Diagnostic Imaging brings you daily coverage of RSNA 2021 live from Chicago. 

RSNA 2021 - Live Coverage From Chicago

The prompt and accurate results of an imaging study can be the difference between life and death. Yet the number of practicing radiologists in the United States is likely insufficient to meet the growing demand. This shortage is also projected to worsen, and according to the Association of American Medical Colleges (AAMC), the U.S. is projected to have a shortage of 17,000 to 42,000 radiologists, pathologists, and psychiatrists by 2033.

 On average, the number of imaging studies increases by a rate of up to five percent per year, but the number of radiology residency positions only increases by two percent.

This shortage has a dramatic impact on patient care. A shortage of radiologists leads to increased turnaround times for study results to be generated, which can negatively impact patient outcomes. One recent study found that radiology imaging delays are an independent predictor of length of hospital stay.

 Another recent study found it took more than three hours to generate a report from an average CT study with the radiologist being the bottleneck 42 percent of the time.

 This delay in clinically pertinent information can have a significant impact on patients. These delays needlessly burden health systems, patients, and insurers, and could be resolved by improved imaging coverage. Although the situation might appear dire, there are many potential solutions.

The first and most immediate solution would be an increase in the number of radiology residency positions.

 The number of positions is largely determined by Medicare, and there have only been marginal increases over the years in radiology. In 2006, there were 1,011 radiology residency positions. In 2020, there were 1,113 radiology residency positions. However, this slow increase has begun to change as Congress passed a graduate medical education (GME) expansion in 2020, which added 1,000 new residency spots, many of which are likely to be radiology positions.

Increasing medical student awareness of the specialty is another promising way to combat the shortage.

 Medical students often have little exposure to the radiology specialty, and a recent study found that only 35 percent had completed a radiology rotation by graduation.

 This could be remedied by promoting radiology as a core rotation in medical school as well as increasing the amount of radiology material that is covered during the preclinical curriculum. One study found that adding just 2.5 hours of radiology lectures to the first year of medical school significantly increased student interest in radiology, and increased the likelihood of students completing a radiology rotation later in the curriculum.

 This is a promising solution since more medical students interested in and matching into radiology could potentially reduce the impact of the radiologist shortage.

Another solution to this workforce challenge is achieving improvements in radiologist efficiency.

 Many studies have investigated potential methods by which radiologists can increase efficiency. Promising solutions include decreasing distraction and integrating the utilization of artificial intelligence (AI) in practice.

 Interruptions due to phone calls, emails, and consent forms can hinder a radiologist’s workflow, potentially leading to decreased performance and less informed consultations.

Reductions of these distractions through the use of reading room coordinators and enhanced communication workflows are often pursued as solutions for improving efficiency. Additionally, AI enables radiologists to perform their roles with greater efficiency. Although unlikely to replace the role of the radiologist, AI can be used to help with measuring lesions, comparing studies, and triaging scans that are at an increased risk of being urgent. One study found that AI could be used to select which chest X-rays were more likely to be abnormal, and then prioritized those for interpretation by radiologists.

 This method was able to decrease the average delay in interpretation of abnormal scans from 11.2 days to 2.7 days.

Although current trends in the number of practicing radiologists may seem concerning, the future is bright. There are many effective solutions to this problem, all of which are simultaneously being pursued. Many medical schools are integrating more radiology material into the curriculum, radiology residency positions are being expanded, and radiologists are becoming more efficient each year. These trends are promising for the future of the field and with sufficient action, the shortage can be addressed and possibly reversed.

1. Dall T, Reynolds D, Chakabarti R, et al. The complexities of physician supply and demand: projections from 2018 to 2033. Washington, D.C. Associaion of American Medical Colleges. Available at: 

https://www.aamc.org/system/files/2020-06/stratcomm-aamc-physician-workforce-projections-june-2020.pdf

 . Published June 2020. Accessed January 20, 2022.

2. Smith-Bindman R, Kwan ML, Marlow EC, et al. Trends in use of medical imaging in US healthcare systems and in Ontario, Canada, 2000-2016. 

3. Courane S, Conway R, Creagh D, et al. Radiology imaging delays as independent predictors of length of hospital stays for emergency medical admissions. 

4. Perotte R, Lewin G, Tambe U, et al. Improving emergency department flow: reducing turnaround time for emergent CT scans. 

 2018 Dec 5;2018:897-906. eCollection 2018.

5. Prezzia C, Vorona G, Greenspan R, et al. Fourth-year medical student opinions and basic knowledge regarding the field of radiology. 

6. Branstetter B, Humphrey A, Schumann J. The long-term impact of preclinical education on medical students’ opinions about radiology. 

7. Dargan R. Artificial intelligence boost efficiency and quality in radiology practice. Radiological Society of North America (RSNA). Available at: 

https://www.rsna.org/news/2019/march/artificial-intelligence-boosts-efficiency

 . Published March 18, 2019. Accessed January 20, 2022.

