Pathology informaticist points out that not every slide in a patient’s case may need to be scanned and archived by a digital pathology system
Is digital pathology ready for prime time in the specialty of anatomic pathology? Many proponents of digital pathology would say “yes,” and the pathology laboratories now using digital pathology systems report significant benefits. But there are pathologists who argue that this is still a developing technology.
Like any new technology in clinical laboratory medicine and healthcare, digital pathology must demonstrate the right combination of cost-to-acquire, speed-in-use, and added-clinical-value, if it is to gain wide acceptance by pathologists. At the same time, “going digital” has transformed radiology, for example, and this example is often cited by boosters of digital pathology systems.
Certainly it is true that digital imaging has been a boon to healthcare, particularly in the field of Radiology. But would it be as beneficial a tool for Pathology as well? During a presentation given at the Society for Imaging Informatics in Medicine (SIIM) at least one presenter stated that such a belief in digital pathology might be misplaced—at least at this point in its development.
Radiologists primarily look to see if previously noted “issues,” such as an internal lesion or a broken bone, have healed or are no longer visible. Pathologists, on the other hand, look to see if the same cells that identified an abnormality, such as the presence of cancer, are “still present” in the tissue. These intrinsic differences in how the two medical specialties produce the images they use needs to be recognized, noted Paul J. Chang, M.D., FSIIM, Professor and Vice Chairman of Radiology Informatics, as well as, Medical Director of Enterprise Imaging at University of Chicago Hospitals. He believes these differences fundamentally alter how digital imaging should be used in pathology versus radiology.
“The most important difference is that radiology is digital at the beginning [of the workflow process]; pathology is digital at the end,” he said in a CMIO article. “This has devastating consequences when designing IT.”
Digital-Based Workflow versus Digital Imaging in Pathology and Radiology
Chang insists that the standard procedures applied to the acquisition and storage of radiological digital images cannot be applied to pathology without major alterations to the workflow.
“Digital pathology imaging is still very immature, especially with respect to workflow,” Chang said. “The good news is that there is a lot of overlap and significant opportunities for collaboration and convergence between pathology and radiology IT, but we need to avoid force-fitting…avoid the traps.”
Vendors want to apply to digital pathology the same digital imaging model that was so successful with radiology Picture Archiving and Communications Systems. However, according to Chang, pathology and radiology workflows differ in fundamental ways.
For example, Chang pointed out that when a radiologist interprets a study, he or she must have immediate access to all prior high-resolution images in order to determine any changes in the patient’s condition and render a diagnosis. To support this need of the radiologist, all images must be stored indefinitely as high-resolution (i.e., enormously large) files. Pathologists, however, do not use prior image studies of a specific patient in this same manner.
Another difference is that, whereas, radiology stores all of the images taken during multiple studies so they can be accessed and evaluated not just by radiologists, but also by oncologists, surgeons, and a plethora of other medical professionals at a later date, pathologist are most interested in reviewing slides that contain a blue dot. The blue dot was placed there by a pathologist during a previous viewing to indicate the presence of diseased cells. Chang’s argument is that slides without blue dots are effectively useless to pathologists, and therefore, should not be digitally stored.
Digitizing Pathology Slides Creates an Enormous Amount of Data
Chang pointed out that, on average, a single pathology slide digitizes to about 15 gigabytes (GB) each and a complete digital study can run into the terabytes (TB). For this reason, immense storage capacities are required to house even a few days of pathology studies, let alone an entire slide archive produced by the typical anatomic pathology laboratory in the course of a normal day’s volume of cases.
“It’s hard for me to justify—on a cost basis—a return-on-investment or total-cost-of-ownership argument for a [digital pathology] archive that’s ‘write-once–read-almost never,’ and that’s what we’re talking about,” said Chang in a video interview taken at this year’s Radiological Society of North America (RSNA) conference.
“A typical day’s work in a moderate sized pathology laboratory can actually generate more digital data than a whole radiology department in a week or a month,” said Chang. He went on to question whether anatomic pathology laboratories could “justify” the creation of digital pathology archives to permanently store slides that did not contain blue dots, when the pathologist’s “dominant use case” doesn’t require all those slides—just a representative image.
Chang believes that pathologists should “embrace a digital-based workflow to improve efficiency, safety and quality.” However, he also notes that it would be better from a cost-effective standpoint to create a digital workflow for pathology where only the single blue-dot slides were captured and stored, rather than the entire image study.