Is Your Lab Ready for an AI Digital Pathology Workflow?
January 3, 2020Efficiency remains a vital healthcare goal, and in the field of pathology, the future will be digital. While cost has been a roadblock for some labs looking to digitize, recent metrics are beginning to reveal the benefits: Error reduction and consistent outcomes, faster turnaround times, slide sharing, and storage benefits are just a few of the advantages. The advancement of integrated artificial intelligence into digital pathology software is rapidly making these benefits all the more apparent. From a competitive position, ignoring the value of DP in your lab is not recommended. Laboratories must embrace the technology as a cost of doing business and seamlessly integrate it into their LIS and associated artificial intelligence applications — or risk being left behind with a shrinking bottom-line.
Why Labs Should Consider Integrating Digital Pathology Into Their Workflow
According to Dan Angress of Angress Consulting, LLC, who has 30 years of laboratory diagnostics experience, the potential business advantage of digital pathology is going to come mainly from efficiencies.
“You can imagine the difference in how laboratories can utilize remote locations that don’t have any of the instrumentation,” he said. “For example, if a west coast lab hires an east coast pathologist who can look at slides via a monitor, the lab can extend working hours and add tremendous productivity to the laboratory simply by delegating work remotely. With workflow tools specifically designed for digital pathology and integrated into existing Laboratory Information Systems, the workflow can be managed seamlessly.”
Joe Nollar, XiFin AVP, LIS Product Development, said with the quickly evolving use of artificial intelligence tools, the ability to triage cases into high or low priority is steadily increasing. For example, Proscia’s DermAI™ sorts and classifies cases to ensure that high-impact cases are always reviewed first.*
“In preparation for the full adoption of digital pathology for primary diagnosis, labs must consider how to integrate AI into the lab workflow and overall IT infrastructure,” he said. “Successfully integrating a digital pathology system into an LIS lets the lab take advantage of improved case throughput, case triaging via artificial intelligence, integrated image analysis, streamlined slide logistics, integrated reporting (with images and analysis), digital secondary consultations and precision medicine collaboration, and pathologist workplace ergonomics.”
Also, Nollar said that pathologists could use these innovative tools to prioritize cases that have been pre-selected as high risk. While every case is important to review, this technology represents a significant opportunity to use technology to reduce a lab’s risk of losing or misdiagnosing a critical case and is a significant process improvement to the way tumor boards are managed today.
Angress said one of the problems is that many labs are sitting on instruments that are rarely used for anything workflow oriented.
“They’re probably using them to scan some slides for a consult or for research projects,” he said. “But if you’re not looking at digital pathology as an option for lab workflow, then you’re going to get left behind.”
Pointing to the future of digital pathology, Angress said that two large scanning companies, Philips and Leica, have FDA approval for primary diagnosis. Also, there are artificial intelligence tools available that labs can implement into the workflow environment to add even more value to the use of digital pathology.
And for labs concerned with the cost of the technology, Angress shows in a published case report, “Use of Digital Pathology to Drive Revenue to Labs,” that digital pathology can be used in a TC/PC environment to grow business outside of a lab’s community, providing labs additional revenue from remote clients. In addition, he discusses the push toward expanded CPT codes for billing.
Integration: AI, Digital Pathology, and the LIS
The integration of DP into the lab, and specifically into the LIS, is a critical step to ensure seamless workflow and improved efficiency. AI tools focused on workflow, as well as diagnostic accuracy, are making this easier.
“Using innovative AI tools, pathologists can prioritize cases that have been pre-selected as high– risk,” Nollar said. “While every case is important, this technology represents a significant opportunity to use technology to reduce the lab’s risk of losing or misdiagnosing a critical case. If its Laboratory Information System is not enabled to support the digital pathology workflow, the lab will be limited in its ability to take advantage of numerous opportunities. The lab of the future must be prepared with a Digital Pathology LIS (DPLIS).”
“The LIS can be fully integrated with histology automation, such as the Leica Cerebro or the Ventana Vantage platform; be able to pass case information to the digital pathology system, and prepare barcoded slide labels that can be read by the imaging device,” said Nollar. “Once slides are scanned and the whole slide images are ready for review, pathologists should be notified of their availability via the LIS.”
The ideal digital workflow includes access to all case-related whole slide images via an image viewer integrated with the LIS. Pathologists can then capture regions of interest and include them in the LIS report if desired. If pathologists are utilizing automated image analysis, they should be able to seamlessly incorporate those results in the LIS and the final report.
“A future-proof LIS must be capable of managing more than just the digital pathology workflow,” said Nollar. “In some cases, additional testing may be necessary. Pathologists should be able to order test add-ons in the LIS and monitor on a test-add-on worklist as slides are prepared and digitized. Occasionally, a secondary consultation is necessary. An LIS that allows the ability to digitally share cases with colleagues is essential.
“Additionally, many forward-looking labs understand that their lab’s data is valuable to precision medicine programs. An LIS that facilitates the test results data being pushed from the LIS to a precision medicine collaboration or informatics platform, will be paramount.”
Getting Started in Digital Pathology
Angress recommends three steps to prepare your lab for digital pathology:
- Establish a business case for your laboratory
- Talk to your LIS vendor about whether it can integrate digital pathology into your system
- Consult several hardware scanning vendors to establish what instruments can meet your lab’s throughput, budget, and digital pathology objectives, short and long term
Interested in learning about XiFin LIS for integrated digital pathology workflow?
About The Experts
Joe Nollar, AVP, LIS Product Development, leads the development of XiFin’s LIS solution. He has nearly 20 years of experience designing laboratory systems. Nollar served as CIO of PathCentral, directing all commercial LIS software development. In 2001, Nollar founded eTeleNext, a full-service provider of web-based LIS solutions, until acquired by PathCentral. He has a bachelor’s degree in Political Science from UCLA and a Masters in International Management, Economics, and Politics from the University of St. Andrews in Scotland.
Dan Angress of Angress Consulting LLC has over 30 years of experience in laboratory diagnostics in various capacities, from marketing and sales to operations and business development. He has built and sold various laboratory LIS software products used across the country and is well versed in the use of digital pathology and artificial intelligence in the lab space. He has in-depth knowledge of healthcare practices and procedures, laws, regulations, and guidelines about healthcare. His capabilities extend across the spectrum of healthcare, including laboratory, pathology, software LIS, digital pathology, IT, medical devices, personalized medicine, pharma, and radiology. Dan currently consults with Proscia on the application of AI and digital pathology.
*Proscia’s DermAI is for research use only.