8. Lee MH, Schemmel AJ, Pooler BD. Radiology workflow dynamics: how workflow patterns impact radiologist perceptions of workplace satisfaction. 

Articles

While current projections suggest a shortage of radiologists in the future, promising developments with radiology residency positions, medical student awareness of the field and AI-aided efficiencies could be game changers. 

Are We Prepared for a Looming Radiologist Shortage?

Let’s get into some of the details of the paper. If we look first at the patient characteristics on this study. There were 58 patients with biochemical recurrence of prostate cancer after a definitive primary therapy. What’s that? That’s either surgery or radiation therapy. And both scans were performed within a mean interval of 9.4 days. There wasn’t a great stretch where some new development, some new metastasis, was probably going to happen. A very short interval between the 2 scans. All scans were visualized and analyzed independently on a patient, on a region, and on a lesion-based analysis. All exams were performed in the same medical department using identical scanners. It wasn’t like this place had this kind of scanner [and another place] had a different kind of scanner. The scanners are the same, and the people looking at it are the same. What changed is the 2 different modalities. Louis, do you want to walk through the patient characteristic slide a little?

One thing that you first have to look at when you look at this paper is that we weren’t dealing with a population that was stratified to a very low PSA [prostate-specific antigen]. This was a wide range of PSA, both at diagnosis and at the time of scan importantly. You had PSAs that varied from a very low 0.2 ng/mL to 230.4 ng/mL. That’s incredibly important from a perspective of understanding the data. We’re not localizing to a very low PSA.

The other thing that’s interesting is that we have a range of Gleason scores. Among those included were 4 patients who at the time of diagnosis had Gleason 3 disease, which is interesting that they went on to definitive therapy. That’s something that wasn’t clarified in the paper that I’d like to learn a little more about. Otherwise, we have patients who have a Gleason score anywhere from 7 to 10, so presumptively had clinically significant prostate carcinoma.

We had no patients who had metastatic disease. We had patients who at the time of diagnosis had nodal disease, but none who had disease within the bones. That’s important. The other thing that’s important is that, as you mentioned, there’s a difference in how these were treated. The majority of patients were treated with radical prostatectomy and a smaller number of patients were treated with radiation therapy.

That’s excellent. When I look at some of the patient characteristics, I echo your comments. When you look, typically people are going to present in the recurrent setting or with more advanced disease. There are very few patients who by their Gleason score would be thought of as low risk: classic D’Amico low risk or NCCN [National Comprehensive Cancer Network] low or very low risk. Most of the patients would be felt to be favorable, unfavorable, intermediate risk, or higher. However, some of the nuances you alluded to are still not known. When we look, this group of patients represents that 72% of the patients underwent radical prostatectomy, whereas only 28% underwent radiation therapy.

We know that from the American Cancer Society data that with older men, the overwhelming choice of therapy is radiation therapy at a 3:1 ratio. This doesn’t represent exactly what we’re seeing from a patient characteristics perspective with older men or people with recurrence. This is unique to their experience in Europe.

Let’s go on and talk about localization of positive lesions. Louis, do you want to kick us off about what we’re talking about and how they broke up where a recurrence might be?

They decided to determine where they were based on the 4 regional areas that you would think of when you think about recurrence. Of course, you’d think about disease locally in the prostate or prostate bed. Of course, it’s a prostatectomy, so we’re thinking about the prostate bed. If it’s local pelvic disease, so not nodes outside the pelvis. If it’s extrapelvic disease, extending in the other known target regions for disease, the retroperitoneum. And then of course the bone. Not viscera but the bone. In terms of positivity, in terms of localization, what they identified was that a greater number of patients were identified with 18F-fluciclovine at the bed compared with gallium PSMA [68Ga-PSMA-11]. But when we looked at other sites of disease, the pelvis itself, extrapelvic lymph nodes, and bone metastatic disease, gallium-68 demonstrated a higher percentage of positivity. That being said, when we look at the 

 value would be at the local bed, where 18F-fluciclovine in the context of this study did better at the local bed than PSMA [prostate-specific membrane antigen]. The numbers otherwise were close. They weren’t necessarily significant but did favor 68Ga-PSMA-11.

Perfect. I echo your comments about the previous slide. When we look at the number of patients who had radiation therapy, it’s only 1 of 4. When you’re looking at a local recurrence in the prostate, most patients don’t have prostates to recur. It does affect how we’re looking at the data set a little. Irrespective of that, when we look at local recurrence—i.e., in the postprostatectomy area, which represents over 75% of patients—you’re looking that there was more detection using Axumin [18F-fluciclovine] than there was for the PSMA scans. Why is that? This gets into some of the nuances of reading the scans. Louis, do you want to talk a little about when you’re looking locally in the bed vs looking for distant disease? What are some of the nuances—based on what happens in technique—that can throw off your ability to read?

