AI Will Supplement Practices, Not Replace Radiologists

As deep learning neural networks continue to become more sophisticated, radiology professionals are beginning to worry that artificial intelligence will rob them of their jobs. While deep learning technology may be able to undertake specific tasks normally performed by a trained radiologist, professionals in the field of radiological science shouldn’t expect the technology to replace them.

While research shows that AI technology can identify pathologies in medical images sometimes better than a radiologist, deep learning systems can’t duplicate the human aspects that the position requires.

Instead of perceiving AI as a danger to their careers, radiologists should view deep learning systems as an enhancement rather than a hindrance. The volume of patient medical images has increased much faster than the number of radiologists available to interpret them. Any system that can quickly assist in the reading and interpretation of multiple images can not only help radiologists, but improve relations between these doctors and their patients as well. Let’s explore some of the ways AI can assist radiologists, rather than replace them:

AI Hasn’t Been Developed Enough For The Clinical Setting

In the field of radiology, deep learning is still restricted to the research stages. According to a 2019 report in Acta Radiologica Open, there is no auxiliary method of using artificial intelligence in radiology practice in Western countries as of yet. However, the same report further states that AI-dependent medical imaging in daily practice may not be far off into the future.

Dr. Emre Pakdemirli, a radiology consultant, believes that the increase of applications for U.S. Food and Drug Administration (FDA) sanctioned AI-driven imaging equipment may be proof that the technology is making a transition into the clinical setting. While this technological advancement is on the horizon, radiologists should not view the AI as a threat to their jobs, but rather as an ally that will lighten their workload. Acting as a provisional tool for imaging and diagnosis, the AI will serve more as a dependable friend to radiologists rather than as their replacement.

Image Interpretation Is One Part Of A Radiologist’s Job

As technology continues to become more advanced, one task AI may eventually usurp from radiologists is the role of image interpretation. As it currently stands however, AI radiology systems are too limited to undertake the job.

Deep learning radiology systems are created to perform single image recognition tasks, such as detecting a brain hemorrhage on an MRI. While this does assist radiology practices, thousands of single detection tasks are required to identify all possible findings in images. To date, AI are not capable of handling more than a few single detection tasks at a time.

If AI ever becomes advanced enough to take over the task of image interpretation, one fact that can comfort radiologists is that it is only one portion of their job. There is more to radiology than image interpretation and radiologists could instead channel their expertise into other aspects of their job, such as:

• Treating patients

• Consulting with other doctors on diagnosis and treatment

• Performing image-guided medical interventions

• Relating findings from images to other medical records and test results

• Discussing recommendations, procedures, and results with patients

Machines Can’t Replicate Necessary Human Characteristics

As deep learning solutions evolve into the clinical arena, trained professionals will still play an integral role in radiological science. While machines handle the more technical aspects of the job, people will always be in charge of the tasks that involve human interactions, something a machine can’t replicate.

Since a vital part of a radiologist’s job includes communicating with patients and discussing possible actions and recommendations for treatment, machines can’t be expected to build the human connections necessary for patient-centered care. According to a 2018 report by European Radiology Experimental, “So far, AI is neither astute nor empathic. Thus, physicians (i.e., we say here radiologists) remain essential for medical practice, because ingenuity in medicine requires unique human characteristics.”

Ironically, the implementation of AI systems in radiology actually highlights and enhances a radiologists’ value in their field; with AI systems handling the workload of image interpretation, connections between patients and radiologists will improve now that the doctors have more free time for their patients.

AI May Not Be As Impactful As Predicted

If radiologists are anxious about the encroaching technological age of AI, then they should pause to reflect on their industry’s history. In the field of radiology, history tells us that technological innovations don’t always turn out to be as impactful as originally predicted.

Around the 1990s, a computer-aided detection network was used widespread in the field of radiology, but research indicates that the machines didn’t have much of an effect on radiologist accuracy.

According to Curtis P. Langlotz, a professor of radiology at Stanford University, “The recent rush of novel AI algorithms should prompt introspection about past failures of AI to live up to its promise. Today’s AI tools have achieved regulatory clearance based on their performance at a small number of health care organizations. Perhaps the incremental accuracy of these new AI methods will reduce false-positive findings and blunt the ‘cry wolf’ effect, but the generalizability of these algorithms to the diversity of radiology practices remains an open question.”

While AI systems are expected to supplement the field of radiology, it doesn’t look like machines will be leading the industry any time soon.

References:

1. Pakdemirli, E. Artificial intelligence in radiology: friend or foe? Where are we now and where are we heading? Acta Radiologica Open. 2019. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6385326/ Web. May 22, 2019.

2. What is a radiologist? American College of Radiology. https://www.acr.org/Practice-Management-Quality-Informatics/Practice-Toolkit/Patient-Resources/About-Radiology Web. May 22, 2019.

3. Pesapane, F., et. al. Artificial intelligence in medical imaging: threat or opportunity? Radiologists again at the forefront of innovation in medicine. European Radiology Experimental. 2018. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6199205/ Web. May 22, 2019.

4. Walter, M., 4 key reasons AI won’t replace radiologists. https://www.radiologybusiness.com/topics/artificial-intelligence/4-key-reasons-ai-wont-replace-radiologists Web. May 22, 2019

[Originally produced for a leading healthcare company associated with Marketing Insider Group in May 2019]