Diagnostic Imaging brings you coverage of ARRS 2022 annual meeting.

Coverage of ARRS 2022 Annual Meeting

's Weekly Scan is your chance to catch up on the latest radiology content, from new research and device news to the latest case quizzes and conference reports. Review the slideshow below to see this week's top content:

Diagnostic Imaging's Weekly Scan: July 31-August 6

Slideshows

Catch up on the top radiology content of the past week.

C) Progressive multifocal leukoencephalopathy 

Day 1 images

Day 2 images

Articles

Image IQ Quiz: 36-Year-Old Patient with Six-Month History of 'Brain Fog'

Recognizing the potential benefits and daunting challenges of incorporating artificial intelligence (AI) models into radiology workflows, researchers have published a guidance that examines four key questions on the process for evaluating and implementing AI-enabled modalities into practice.

, the authors discuss the makeup of an AI governance and management structure, pertinent factors for assessing AI-powered imaging modalities, parameters for implementation into practice as well as monitoring of AI tools for effectiveness.

Here are seven key takeaways from the guidance.

1. Radiology AI algorithms currently comprise the biggest amount of AI models that have been cleared by the Food and Drug Administration (FDA), according to the article authors. Not only are radiologists represented in organizational AI governance structures, the study authors also pointed out that radiology groups are leading the implementation of AI imaging at many health-care institutions.

2. In describing the makeup of an AI governing body, the authors said those defining priorities, strategies, and the scope of evaluating and implementing AI models at larger institutions may include leadership from multiple imaging departments, electronic health record (EHR) managers, information technology (IT) management, a legal representative, AI experts as well as institutional administration liaisons. However, the involvement of end users of the AI software is particularly critical, according to the article authors.

“Incorporating end users into the governance structure is of utmost importance to consider their needs and concerns about an (AI) algorithm and to include (them) into the decision-making process,” wrote study co-author Curtis Langlotz, M.D., Ph.D., the director of the Center for Artificial Intelligence in Medicine and Imaging at Stanford University, and colleagues.

(Editor’s note: For related video content, see “

Assessing and Implementing Artificial Intelligence in Radiology

Essential Questions for Assessing Artificial Intelligence Vendors in Radiology

3. Determining who pays for the installation and maintenance of AI models is an important consideration for radiologists in community hospital settings. While the AI models would be monitored by the radiology department, the authors maintained that radiologists would need to emphasize how the given AI model would benefit the overall health system as well as radiologists in order to prevent the costs from being fully absorbed by the radiology department.

“If a health system bears the financial burden for AI, radiologists must develop the value proposition for each model. If models are seen as only improving radiologist efficiency or accuracy, then the radiology group may be asked to bear some or all of the financial cost” noted Langlotz and colleagues.

4. Does the data set used in the development of the AI model align with clinical use in your practice’s patient population? The article authors emphasize assessment of inclusion and exclusion criteria in the model development for potential bias. Another important consideration is whether there has been external validation of the AI model. In order to gauge the accuracy and reproducibility of the AI model, Langlotz and colleagues suggested that institutions perform a test of the model prior to incorporating it into daily use.

5. For risk assessment of AI tools, the study authors recommend a risk categorization framework issued by the International Medical Device Regulators Forum (IMDRF), which considers the nature of intended use cases as well as the potential impact of the software upon clinical decision-making. For lower-risk AI algorithms, such as those that aid in worklist prioritization, the IMDRF system can also be beneficial, according to Langlotz and colleagues.

6. Ensuring appropriate data security is a key aspect of implementing AI tools into practice. Noting the use of data maps to convey data location, flow and encryption status for projects, the article authors emphasized appropriate data security and sharing practices when vetting AI tools. Langlotz and colleagues added that those assessing the AI tools should be especially vigilant for models that are cloud-based or those developed internally at an institution for clinical use.

“This detailed analysis and accountability for data security and robust governance of data use are vital steps toward ensuring patient protection and public trust,” explained Langlotz and colleagues.

7. Acknowledging that post-AI implementation monitoring can be challenging for smaller practices and those without dedicated informatics support, the article authors suggest the use of automatically populated registries to ease monitoring burden and looking for AI vendors that can provide monitoring data capture in their products.

Researchers discuss key parameters for the assessment, implementation and post-implementation monitoring of emerging artificial intelligence (AI) tools in radiology practices large and small.

Seven Takeaways from a New Guidance on AI Implementation in Radiology

The use of electronic health record (EHR) interventions may have a considerable impact in curtailing the use of iodinated contrast media (ICM) for computed tomography (CT) examinations, according to a new study published in the 

, researchers examined the use of two EHR interventions to conserve ICM use in a 

retrospective study involving 78,792 patients

 who had a least one CT exam between April 1, 2022 and July 3, 2022 at a multi-site healthcare system. The study authors said the EHR interventions resulted in a 12 percent decrease in contrast-enhanced CT exams per day and a 15.2 percent decline in orders for CT with ICM per day.

“The study showed a significant decline in utilization of contrast-enhanced CT after implementing two EHR order entry-based interventions designed to conserve iodinated contrast media … . The findings indicate the ability to achieve rapid changes in practice after system-wide EHR implementation,” wrote lead study author Daniel I. Glazer, M.D., the medical director of computed tomography and director of Cross-Sectional Interventional Radiology (CSIR) at the Brigham and Women’s Hospital, and colleagues.