That’s the crux of the argument in this paper. Let’s talk little about physiology: 18F-fluciclovine vs PSMA. If you have that image up, it helps to illustrate it a little. First and foremost, we talk about urine excretion. When we talk about PSMA agents, they’re genitourinary excreted. Therefore the amount of radiotracer that you’re going to find in the kidneys and the bladder is going to be greater than you will with 18F-fluciclovine. As an aside, I want to comment about uptake in other places. As I mentioned, there’s going to be more background marrow uptake with 18F-fluciclovine compared with PSMA agents. There’s going to be more uptake in muscle with 18F-fluciclovine than there’s going to be with PSMA agents.

What’s crucial is technique. When we look at the technique, and we’ll talk a little about this in the papers, how did [Birgit] Pernthaler do this, in terms of gallium-68 vs 18F-fluciclovine? Typically, when we do an 18F-fluciclovine study, we acquire those images approximately 4 to 5 minutes status post-injection of radiotracer. We’ll begin an acquisition from the thighs and head to the vertex. That’s your standard 18F-fluciclovine acquisition. What’s interesting is that there’s a debate about how to acquire PSMA or 68Ga-PSMA-11, or 18F-fluciclovine. How do you do that? In the Pernthaler article, they did dynamic images initially through the pelvis, which were about 5 or 6 minutes. Then they went back at 60 minutes and did an acquisition from the vertex to the thighs. They didn’t start from the thighs to the vertex, and I say that only because over the acquisition time you’re going to be accruing more radiotracer within the bladder. Many sites, and this is just anecdotal, will start from below and then head above.

The other thing that wasn’t done in the Pernthaler article, which some places do, was acquiring delayed images through the pelvis with high resolution. In other words they’ll acquire the images over a longer bed time. Interestingly, on the 18F-fluciclovine side, the number of minutes of bed acquisition for the prostate was 5 minutes for 2 beds within the pelvis. For the 68Ga-PSMA-11 agent, it was 2.5 minutes through the pelvis, so there was a difference in acquisition. Those things are crucially important.

Just as an aside, when you interpret the images, 1 thing that’s crucially important—and it’s pointed out on this slide—is that this is an example of a patient who had a recurrence that was identified with 18F-fluciclovine but not with PSMA. Why was that the case? In this case, we can see that here’s the standard grayscale imaging. This is the low-dose noncontrast CT scan. Here’s the PET [positron emission tomography] without the fusion, and of course we have the fusion images. What’s happened is that the bladder is obscuring the uptake within the region of concern or disease recurrence within the prostate. This is the crux of where a difference in local recurrence may occur. I’ll make some more comments, but what are your thoughts when you look at these images? [How are you] feeling about the way the images look?

The hardest area for us is around the bladder. They asked Willie Sutton once, “Why do you rob banks?” He said, “That’s where the money is.” Where the prostate was is probably the area with highest risk of recurrence. That’s why postprostatectomy radiation therapy has been so successful. When the RTOG [Radiation Therapy Oncology Group] put together a consensus about what to aim at, we created what’s called an RTOG atlas consensus statement—here’s where we need to aim. That was the area where the prostate was and into the bladder a little. That would represent what we’re seeing here: 1 of 6 patients with a positive local recurrence on Axumin [18F-fluciclovine] that wasn’t demonstrated on a 68Ga-PSMA-11 scan. Why? Because the bladder uptake obscures your ability to detect it.

In the old days, before we had this fancy next-generation imaging, we’d cover where we thought the prostate cancer occurrences could be based on historical data and historical information. We were somewhere between 40% and 70% of the time, but it would be much better to be able to target areas of specific recurrence. Then we were doing Axumin [18F-fluciclovine] scans, able to find where these sites of recurrence were. Some of them in the areas of the seminal vesicles remnants, some near the area of where the anastomosis was, but in general we were finally able to see where these areas of occurrences were and how that could potentially lead to improvements in outcome. In fact, EMPIRE-1 was a trial that looked to see if imaging could be a relevant biomarker to drive care and show that an improvement in care using next-generation imaging to guide radiation therapy in those patients. It’s obviously very exciting. PSMA unfortunately has its issues with being obscured by the bladder uptake. From my standpoint as a radiation oncologist, if it’s obscured and the radiologist or the nuclear medicine physician doesn’t read that there’s disease there, that could cause the patient to not have the best outcome with respect to getting the right next therapy.

As we go into the next slide, we’re going to see something that’s remarkable for PSMA imaging agents, which is the ability to detect bone metastases. Louis, talk me through a little about how you think the 18F-fluciclovine and PSMA agents have done from your experience and from this trial?

