Integration and Interoperability Are Essential for the Growth of Digital Pathology

There are multiple hurdles that digital pathology must overcome before the widespread adoption of this technology. One key hurdle is that consumers of digital pathology are rightly wary of staking a lot of money and time on the often frustrating process of installing and adjusting to expensive digital pathology systems. A large portion of this frustration is due to sub-par standards and practices with regards to digital pathology hardware and software implementation and integration.

There are two elements that make Implementation and integration particularly crucial to the success of digital pathology. First, digital pathology is just one (albeit crucial) part of the overall pathology and diagnostic workflow. Integration with existing laboratory hardware and software, including the Laboratory Information System (LIS), is vital to ensuring that the full benefits of digital pathology can be achieved. If that doesn’t happen, or if it doesn’t happen smoothly, consumers are justly concerned that the efforts spent on this expensive endeavor won’t pay off. Rather, they will be left with just another tool whose use consumes time and effort that pathology labs simply don’t have.

Secondly, digital pathology technology is evolving quickly. In theory, this rapid evolution provides the consumer with an improved experience, but poor implementation and integration can make it challenging for an existing setup to evolve alongside the technology. As new hardware, software, and accessory tools are developed as part of the pathology workflow, pathologists want to feel assured that their systems will never be outdated and that adopting and integrating new technologies will be straightforward.

Digital pathology is not meeting consumer expectations for integration and implementation due to issues concerning interoperability. Closed hardware and software systems leave little room for independent, third party integration. What’s more interesting and troubling is that, to date, even the most open hardware and software providers have yet to successfully establish standards or ensure compatibility between their products. Consumers want a seamless experience from start to finish. What they’re often getting is, at best, a disjointed connection between disparate products, and at worst a completely separate set of technologies and workflows.

There are several noticeable consequences of these poor integration and implementation processes. First, adopting digital pathology becomes increasingly expensive. Complex integrations require customization’s and additional labor during setup. They also run into problems more regularly as pathology labs grow and expand. Another notable consequence is that many of the promises of digital pathology are not being fulfilled. Pathology workflows are diverse, with each lab having its own set of needs and constraints, and the best way to address this is through a network of technologies. The lack of standardization in integration and implementation has made addressing each need difficult. Finally, the long-term consequence of poor integration and implementation is that consumers feel locked in to a particular vendor. With all the time and money spent on the initial implementation of a digital pathology system, the first system a lab buys is likely going to remain a fixture of the lab for a long time.

The industry, specifically hardware and software companies, must take action to correct this, not only for their consumers, but also for the patients who will greatly benefit from digital pathology technologies. Opening up a hardware or software technology for better interoperability is a technical process that need not compromise the integrity of a system. The development of a universal standard for data movement and management should be the primary goal of digital pathology companies moving forward. This would facilitate communication and integration between not only digital pathology systems, but also with the other elements of a pathology and diagnostic workflow.

So why is this not being done? There are several possible answers to this. One is that the recent explosion of digital pathology technology, including vast improvements in scanners and software, is still relatively young. Companies simply have not had the time to work together to ensure interoperability during implementation.

Another is the balance between the cost to the customer and the cost to the technology provider in accomplishing integrations and interoperability. The development of standards of integration will cost the companies money upfront that could otherwise be spent on growing and optimizing their product portfolio.

A final and more cynical possibility is that digital pathology companies have not developed or adopted integration and implementation standards as part of a sales strategy to ensure vendor lock-in and improve customer retention statistics. In the longer term, however, this results in frustration to the consumer and impacts both the likelihood of expanding a digital pathology workflow as well as the companies to which they will turn when looking for technology.

The digital pathology industry has its fate in its own hands. The decision to improve integration and implementation is one that is technological and strategic in nature, often with only minimal resources needed to make critical changes to interoperability. These changes can be narrow or vast, depending on the strategic goals of the companies. Ultimately, interoperability, integration, and implementation are crucial for the continued adoption of digital pathology.

 

In seeking to foster discourse on a wide array of ideas, the Digital Pathology Association believes that it is important to share a range of prominent industry viewpoints. This article does not necessarily express the viewpoints of the DPA; however, we view this as a valuable point with which to facilitate discussion.

5 comments on “Integration and Interoperability Are Essential for the Growth of Digital Pathology

  • Erio Barale-Thomas says:

    There are several car makers, but they all provide a motor, steer and 4 wheels; and they all use the same gasoline or diesel. I always wondered why the digital pathology industry didn’t take the same approach. I hope this lack of maturity (in their mind and in their tools) will be bypassed soon.

    I don’t agree with Nathan that the fate of digital pathology relied in the industry’s hands; it’s not exactly what he says, I recognize, but ultimately that is the outcome for us the users. We must be far stronger and far more demanding to our suppliers, in order to get what we need.

    I have also the experience that young actors are more reactive and have more useful tools than old elephants…

    Reply
    • Erio, I couldn’t agree more. The push for this has to come most strongly from the consumer. Ideally, the actions of industry follow customer demand. How industry chooses to responds to this clear demand will, in my opinion, heavily influence the rate of technology adoption on the one hand and the pace of technology evolution on the other. Ultimately, though, interoperability and integration will happen only when companies open their eyes to these benefits and to customer sentiment. Here’s to hoping that that day is sooner rather than later!

      Reply
  • Thanks for the thoughtful article. I agree with much of it. I think part of what might be missing is an industry blueprint or vision (perhaps with DPA leadership) that breaks DP into self-contained components e.g., LIS, scanner, workflow manager, workstation, etc. This is not dissimilar to what happened in Radiology with the IHE work. My impression is that another challenge is regulatory in that approvals are being provided for end-to-end systems vs. the components with clear requirements for interoperability. Do others see this is a barrier? It also seems to me that several vendors are also using proprietary image approaches – which no doubt are in part aimed at improving the overall user experience – but which also end up undermining true interoperability. Hopefully the DICOM 145 work will reduce this going forward. Should the path to interoperability start perhaps with an image standard?

    Thanks for getting the discussion started!

    Reply
    • You’re right – this is missing the blueprint or next-step-forward element of the conversation which is key to seeing the level of interoperability that digital pathology needs. I’d be all for pushing harder at developing and executing a plan like that. With regards to regulatory – Had a standard existed prior to regulatory approval, would end-to-end have been required or could a component-based process have been sufficient? Without a standard, can we ever expect a component-based approach? And what baseline standard can we set that that allows both for confidence in broader component-based approvals and use while enabling vendor-specific technology improvements? I’m not asking these as questions to which there is no answer – I believe that there is an answer to each of them. The challenge is to come to a consensus as an industry. An image standard would be a great start.

      Thanks Brent!

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

1 × two =