Will Digital Pathology Finally Join Radiology in the New Digital Healthcare?
The publication of a peer-reviewed article in the American Journal of Surgical Pathology has opened the door to the widespread use of digital pathology for diagnosis, pointing to a new era for that important specialty — and the broad inclusion of digital pathology images into enterprise-wide digital archives
Even as diagnostic imaging procedures in the U.S. have shifted from being analog film-based to being entirely digitally based-and therefore capable of being stored and shared, the specialty of clinical pathology has lagged behind its sister specialty of radiology in moving fully into the digitally facilitated new healthcare.
There are a number of reasons for that lag, but one has been particularly challenging: until now, pathologists’ diagnosis using digital images has been considered inferior to diagnosis based on the use of glass slides. That factor has inhibited the widespread conversion of glass-slide images to digital ones, and their storage in digital archives and their sharing among clinicians.
But now, the newly published, peer-reviewed article in the American Journal of Surgical Pathology is set to change that dynamic. The article, “Whole Slide Imaging Versus Microscopy for Primary Diagnosis in Surgical Pathology: A Multicenter Blinded Randomized Noninferiority Study of 1992 Cases (Pivotal Study),” appeared online on Sept. 28 in that journal.
As the article’s authors write, “Whole slide imaging (WSI), also known as digital pathology or virtual pathology, is a technology that involves high-speed, high-resolution digital acquisition of images representing entire stained tissue sections from glass slides in a format that allows them to be viewed by a pathologist on a computer monitor, where the image can be magnified and navigated spatially in much the same way as standard microscopy [that is to say, pathologists relying on traditional glass slides]. These images can be utilized for diagnosis by pathologists, creating a digital workflow that obviates the use of conventional bright field light microscopy (henceforth referred to as “microscopy”). Before substituting the time-honored, familiar and versatile microscope with WSI,” the authors write, “several valid concerns need to be addressed. The most critical one is whether pathologic diagnoses rendered using WSI are comparable with (i.e., noninferior to) diagnoses made by microscopy.”
Essentially, what these researchers determined, through a very rigorous clinical study, was that the quality of digital pathology images is high enough to provide reliable diagnosis using digital images only.
As the authors write at the conclusion of their article, “In summary, this study demonstrates that WSI is noninferior to microscopy for the purpose of making a primary diagnosis in surgical pathology. This conclusion applies across a wide range of organ systems, sampling methods, specimen types, stains, and practice settings. Our findings have the potential to significantly alter the workflow of surgical pathologists in coming years and pave the way for a purely digital workflow analogous to the process currently used by radiologists.”
The study received financial support from the Netherlands-based Philips, whose IntelliSite digital pathology solution has received FDA (Food and Drug Administration) clearance, based on the evidence confirmed in the study.
One of the article’s authors, and its principal investigator, was Clive Taylor, M.D., Ph.D., a professor of pathology in, and the former chair of, the Department of Pathology in the Keck School of Medicine at the University of Southern California, in Los Angeles. Dr. Taylor spoke with Healthcare Informatics Editor-in-Chief Mark Hagland regarding the study and its implications for the advancement of the use of digital pathology — and for the increasing storage of digital pathology images in enterprise-wide digital archives. Below are excerpts from their interview.
What was the timeframe for this extensive study?
We began the study in a serious fashion in 2013, though there had been some discussions prior to that. The actual formal study took 16 to 18 months and involved nearly 2,000 cases, with each participating pathologist viewing at least 500 sets of slides. The article was published in the August/September issue of the AJSP online.
What did you essentially find?
Surgical pathology functions on the basis of pathologists’ being trained in pattern recognition using glass slides for recognition of disease entities; and cancer is still diagnosed in that way, even in the molecular age. And that expertise, that pattern-recognition expertise, is based on the collective experience of many pathologists over many years. And in comparing digital images with glass slides, the concern was that it might work well in one organ system that was easy, but not at all in another organ system that was more difficult. So the intention was to do this study across all organ systems. And that was the difference between this study and other previous ones, which were usually focused on one or two organ systems. And the FDA wanted to see this across all.
So essentially, you and your fellow researchers were able to affirm the validity of diagnosis using digital pathology images, correct?
Yes, we found that it was clinically sound. And to get 300 prostate cases, we had to go to numerous organizations. That’s why the FDA wanted this done across both academic medical centers and small hospitals.
How do you see digital pathology moving forward as a result of this?
That’s a really good question. Radiology went digital with a bang, as it were, around the turn of the millennium, and you went from silver deposits on films, to digital, and it all happened incredibly quickly. And when that happened, some of us pathologists said, well, it will happen for us as well, eventually. But we had an intervening step, that you couldn’t just take a digital image to begin with, you had to transform a glass slide image into a digital image. And that’s what this study did, it validated that you can diagnose from a digital image.
Given that, how quickly will this take off?
I think in major centers that this will take off quite quickly. There’s a cost/economic issue that wasn’t an issue in radiology. Radiology encountered significant economic savings by shifting from analog images to digital ones. Pathology doesn’t have that savings; we still have to make the glass slide. So any economic benefits will have to come from increased efficiency in handling and sharing the images and in leveraging artificial intelligence in diagnosis. So that will take some time. And all of those things are already going on.
Pathologists are under time pressure, as are radiologists, correct?
Yes, because of a national and international shortage of pathologists. And that is exacerbated by the fact that it takes five years of practice to reach competency, and 10 to 15 years to achieve expertise.
As we connect clinicians and patient care organizations across healthcare, the shift towards digital pathology will prove to be a boon, correct?
Oh, I believe that this will be a huge boon for pathology. Pathology has been a very isolated, parochial specialty, one somewhat disconnected from other processes. So I think that by putting digital pathology images into the medical record through links, and made available as appropriate, that could not only make images available to clinicians, but also to other pathologists, as needed. And all that time savings will be a huge benefit.
Is there anything you’d like to add?
I’ve been involved in digital pathology in one way or another for about 20 years. And among the challenges have been the time it took to scan images, and the storing of the huge files. But now, you can put them into the cloud. And the piece that makes this possible is digital analysis, or providing clinical decision support. It’s like having an intelligent microscope, really. Within a few years’ time, digital pathology images will come to pathologists already pre-analyzed by a digital clinical assistant, which will have highlighted the areas that are abnormal. And in terms of how that clinical decision support might evolve forward, look at what’s happened to the iPhone: when it started, it had five apps, and now, it has two million. And we’ll end up with 100 or so in pathology.
Copyright Vendome Group LLC Nov/Dec 2017. This article was written by Mark Hagland from Healthcare Informatics and was legally licensed through the NewsCred publisher network. Please direct all licensing questions to firstname.lastname@example.org.