Why DICOM compliance is essential for digital imaging in pathology and beyond
Alan Byrne is Director of Global Marketing for Pathology at Agilent and serves on the Board of Directors of the Digital Pathology Association. His experience working both for manufacturers and in frontline lab management gives him a unique perspective into diagnostics and the technology that drives it.
Taking full advantage of his insights, we picked Byrne’s brains on the importance of digital imaging and communications in medicine (DICOM) in enhancing interoperability, affordability, and sustainability in digital pathology.
What is your professional background?
I’m a scientist by training, with higher qualifications in both precision medicine and business administration. I actually started my career as a salesperson, progressing to local product management and general management positions within the pharma and in vitro diagnostics sectors in Ireland and further afield.
During that period I became very interested in clinical laboratory management and, in particular, the interface between testing and treatment, and I took a role at Beaumont Hospital, Dublin, as a business manager for the Directorate of Laboratory Medicine.
My role was essentially to work at the interface of the clinical and scientific leadership teams and the hospital group’s C-suite teams. I ensured that the lab was adequately resourced to achieve the correct mix and volume of testing in accordance with quality, cost, revenue generation, and modernization targets.
Being equipped with both industry and in-lab experience allows me to see both sides of the coin and, I can tell you, the customer’s needs always come first for me. This experience and mindset led me to Agilent and, in particular, to the pathology division. Now, as a director of product marketing I have responsibility for two areas: pricing and analytics, and digital pathology. I've been in the role here for a little over three years.
How does DICOM compatibility enhance the integration of digital pathology data with other imaging modalities in a clinical diagnostics workflow?
It provides a standardized framework for managing and sharing images across modalities such as computed tomography, R-ray, magnetic resonance imaging, and now digital pathology too. Simply put, it allows images to be shared across different systems and devices, regardless of the manufacturer. Essentially, DICOM is a standard.
Here is a real-world example of DICOM in action. Imagine a pathology laboratory that has, over the years, installed whole slide imaging (WSI) scanners from several different manufacturers. Now imagine each scanner had DICOM compatibility – this would mean that they could all work together with an image management system (IMS). This interoperability enhances efficiency, accuracy, and comprehensive diagnostics, and also, in an ideal world, allows labs to utilize existing onsite technology.
Another key point is that DICOM also allows the integration of WSI scanners with other technologies – namely picture archiving and communication systems (PACS) and Vendor Neutral Archives (VNAs). This allows pathologists to share images with colleagues in other departments, such as radiology, to better facilitate multidisciplinary team meetings and tumor boards. This integration points to enhanced efficiencies, accuracy, and more comprehensive diagnostics.
What challenges do laboratories face when implementing DICOM standards in digital pathology, and how can these be addressed to improve interoperability?
One challenge is around product selection and procurement. Some instrument and software manufacturers may offer DICOM functionality as an add on, rather than an off-the-shelf offering. Some may claim that their technology is DICOM compatible, but interoperability may not yet be proven.
This is where organizations like the Digital Pathology Association are supporting the community. They organize “connectathons” to challenge the compatibility of different systems. Having said that I think we are moving towards a situation where DICOM becomes a sort of lowest common denominator for digital pathology systems – if suppliers don’t have it, they won’t be in the game.
Another obstacle is the preparation for the actual integration at the institution. There is little point proceeding with procurement until existing systems have been fully evaluated. This may include evaluation of the hospital information system itself, as well as the laboratory information system, to understand their capabilities and compatibility with DICOM. Whilst interoperability is the way forward, the institution may not be ready for it, since not all PACS or VNAs in use today can accommodate the file sizes and intricacies of whole slide imaging, but that’s another conversation.
With the increasing adoption of AI tools in pathology, how does DICOM compatibility facilitate the use of machine learning algorithms and decision support systems?
Digital pathology generates huge amounts of image data and metadata. DICOM's standardized approach simplifies the storage, retrieval, and management of these images, making it more efficient to ingest the datasets required for training machine learning (ML) models
Looking to the future, I feel that this standardization will increase user confidence and adoption of AI tools for digital pathology.
Another consideration here is the increasing importance of digital pathology within companion diagnostics. As an example, there are therapies under development that are using digital tools as part of the biomarker discovery process and beyond. I believe it is realistic to expect that we will soon see companion diagnostic assays used in conjunction with digital decision support algorithms in the clinic. I say this because some of these emerging therapies (anti body-drug conjugates as an example) may require very complex pathological interpretation. In these cases, the use case for DICOM standardized AI-driven image analysis is compelling.
Given the massive file sizes associated with whole slide imaging, what are the critical factors laboratories should consider when selecting data storage solutions?
This is a very common question. For me there are a few key factors. First, it depends on how many slides the lab processes per year and how many it plans to digitize now and into the future. Data storage is important for all, but it is a major concern for high-throughput digital labs, or those that process a lot of immunohistochemistry, owing to increased file sizes.
Second, we need to consider the local set-up. The institution may have an existing on-premise or cloud-based system available, particularly if they have a modern radiology unit, or a new system may need to be procured that complies with local storage protocols. Hybrid solutions, combining on-premise and cloud-based storage, can be effective in managing costs and ensuring compliance with local storage protocols. We are seeing more customers opting for this data storage solution.
My advice is to include laboratory IT departments early and continuously throughout your digital journey. I believe it is fair to say that digital pathology vendors are supporting these storage challenges with novel solutions but that this will remain a relevant concern for labs into the future.
What measures might need to be introduced to ensure affordability and sustainability of digital pathology?
There is a trend for WSI scanners to be released to the market with greater and greater slide capacity. They are an expensive outlay and annual software subscriptions can mount up too. As with all lab investments it is important to consider the return on investment. Some of these returns can be tangible. The Digital Pathology Association has published a return on investment calculator for its members . It considers the upfront costs of procurement, project costs, staffing, training, and so on, offset against savings generated by digitization and revenue gains from bringing tests in house etc. It can be useful in business case discussions. Some pathology staining vendors are reducing the upfront financial barriers to digital pathology by offering laboratories solutions that see outlays blended into staining reagent rental programs.
Reimbursement of digital pathology has been slow and is very region specific. Advancements in this area would be welcomed since it would both reduce the overall cost and also stimulate the broader adoption of these technologies, to benefit patients.
Emerging technologies such as virtual staining have the potential to increase sustainability. Essentially it uses algorithms to predict what staining would look like if it were to be undertaken on a tissue. It has multiple use cases, but a clear and obvious one that comes to mind is special staining. Here algorithms run on an H&E image could be used to predict special stain outcomes thus reducing reliance on the use of toxic chemicals, and ultimately reducing waste.
Newsletters
Receive the latest pathology news, personalities, education, and career development – weekly to your inbox.
