Among the oft-asked questions about AI is, “Will it put doctors out of business?”
And while the question is of interest to pretty much everyone, no one has a more immediately vested interest in it than doctors themselves, except perhaps medical students still training to become doctors.
Not surprisingly, then, AAMC News, a media outlet for the Association of American Medical Colleges, recently ran a long article that looked at the question in depth.
“An array of studies have offered glimpses of AI’s enormous potential,” writer noted early on. From algorithms out-performing radiologists in identifying myriad forms of cancer, to AI detecting rare hereditary diseases in children, to predicting the cognitive decline in Alzheimer’s patients, the triumph of AI over human counterparts has been documented far and wide.
But when will AI become an everyday tool for diagnosis? Sometime in the future, definitely, but not necessarily anytime really soon.
One challenge the writer describes is putting AI to its highest and best use. For example, AI could be particularly beneficial in places with limited access to health care.
According to Sameer Antani, PhD, Staff Scientist and Acting Chief for the National Library of Medicine’s Communications Engineering Branch and Computer Science Branch, “Machine intelligence presents us with an opportunity to significantly improve the delivery of health care, particularly in high-disease or low-resource settings.”
For example, a study published in March 2019 by the American Academy of Ophthalmology, found that a Google algorithm improved doctors’ ability to accurately diagnose diabetic retinopathy, and the algorithm has been tested in India, which is the type of country that could benefit from AI screenings, since it suffers from a shortage of doctors and ophthalmologists.
Still, obstacles remain, including the challenge regulatory agencies are encountering measuring and validating AI’s performance in medical diagnostics, given the rapid pace of change in AI technology, as well as the availability, quality, and completeness of training data on which AI can be tested.
“Factors that affect data strength include social, geographic, or economic biases, as well as simply acquiring data,” the writer explains, (and while) computer scientists are developing AI architectures that produce compelling results with less data,” experts say such advances down’t necessarily address the overall shortcomings.
But, in the end, the most personally pressing question for doctors and medical students concerns the impact of AI on their jobs. And, there again, the expectation is perhaps not so dire as some would expect.
“Radiologists are being trained to recognize AI’s shortcomings and capitalize on its strengths,” wrote Curtis P. Langlotz, MD, PhD, professor of radiology at the Stanford University Medical Center, in a May 2019 editorial for the journal Radiology. “An AI algorithm that diagnoses common chest conditions at the level of a subspecialty thoracic radiologist is a major step forward, an incredible asset to underserved regions, and could serve as a valued assistant for a subspecialty radiologist.”
To be sure, the jobs of specialists such as radiologists and pathologists are bound to change, but change is not the same as elimination.
Over the next 5 to 10 years, the most successful radiologists and pathologists will be those “who are well equipped and eager to participate in data management and integrated diagnoses,” said Frank J. Rybicki, MD, PhD, vice chair of operations and quality with the University of Cincinnati Department of Radiology and medical director of Imagia Cybernetics Inc. When an algorithm produces unexpected results, radiologists will need to understand why.
Many stakeholders believe this will lead to a deeper role for providers in patient care, which is one reason why they’re actually more excited than feeling threatened by the rise of AI in healthcare.