Radpathology/Pathradiology: The Information Specialty of The Future?
Radiology and Pathology will likely merge into a hybrid medical specialty
Imagine that it's 2045, in the information specialty department of the hospital.
*The same physician detects a lung mass, biopsies it, and reviews the specimen to make the diagnosis of squamous cell carcinoma. Soon thereafter, she calls the patient with the results and dictates a unified imaging and pathology report.
*At tumor board the same day, another colleague is presenting both the CT and pathologic findings of a papillary renal cell carcinoma.
*Meanwhile, a machine learning algorithm is reading ICU films.The computerized helper allows the physician to spend more time on the more rewarding and consultative parts of work.
Welcome to the Imaging 5.0 uber-super physician who is both radiologist and pathologist.
The image centric disciplines of radiology and pathology (rad-path) are ripe for disruption. Analyzing images is a perfect fit for AI, potentially redefining the two disciplines. Artificial intelligence (AI) will change radiology and pathology, likely forcing them to consider merging.
One audacious and futuristic proposal to survive the AI robopocalypse is the merger of the two specialties. In a 2016 JAMA editorial, Drs Saurabh Jha (Associate Professor of Radiology at the University of Pennsylvania) and Eric Topol (Cardiologist and Director of the Scripps Institute) proposed a fusion of diagnostic radiology/molecular imaging and pathology/laboratory medicine into a unified "information specialist" discipline. Surprisingly, this was not a new idea. Richard Friedberg MD championed the idea as far back as 1997 when he created a unified Diagnostic Medicine Service (combining pathology, radiology, and nuclear medicine) within the Department of Veteran Affairs, known as the VA Southeast Network. In 2006,Bruce Friedman, a prominent pathologist, made a compelling argument for unification and presented "Ten reasons for merging pathology/lab medicine with radiology." Among the benefits he envisioned are "integrated reports of pathologists and radiologists working collaboratively" and "higher levels of quality." A former executive of General Electric made a similar argument in 2007.
Under the current paradigm, radiologists and pathologists work in a siloed environment, with no linkage between their reporting systems or workflows. Advocates of an integrated diagnostic service expect that unification will add value by increasing accuracy, speeding up diagnosis, and improving patient outcomes. There is also value in pairing high-definition advanced imaging from radiology with the latest high-tech tools of the pathologist.
An emerging strategic alliance could evolve over the short, medium, and longer term.
1.Short term: Update existing legacy information systems. Create an "integrated diagnostic server" controlled by pathology and radiology.This enhanced system would scoop up data and images for selected patients from the RIS,PACS,and LIS (laboratory information system). The resulting combined workflow would produce single holistic rad-path report that combines text, sentinel images, and histopathologic data, viewable on a web portal. The UCLA Radpath portal is an excellent model for such a unified report.
2. Medium term: Digitize pathology. As pathologists increasingly give up reviewing images on slides and microscopes and join the digital revolution, it will be easier to provide a truly integrated diagnostic worklflow. Identical PACS systems for radiology and pathology will allow viewing of both types of images and allow tighter integration between the two departments. For example, after a university hospital in Finland digitized its pathology department, the latter became part of an integrated Diagnostic Imaging Center, which included radiology and nuclear medicine
3. Longer term: Merge. Combine the two specialties into a single professional identity with revised residency/fellowship training. According to Drs. Jha and Topol, "information specialists should train in the traditional sciences of pathology and radiology. The training should take no longer than it presently takes because the trainee will not spend time mastering the pattern recognition required to become a competent radiologist or pathologist. Visual interpretation will be restricted to perceptual tasks that artificial intelligence cannot perform as well as humans.....Information specialists should be taught Bayesian logic, statistics, and data science and be aware of other sources of information such as genomics and biometrics....."
As noted in the same JAMA editorial, the history of automation in the broader economy typically results in a redefinition of roles. Jobs are not necessarily lost, but repurposed. The information specialist is likely to emerge as a repurposed physician, a reborn phoenix that has risen from the fiery ashes of the individual predecessor specialties.