In 1854, Henry David Thoreau wrote, “If a man does not keep pace with his companions, perhaps it is because he hears a different drummer.” The metaphor of marching to the beat of a different drum is an apt one for the Sperling Prostate Center and Sperling Medical Group, especially when it comes to incorporating Artificial Intelligence (AI) into clinical practice. How do we know that our use of AI is the exception, not the rule?

A minority of radiologists embrace AI, according to survey

Our Centers are part of a relatively small subset of clinical radiologists who employ the tools of AI to improve what they can offer patients. We know this because the American College of Radiology (ACR) conducted its first survey “to understand how radiologists are using artificial intelligence (AI) in clinical practice and to provide a baseline for monitoring trends in AI use over time.”[i]

Before enumerating the survey results, a word about the ACR. It is nearly a century old, having begun in 1923, and now includes almost 40,000 radiologists, radiation oncologists, nuclear medicine physicians and medical physicists. Delving into clinical applications of atomic and subatomic advances demands the brilliant, dedicated minds that characterize its member researchers and practitioners. The brief survey, distributed by email to roughly 20,000 members, was designed to find out if and how respondents were currently employing AI in their practice, evaluate how well their specific algorithms were performing, and assess future needs.

I found the results surprising. Out of the majority of the 1,427 responses, only 493 indicated they were using AI. Of those, 366 completed a follow-up supplemental survey: which specific FDA-cleared models were most commonly used, and for which purposes; how AI could enhance their practice in the future.

What the data shows

According to the survey results, larger practices and academic centers were more likely to use AI, “although nearly 20% of practices of fewer than five members reported using AI.” For those utilizing AI, all subspecialty areas were represented, but breast imaging was most common, followed by chest/lung and neurology. Likewise, all imaging modalities were included with CT the most common. In fact, 52% reported using AI for image enhancement, though other uses included worklist management, departmental operations, etc.

Regarding future use and development:

All survey respondents were asked what they would like AI to do to enhance their clinical practices. Lesion detection (73%) and anatomic measurements (71%) were the most important; however, worklist prioritization, image quality improvement, and departmental process improvements were all chosen by greater than 50% of respondents.

As hopeful and visionary as that sounds, it seems the naysayers outnumber the proponents. According to an April 21, 2021 news article in Radiation Business:

Out of those not deploying the technology, 80% said they “see no benefit.” One-third of respondents said they cannot justify the expense, or that the purchasing decision was beyond their control. Some expressed concerns about decreased productivity. More than 70% of respondents hold no plans to pay for AI, while 20% see themselves doing so within the next five years.

Our mark of distinction

Unlike those who don’t plan to incorporate AI into their work anytime soon, the fact that our relatively small community practice has already embraced AI in our MRI detection and diagnostic services places us within the 20% (one-fifth) of our professional colleagues whose practices are distinguished by this forward-looking technology.

While we are marching to the beat of a different drum compared to the 70-80% of non-users, our pioneering incorporation of AI is a very positive mark of distinction.

We are proud to be in the vanguard of AI users whose patients are already benefitting from more efficient, accurate imaging results. We will continue to add AI tools as they become tested and available. I am fortunate to be personally connected with AI research and development.

I hope that ACR continues to conduct annual AI surveys, as I expect the pace of adoption to increase as AI continues to be developed and refined. Its advantages will become ever clearer within the field of radiology. The proof will be in the numbers.

NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.

References

[i] Allen B, Agarwal S, Coombs L, Dreyer K, Wald C. 2020 ACR Data Science Institute Artificial Intelligence Survey. J Am Coll Radiol. 2021 Apr 20:S1546-1440(21)00293-3.