One thing you know when you look at the fused images and when you look at the PET-only images is the uptake that’s a baseline uptake within bone. When we look at this vertebra and the uptake in the bone, we see that there’s clearly some uptake within the marrow; therefore, that could easily obscure the identification of a small or not particularly avid metastatic focus. Whereas when we look at the PSMA study, we see a large difference in terms of the amount of uptake within bone. Clearly, when the background is very low, the ability to identify a lesion is going to be easier and become more overt.

One comment that you can make about 18F-fluciclovine before we had PSMA was that we did note that lesions that were particularly sclerotic could be very difficult to see and distinguish as being neoplastic. Anecdotally, in my experience with PSMA, the background of uptake in the marrow is so low. Perhaps 1 of many reasons why PSMA outperforms 18F-fluciclovine within the skeletal system is the absence of background uptake. That will make a huge difference.

There are other things that are at play, and we’ll talk about those. Sometimes, simply, it’s about density. With PET, we’re all about resolution. How small can something be identified as positive is dependent on the concentration of radiotracer within it. If something is on a micromolecular level, we’re not going to see it regardless of how good our radiotracer is because it doesn’t have enough resolution. But even very small lesions, because the density of PSMA is so high within these lesions, we can identify these very small lesions. In this case perhaps 18F-fluciclovine because it’s a different mechanism of an action, it’s an amino acid transporter, it doesn’t have the density of uptake to be shown as being positive outside the issue of background. Those are the 2 comments I would make when I look at this comparison.

That’s perfect. When we talk about this for the radiation oncologist looking at these images, unfortunately we’re seeing that where there’s probably a lesion, their PSMA is detecting that lesion but 18F-fluciclovine did not within a distant site of metastasis. With that, we’re getting better visualization of smaller amounts of disease using a PSMA-imaging agent.

What about lymph nodes? Our next image is a set of positive lymph node metastasis looking at 18F-fluciclovine on the top and 68Ga-PSMA-11 on the bottom. Both modalities are detecting the lymph nodes, but you can see a much more robust pattern. I’ll let Louis as the imager give a little more nuanced comment to it.

Here you get a good sense of it, just to talk about physiological uptake. Both radiotracers have physiological uptake within the liver, and you can see radiotracer uptake within the kidneys but notice the difference in the PSMA study of the kidneys compared with the 18F-fluciclovine study. Here we have negative uptake within a renal cyst. But we have positive nodes, which are in the retroperitoneum and are clearly more identifiable and more scintigraphically evident on the PSMA study. We have to wonder why that is. We have to look at the mechanism of the radiotracer and understand that perhaps it’s because of the density of radiotracer being accumulated. In other words, does this tumor that has metastasized the retroperitoneum have a higher concentration of PSMA on its surface, which is allowing the radiotracer to outperform the 18F-fluciclovine? Regardless of the ability of the 18F-fluciclovine to identify the nodes, it’s a matter of concentration of radiotracer based on the molecular behavior of the tumor. In general, we’ve seen that largely be the case, that nodes—especially extrapelvic nodes—are better identified with PSMA than they are with 18F-fluciclovine. But that’s not entirely the case. As you see from some of the data within this paper, there are cases where 18F-fluciclovine outperformed PSMA on individual patients within the nodal beds, which is an interesting commentary on disease heterogeneity; that’s interpatient heterogeneity. There’s a lot going on.

Let’s get a into some of the specific data of the study. When we talk about detection of local recurrence and metastasis, the first thing is the overall detection rate for prostate cancer recurrence in this study: 79.3% in the 18F-fluciclovine arm and 82.8% in the 68Ga-PSMA-11 arm. Local recurrence was detected in 37.9% of patients on the 18F-fluciclovine arm and 27.6% on the 68Ga-PSMA-11 arm. Local pelvic lymph node recurrence was detected in the 18F-fluciclovine arm vs the 68Ga-PSMA-11 arm in 46.6% of patients vs 50%, and in extrapelvic lymph node metastasis in 41.4% vs 51.7%. Finally, when we look at bone metastasis, we’re looking at 25.9% vs 36.2%. Louis, walk us through this slide, which shows it in graphical form. It shows us issues with respect to 1 study, the other study, both, or none.

We get back to the idea of local recurrence to start. First, we see that there were no patients where local recurrence was identified only by PSMA and not by 18F-fluciclovine, but the converse was true where 18F-fluciclovine was able to identify a local recurrence and PSMA could not. When we look at the other sites of disease—the pelvic lymph nodes, the extrapelvic spaces, within the bone—we see that there are patients for whom only PSMA could identify the disease within the pelvic nodes, within the bones, and within the extrapelvic lymph nodes. That gets you to the large conclusions we’ve made, which is that the radiotracers perform very admirably. They perform admirably, but PSMA outperformed 18F-fluciclovine when it came to spaces outside the local bed. In this case, the local bed favored 18F-fluciclovine in terms of the ability to identify disease.

