Artificial intelligence, or AI, is a term that is often associated with robots or computers that “learn” by having material introduced to them. Different people have different reactions to the idea of AI, especially with its potential use in medicine. AI can perform many tasks at the same time and in the same way producing results that will always be precise, but not necessarily accurate. This is relevant to fields like diagnostic radiology, where the importance of recognition of pathology on imaging is essential to providing a diagnosis. There is an art to identifying pathologies, variants and “normal” reads, so the thoroughness and attention to detail necessary to perform accurate reads cannot be emphasized enough.
After four years of college, possibly a year or two of graduate school, and four years of medical school, a student transitions to a resident physician. After another four or five years, a resident becomes a fellow and then an attending physician. This is quite the journey; so many factors have to be taken into account when choosing a medical specialty. So does the possibility of integration of AI into a field like Radiology influence a medical student’s decision when choosing a career? Of course it does.
One of the first questions you are constantly asked as a third-year medical student on rotations is: “What do you want to go into?” There are many different ways to approach that question, but I choose to answer honestly, but respectably, “diagnostic radiology.”
In most cases the response sounds something like: “Why would you want to pursue a field that’s actively working to replace you?” The answer, for me, is quite simple. I’m pursuing radiology because I am truly excited about the future and honestly believe any changes in technology can only benefit radiologists and their patients.
In medicine, the importance of accuracy can never be overlooked. Diagnostic radiologists across the world read millions of exams each year and have to systemically make decisions based on the images that they are presented. AI programs have the potential to help with pattern-recognition and identify potential pathologies and variants. This is could end up being more helpful than harmful because AI can assist in the triage of exams as they are entered into the medical record system. The assistance with triage could speed up productivity by prioritizing suspicious reads based on pattern recognition coding.
How special would it be to have an active role in advancing a technology that has the potential to save time and save lives? The opportunity to learn more and get involved with the next phase of radiology is a very attractive part of pursuing this specialty as a career choice. Although I am in the early stages of my training, I don’t believe that the advancement in technology has to come at the expense of jobs in radiology.
The likelihood of AI completely taking the place of diagnostic radiologists is debatable. While some models have shown promising results, the margin for error in medicine will always be small. Such a small margin that in some cases error could result in life or death. For example, an exam with a few pixels containing artifact (radiology terminology referring to things like jewelry on the patient or a shadow or part of the image that is over or under exposed) could be the difference between a new diagnosis and a patient positioning error.
Radiology was fully matched this year in the National Resident Matching Program, so I think it is safe to say that interest and excitement is definitely growing. This specialty is definitely thriving, with more and more applicants who are just as excited as I am about the opportunity to be at the forefront of the radiology paradigm shift. I do, however, think that the idea of major change makes a lot of people uncomfortable. The thought of losing any control, especially as it relates to your career, is enough to make anyone think twice when choosing a medical specialty. This is a unique time for radiology and I am excited for the future of medicine no matter how it looks, as long as patient outcomes and wellness remain the priority.
Lewis Jordan is a medical student.
Image credit: Shutterstock.com
