but today, tools and techniques are advancing rapidly – and so are the roles of pathologists, who are increasingly working as part of a broader patient care team. In this respect, AI is enabling them to gather more, and better, patient data to inform diagnosis, treatment and monitoring.
With the NHS’s chronic understaffing and increasing patient backlog additional pressure has been put on the need to provide rapid diagnosis of cancerous tissues for a pathologist’s already heavy workload. Rather than replacing expertise, AI serves as a powerful aid to pathologists, helping them work more efficiently and accurately.
Digital techniques can help minimise analytical errors while also freeing people from repetitive lab work, as well as evaluate images and identify details that the human eye could miss. It can also be incredibly cost-effective – a 2020 Deloitte research report estimated that in Europe, “AI could save up to 53 million hours of routine analyses for clinical technicians, linked to potential savings up to £755m (€883m) per year”.
Such eye-catching figures are notable in the UK for two reasons. First, an under-resourced NHS is desperately trying to optimise its expenditure as an ageing population places ever more demands on its services. Second, NHS managers have typically seen traditional pathology methods as excellent value for money, especially compared with more complex imaging techniques. These decision-makers are therefore likely to consider any move towards a process like molecular testing to be more costly. However, while this may be true in relation to the immediate, upfront cost, where AI-aided methods are being used to determine whether hugely expensive treatments are needed, the slight increase in the cost per test would be offset by wider, long-term efficiency savings.
The power of AI – offering a new perspective
While some people still need to be persuaded of AI’s potential in pathology, the pace of innovation is exciting. In a recent issue of Diagnostic Pathology, researchers at Ohio State University noted that advances in the field were unlocking opportunities across “anatomical, clinical and molecular pathology” while catalysing new solutions ranging from biomarker screening to outcome prediction.
Indeed, the bright future of digital pathology in the UK prompted the Royal College of Pathologists to issue a statement noting the “great potential for the development of AI to support the diagnostic process in pathology, especially image analysis in histopathology”.
The ability of AI systems to detect patterns, identify anomalies and accurately predict outcomes is remarkable. It’s an area in which Owkin, an AI innovator, is well placed to drive change. Two of the company’s more exciting AI diagnostic developments integrate digital pathology workflows to support accurate decisions at a fraction of the time and cost of existing tests.
The first is MSIntuit® CRC, the first CE-marked AI diagnostic that prescreens for MSI, a key biomarker used in the management of patients with colorectal cancer. It aims to have a significant impact on doctors and patients by decreasing workload and turnaround time and preserving tissue material and resources. By using AI, this innovative tool supports reproducibility by potentially addressing inter-observer variability, with the end goal of optimising quality and efficiency for critical tests and helping to facilitate better access to immunotherapy.
Owkin is also developing RlapsRisk® BC, a risk assessment tool for the recurrence in early breast cancer, designed to help pathologists and oncologists determine the right treatment pathway. Iain MacPherson, professor of breast oncology at the University of Glasgow, believes that this “innovative AI technology has the potential to address an important unmet medical need that could ultimately lead to better outcomes for patients with early breast cancer treated in the NHS”.
The digital transformation of the NHS
The UK pathology profession faces several obstacles, including talent shortages, growing caseloads and the need for more precise diagnostic capabilities. Demand is growing rapidly, with NHS hospitals’ pathology test volumes rising by a mean annual rate of 2.4% between 2012 and 2021, according to Source BioScience, a provider of histopathology lab services.
The digital transformation of the pathology ecosystem could solve many of these hurdles. A healthcare economic model (HEM)5 proposed by Source BioScience indicates that digital workflows can reduce the average pathology turnaround time by two days. This has been verified by testing at East Kent Hospitals University NHS Foundation Trust and validated by other trusts. The HEM also predicted that over a span of five years, a digital workflow would facilitate savings equivalent to over 8,000 patient lifeyears when compared to the previous traditional pathology workflow.
As the demand for their services continues to increase, embracing the latest technology will help ease pathologists’ workloads, reduce turnaround times and enable greater efficiencies without compromising patient care. This will demand a shift of mindset in the profession and encourage the adoption of digital methods, especially by newcomers to the field.
Owkin is committed to enabling this transformation. Working closely with its extensive academic network to develop robust digital solutions, the techbio aims to empower pathologists to work more effectively while making precision medicine more accessible to more patients.
To find our more about MSIntuit® CRC and RlapsRisk® BC, manufactured by Owkin France, please visit: owkin.com/diagnostics-approach
The rapid development of AI and deep learning tools has prompted many pathology specialists to grow increasingly optimistic about the role that digital pathology can play in assisting their work and enhancing their capabilities. Such innovations herald a new era of healthcare powered by tech-enabled diagnostic precision.
The transformation of pathology
Routine diagnosis in pathology involves the application of a stain called H&E (hematoxylin and eosin) to tissue samples on microscope slides. This highlights and distinguishes cellular structures, helping pathologists spot any abnormalities. While H&E staining is a crucial diagnostic tool, the technique has its limitations. Not only is it time-consuming, there is a high probability of varying interpretation and diagnosis from even just a single slide.