Videos

Drs Steven Finkelstein and Louis J. Mazzarelli highlight data revealed in the article “A Prospective Head-to-Head Comparison of 18F-Fluciclovine With 68Ga-PSMA-11 in Biochemical Recurrence of Prostate Cancer in PET/CT,” by Birgit Pernthaler, et al.

Presentation: Head-to-Head Comparison of 18F-Fluciclovine and 68Ga-PSMA-11 in Prostate Cancer

The addition of new 3D imaging capabilities to a mobile C-arm system may enhance procedure planning, improve real-time intraoperative guidance, and significantly reduce procedure times for endovascular surgery procedures.

Endovascular surgeons who use the Zenition image-guided therapy mobile C-arm system (Philips) will now have 

advanced 3D image guidance capabilities thanks to a new partnership agreement

 between Philips and the United Kingdom-based company Cydar, which specializes in cloud-based procedure map software.

Philips says combining the Zenition system’s versatility and innovative image capture with the artificial-intelligence (AI)-enabled technology and patient-specific 3D mapping of target vasculature offered with Cydar EV Maps may lead to greater efficiency, safety, and improved outcomes. Not only does the technology bolster procedure planning, Philips notes the real-time updates of the 3D mapping can provide enhanced guidance for addressing findings encountered during endovascular procedures.

The integration of these technologies could lead to a 50 percent reduction in radiation exposure and a greater than 20 percent reduction in procedure times, according to Philips.

“With Philips’ integrated portfolio, using validated AI and cloud technologies, we can facilitate collaborative care to optimize surgical pathways,” said Karim Boussebaa, the General Manager of Image Guided Therapy Systems at Philips. “Philips has a strong global network of mobile surgery systems and recognizes that Cydar’s EV Maps solution can play a key role in further improving our integrated offering for endovascular procedures.”

Emerging technology could provide key advances in pre-op planning and intraoperative management in endovascular surgery.

New AI-Enabled 3D Mapping with C-Arm System May Improve Efficiency and Outcomes of Endovascular Procedures

Case submitted by Rathachai Kaewlai, M.D., via 

Image IQ Quiz: What is Your Diagnosis?

Although a concept that seemed unfamiliar to many at the onset of the pandemic, working from home--especially in health care--is nothing new for radiologist Elizabeth Hawk, M.S., M.D., Ph.D., an assistant professor of radiology at Stanford University Medical School. Hawk, who is also Director of Innovation Engagement and a Matrix Pod President for 

, has been working from home using teleradiology for years. 

, Hawk led a panel discussion focused on flexible work arrangements and their impact on career advancement for women radiologists. Hawk, who is a mother of 3, told 

 that a flexible work arrangement, whether it be working from home or on a personalized schedule, has clear positive effects on work-life balance and shouldn't be looked down upon as a disadvantage when it comes to professional advancement. 

sat down with Hawk to learn more about how radiologists can incorporate more flexible work arrangements into their career plans. Watch below:

During the pandemic and beyond, working from home as a radiologist can present clear work-life advantages.

Navigating Flexible Work Environments in Radiology: Elizabeth Hawk, M.S., M.D., Ph.D

In a new distribution agreement with the manufacturer UE LifeSciences, Siemens Healthineers plans to launch the iBreastExam, a handheld cancer screening modality that helps clinicians identify non-palpable breast lumps in a matter of minutes.

In addition to reportedly being a painless alternative to help alleviate patient anxiety about breast examinations, the iBreastExam may serve as an effective tool to facilitate early screening for 

, a disease that is currently expected to claim the lives of approximately 43,250 women in the U.S. in 2022. The 

launch of the iBreastExam in the United States

 will begin this month, according to Siemens Healthineers.

“We’re committed to improving breast cancer-related outcomes by recognizing unmet needs in healthcare delivery, and UE LifeSciences’ iBreastExam clearly addresses a well-defined niche in primary care,” said Niral Patel, the vice president of X-ray products for Siemens Healthineers. “With iBreastExam, we’re making women’s health more equitable and inclusive for all populations.”

, researchers said the iBreastExam and clinical breast exams can be used synergistically as triage tools to “significantly reduce the numbers of patients who need additional diagnostic imaging in resource-limited areas.” The iBreastExam had a specificity of 80.3 percent and a negative predictive value of 94 percent, according to the study.

Siemens Healthineers added that the iBreastExam provides automated digital documentation that can be linked to a patient’s electronic health record.

“An objective, quick, and painless documentation of routine breast exams informs and empowers care providers and women, especially those younger than 40, with dense breasts, or with family history,” noted Brian Englander, M.D., the chairman of the Department of Radiology at Pennsylvania Hospital.

The iBreastExam, a handheld device cleared by the FDA, reportedly enhances routine clinical breast examinations.

Emerging Breast Cancer Screening Tool to See Wider Distribution in U.S.

What kind of impact do fatigue, and experience level have in the assessment of digital breast tomosynthesis (DBT) images?

 demonstrated that experience matters when it comes to interpreting DBT images as does the time of the day when they are interpreted.

In the retrospective study involving over 97,600 mammograms

 interpreted by 18 radiologists over a nearly two-year period, researchers assessed recall rates and false-positive rates for DBT imaging and digital mammography (DM). For radiologists with five years or less of post-training experience, recall rates for the use of DBT increased 11.5 percent with every additional hour in the day but researchers did not see this for patients who had DM. The study authors also found no increase in recall rates for either DBT or DM among radiologists with more than five years of post-training experience.

However, they also acknowledged considerably more detail with DBT, noting “hundreds of images per bilateral screening DBT examination, as compared with four images per bilateral screening DM examination.”

“(Due to) the higher number of images to review, DBT may require greater cognitive resources than DM, which could exacerbate the association between fatigue and radiologist performance,” wrote Ana P. Lourenco, M.D., a professor of diagnostic imaging at the Warren Alpert Medical School of Brown University, and colleagues.

The study authors noted that other researchers had found that DBT had an 

between a 15 to 40 percent reduction in recall rates

. In the new study, the authors also found that patients in the DM group had a higher false-positive rate in comparison to those who had DBT (9.8 percent vs. 8.6 percent) and a higher overall recall rate (10.2 percent vs. 9.0 percent).

“Although DBT achieves a superior (true positive) rate, more junior radiologists appeared to compensate for their fatigue later in the day when using DBT by recalling a broader range of mammograms, more of which were (false positive) findings,” wrote Dr. Lourenco and colleagues.

The study authors noted limitations of the study, including a small sample size of radiologists as well as a lack of randomization for the imaging technology, patients, and radiologists. They also acknowledged that using the 5-year mark to distinguish experience is a rough approximation that may not account for work volume or the varied effect of experience for each person.

While digital breast tomosynthesis is reportedly more effective at detecting cancer, new research suggests a higher recall rate due to fatigue and less experience in interpreting the images. 

Study Looks at Recall Rates for Digital Breast Tomosynthesis and Digital Mammography

I have written a couple of columns in the past about my lack of thrills over pure per-click employment models in radiology. Even the term “pay per click” rankles me a bit. It’s not the type of job I’d want for myself, all other things being equal.

I know this from personal experience. I spent a little over seven years of my career working for what was then the biggest corporate per-click player in the radiology world. That was partially due to the lousy job market in 2011, which gave me few options. It was also my reaction to previous jobs with salaried systems that failed to reward productivity.

Living by the per-click sword showed me ways I could be cut by it. Even that big telerad entity wasn’t always able to supply me with a full work list. Sometimes there just wasn’t enough imaging to keep all the radiologists busy. Sometimes technical problems or even scheduled maintenance would interfere. Whatever the mechanism, there were more than a few instances when I had no options for clicking.

It didn’t seem right. There I was, a professional manning my station and because my employer wasn’t giving me any work, it was not paying me a penny for my time. The employer was the one having issues, but I was suffering the consequences.

Such circumstances recurred often enough over a course of years. I was far from the only rad pointing this issue out to the group’s leadership but there was never corrective action. It didn’t take much cynicism to note that those leaders, unlike us, were salaried so they weren’t sharing our pain.

They didn’t say anything memorable in defense of the status quo. They claimed that, in the grand scheme of things, there really wasn’t much down time so this was not a big deal.

Maybe it wasn’t a big deal to them but when you are the rad whose livelihood depends on churning out RVUs, any moment you’re prevented from doing so feels like an injustice of the first order. This is especially the case when the company’s whole model is based on the notion of maximum efficiency, which they point to when trying to justify the bargain-basement values they assign to their imaging studies. “You’ll make it up on volume” rings hollow whenever you don’t have access to said volume.

Furthermore, most rads seeking to optimize their productivity will tell you they get into a certain rhythm, a “zone,” when they’re reading one study after another. That gets disrupted if the work list is suddenly empty or unavailable, even if just for a fraction of a minute.

Fast forward to the present day. Regular readers of this blog know that I’ve been on the hunt for a new job, after spending three years with a non-per-click gig that followed the one I mentioned above. For job seekers, the market is currently rife with juicy targets. Is it akin to shooting fish in a barrel? Maybe but one doesn’t want to be job hunting again in a few years. This is even more reason to thoroughly investigate all opportunities. Be as certain as possible that one’s next destination is the absolute best.

That said, I have looked at some per-click gigs. For the most part, they don’t stand much of a chance in my considerations. Every now and then, a potential gig turns up with features that turn my head. Some of these benefits include particularly good values for cases, nice schedules, and future upside potential (partnership, etc.).

But there is another change from when I moved to per-click in 2011: I’m older and somewhat wiser. Approaching a per-click gig, I now ask whether the prospective employer has what I’ve come to refer to as a “safety net.” This would be some logistical mechanism for circumstances in which, as result of issues at my employer’s end, I cannot read cases.

It doesn’t have to be huge. Would I like my downtime to be paid pro rata, equal to my average productivity for the preceding 12 months? Sure, but I recognize that an employer can’t afford to just turn on a faucet of money for rads who aren’t generating RVUs. On the other hand, getting zip for my time is a nonstarter for me.

In these conversations, I bend over backwards emphasizing that I do not expect anything if problems are at my end. If I have a power or internet outage, why should my remote employer bear any responsibility? (It would be another matter if I worked on site.)

A funny thing happens in these exchanges. Unless the group already has some sort of safety net in place, the prospective employer invariably tries to convince me that one is unnecessary. For instance, I tell the rad group about times that my biggest-in-the-industry telerad company sometimes had an empty list … and the prospective employer, a fraction of that telerad’s size, proudly tells me about its annual case volume, saying it “never” has an empty list.

I don’t know about you, but when I tell a prospective employer about a concern of mine and it is dismissed with a wave of the hand, I don’t marvel at the rad group’s Jedi mind-trick skills. I get annoyed that the prospective employer is not taking me seriously.

I try being patient with the rad group. I express understanding that it chose a per-click model for a reason. The rad group may have a certain amount of revenue coming in and it has planned how much of that revenue can be paid out for each case.

Still, there’s always some slack in the system. Nobody pays per-click rads 100 percent of their RVU worth. There is excess to cover operating costs, unforeseen expenses and, of course, profit. Certainly, there is enough for there to be a financial “floor” for the rad’s time in consideration and respect for he or she being part of the team and forsaking other opportunities to be at the post.

I also try to go with the flow. Hey, it’s great that you never have downtime. Most rad groups do. We’ve established that you want me to join your team. Why not reassure me by putting a safety net in place, knowing that your flawless system will never require the net to be activated?

Most groups balk at this. Maybe it’s because their negotiators know darned well that downtime happens or could…and they’d rather try to snow me about it than agree that my time as a professional is worth something for which they should pay, however little in terms of compensation. Alternately, they may realize that setting up a safety net would be more work for them than just hiring someone other than me, someone who doesn’t ask these troublesome questions.

That is just fine with me. I prefer an upfront “No” to a drawn-out “maybe” that has no chance of turning into a “yes.” The less said about a “yes” that gets walked back into a “not really,” the better. I recently wasted several weeks of my time playing out such a drama with a per-click employer.

Would rad group employers consider a compensation model that offers a compromise between productivity-driven renumeration and a reasonable “floor” in the event of downtime? 

Is a Pay-Per-Click Safety Net an Option for New Rad Jobs?

No matter one’s profession, expertise, or experience level in a given field, standardization sets a baseline of knowledge we can all understand. By uniting a community around a set of standards, we find something we all agree upon: a place where everyone can find a starting point for their commitment and performance.

For sonographers like myself, standardization lives both closely and at a distance: When we hold ourselves to a high level of performance, each practitioner seizes the opportunity to not only enhance the immediate experience of their patient but elevate the understanding of care in our global community.

Why Does Standardization Matter to the Current State of Health Care?

For some, standardization can seem like a minimum or an essential bar to meet. This is its simplest impact, creating an exact point of entry into a professional community. Young learners grow and evolve from this point over time, accruing increasing momentum, expertise, and respect in the field. Yet they all begin at the same place and build upon the same consistent foundation. Whether piloting a plane, cooking fantastic food, or treating critical ultrasound patients, we work better together from that common ground.

In the current state of health care, the impact of standardization can be staggering. Over the years in our field, sonographers, such as myself, have seen too many instances of patients being referred to secondary care at another lab or clinic without a commitment to high standards, thereby lacking quality of care. The impact on a patient’s health trajectory can be shocking. Many days, our practice can boil down to resolving situations that have simply been missed by others.

These experiences cost patients valuable time and incite frustration, ultimately degrading trust and impacting quality of life. In some cases, the absence of standards can direct patients on an entirely different course of treatment for a disease. It can even lead to loss of life.

That said, the impact of applying standards creates equally intense and exciting results. For global populations, standardization can open conversations about essential training and patient care. As an example, through the 

Inteleos Ultrasound Proficiency Grand Challenge

, we have the chance to welcome critical health interventions to places in desperate need. With only 17 percent of mothers receiving critical ultrasound support in low resource populations, the Proficiency Grand Challenge is an opportunity to make a huge impact.

How Can Standardization Elevate Our Profession and Professional Relationships?

Standardization sets expectations. Simple, right? Yet the complexity of its impact on our relationships reaches well beyond those three words: Standardization draws us all into one conversation, facilitates clear communication, helps communities determine and agree upon assessment metrics, and establishes accountability to ourselves and each other. It sows the seeds to earning critical trust within any field or industry. Licensed drivers, qualified educators, and certified accountants use standardization and certifications to invite a level of essential trust in their relationships.

There shouldn’t be any exceptions to the rule for health care professionals. After all, patients know the difference. Their perception of care is different when they believe the standards are higher. They feel better in moments when they’re already anxious. They feel valued. A patient can tell when he or she is receiving a higher standard of care. It impacts a patient’s confidence and faith in his or her health-care experience. Even in the event of a bad outcome, that confidence paired with compassion allows us to make a huge difference in patient care.

How Does Standardization Lay The Groundwork For Superior Practice, Both at Home and in Our Global Community?

At its heart, standardization aims to drive the professional conversation. We share and agree upon knowledge and best practices, we work to train new learners, and we collectively improve our craft. No matter how simple these practices may seem, their specialized nature is critical to successful performance. If there’s something you don’t know or never learned, you simply don’t know it. For sonographers, this may mean missing a critical imaging detail or producing a picture without holistic value. In addition to this, not having complete and proper clinical information can cause missed diagnosis.

Proper use of equipment, access to training, and consistent guidelines for interpretation impact every sonographer in every community. We are all one village in one conversation with one goal. We are global stewards of high-quality, consistent patient care, and it’s our responsibility to uphold and share the information, expertise, and resulting standards that bring us together. Just because practitioners are working in underserved communities, that doesn’t mean they can’t provide quality care nor does it mean that standards cannot influence and elevate their practice. As members of this community, it’s our job to demonstrate what that means, what it should be and what it can be for all of us.

What Role Does Standardization Play in How We Build the Future?

As medical professionals, we have to make peace with the fact that we’ll never know everything about the human body. Yet, we already know things we never thought possible and this knowledge improves lives every day. I have yet to stop learning in my career, and I hope that day never comes!

Standardization unites our work in a shared conversation. The resulting training and certifications bond our field and strengthen our work holistically. Our patients, our institutions and our peers are better able to understand and value our expertise, allowing us to build relationships based on genuine trust.

Today, we have a unique opportunity to further that conversation, bringing better quality of care to all communities around the world and elevating the entire medical system for all patients. We’re moving modern medicine forward and standardization is a critical tool to help us do it.

Much more than a minimum bar to meet, standardization is our foundation for elevating the level of care and outcomes for all patients. 

Understanding the Value of Standardization in the Future of Medicine

, researchers found that artificial intelligence (AI)-enabled assessment of combined findings from positron emission tomography (PET), computed tomography (CT) and clinical data may provide “substantial improvement” in predicting heart attack risk for people with coronary artery disease.

In the study involving 293 people with coronary artery disease, the authors noted that 22 study participants suffered a myocardial infarction during the 53-month follow-up period. Employing a machine-learning model, researchers assessed the combination of 

F-NaF) PET and quantitative plaque analysis via CT angiography along with clinical findings and compared it to these individual diagnostic measures for identifying myocardial infarction risk.

• Clinical characteristics: (c-statistic 0.64, 95% CI 0.53-0.76)

• Quantitative plaque analysis: (c-statistic 0.72, 95% CI 0.60-0.84)

F-NaF coronary uptake: (c-statistic 0.76, 95% CI 0.68-0.83)

• Combination of all measures: (c-statistic 0.85, 95% CI 0.79-0.91)

For patients with advanced coronary atherosclerosis, the study authors maintained that predicting risk for myocardial infarction doesn’t depend on cardiovascular risk scores, stenosis severity or CT calcium scoring. They said the study showed that disease activity assessment via 

F-NaF PET and analysis of plaque type and burden by coronary CT angiography are primary determinants in risk stratification of this patient population.

“Our machine-learning approach has overcome the challenges posed by collinearity of these variables and, for the first time, has demonstrated that this information is complementary and additive with the combination of both providing the most robust outcome prediction,” wrote Piotr Slomka, PhD, FACC, FASNC, FCCPM, a professor of medicine and cardiology within the Division of Artificial Intelligence in Medicine at Cedars-Sinai Medical Center in Los Angeles, and colleagues. “If confirmed in future studies, this comprehensive approach holds major promise in refining risk stratification of patients with established coronary artery disease, a population for which such prediction is currently challenging.”

Conceding a limited number of patients and myocardial infarction events in this study, the authors noted larger studies are necessary to validate the findings. They said one ongoing prospective study is assessing the aforementioned modalities for the prediction of recurrent events in patients who had a recent heart attack and multivessel disease.

New research suggests an emerging machine learning model that combines findings from advanced imaging with clinical data may improve risk stratification in people with coronary artery disease.