According to the study, the first EHR intervention, instituted on May 10, was an alert that would appear whenever clinicians would order neck, chest, abdomen, or pelvis CT exams with contrast enhancement. The study authors said the alert would remind ordering clinicians about the ICM shortage and suggest alternative imaging modalities and/or imaging strategies specific to the requested CT order. A second EHR intervention, instituted on May 16, changed a box for detailed clinical information from optional to required for CT orders with IV contrast media. If no additional information was provided, the exam proceeded without IV contrast media, according to the study authors.

(Editor’s note: For other related content, see “

Emerging Strategies for Managing the Acute Shortage of Iodinated Contrast Media

Iodinated Contrast Media Shortage: Practical Solutions and Applications in Radiology

The study revealed that the weekday average number of patients who had a contrast-enhanced CT exam went from 726 in the pre-EHR intervention period (April 1-May 9) to 639 in the second post-EHR intervention period (May 16-July 3). Overall, researchers noted a 5.1 percent decrease in CT exams performed with ICM between the pre-EHR intervention period (53.8 percent) and the second post-EHR intervention period (48.7 percent).

“The findings may provide a model for the use of EHR-based initiatives to impact behavior in the event of future crises that require prompt practice changes,” suggested Glazer and colleagues.

In regard to study limitations, the data was derived from one health-care system, albeit one with multiple sites. Accordingly, the authors note the findings may not be applicable to other health-care settings. Given that the study was a little over three months in duration, Glazer and colleagues noted that long-term sustainability of the changes in CT ordering patterns is unknown. They also pointed out that the study included CT exams that were scheduled and protocoled prior to the use of the EHR interventions.

“The observed decline in contrast-enhanced CT utilization may have been greater if the analysis had been limited to only examinations ordered after the (EHR) interventions were in place,” added Glazer and colleagues.

A recent study found the use of an alert and a request for more clinical information in a multisite health system’s electronic health record (EHR) system led to a 12 percent reduction in contrast-enhanced computed tomography (CT) exams per day and a 15.2 percent reduction in orders for CT with contrast media per day.

Can EHR Interventions Have an Impact in Conserving Iodinated Contrast Media in Radiology?

Offering the promise of improved efficiency with in-room workflows, reduced strain with patient positioning and robust image quality, GE Healthcare has introduced a new fixed X-ray system with Definium 656 HD.

In order to alleviate some of the physical demands with patient positioning and ensure optimal image capture, the 

 offers a variety of features. GE Healthcare said these features include: 5-axis motorization and automatic positioning; a 12” touchscreen on the tube head console that facilitates automated adjustments to in-room workflows; and 3D camera technology through the system’s Intelligent Workflow Suite that enables more consistent image quality and prevents unnecessary repeat X-rays.

The Definium 656 HD platform incorporates artificial intelligence (AI) to provide enhanced anatomic detail and clarity through the use of Helix 2.2 advanced image processing and 100 um FlashPad HD detectors, according to the company.

GE Healthcare said other benefits of the Definium 656 HD system include multi-level image slice capabilities with the VolumeRAD™ Digital Tomosynthesis feature and the enhancement of the Auto Image Paste feature with AutoSpine, which facilitates precise and efficient stitching of long images.

“(The Definium 656 HD system) enables healthcare providers to benefit from the highest levels of motorization, automation, assistive intelligence, and advanced applications offered to date in GE Healthcare’s fixed X-ray portfolio,” noted Tanya Lancaster, the general manager for Fixed X-Ray at GE Healthcare. “Not only does the system provide optimal image quality, but it can also reduce the physical workload for (clinicians) and streamline the overall exam workflow.”

The Definium 656 HD fixed X-ray system reportedly features enhanced, artificial intelligence (AI)-driven image processing, facilitates radiology workflows, and reduces patient positioning time.

GE Healthcare Launches AI-Enabled X-Ray System

While reverse transcription-polymerase chain reaction (RT-PCR) testing is considered the standard for the diagnosis of COVID-19, an emerging artificial intelligence (AI) algorithm may enhance early diagnosis of the disease on initial chest radiography screening.

In a recent retrospective study published in 

, researchers compared radiologist assessment of initial chest X-rays for COVID-19 versus an AI algorithm model comprised from four deep convolutional neural networks. The algorithm was trained with frontal chest X-ray images from 3,647 patients who were positive for COVID-19 and 1,908 patients who did not have COVID-19, according to the study.

In the testing of the algorithm with a subset of initial X-ray images for 225 patients who tested positive for COVID-19 and 255 patients who tested negative for the disease, the study authors found that the algorithm had an area under the curve (AUC) of 85 percent versus a 71 percent AUC for a consensus of five radiologists with experience ranging between 1 to 14 years.

Acknowledging the algorithm’s lower sensitivity for diagnosing COVID-19 in comparison to similar algorithms from other studies, the study authors said the focus of their research was the initial X-ray screening “when pulmonary findings are less conspicuous” than more advanced stages of COVID-19. The researchers said this focus is more conducive to real-world applications involving early diagnosis to facilitate effective triage and timely treatment intervention.

“The advantage of AI may lie in its ability to identify changes in medical images before they become visible to the human eye. That would be the case with COVID-19 patients before abnormal lung densities were conspicuous in (chest X-rays),” wrote study co-author Juan I. Rayo Madrid, MD, who is affiliated with the Department of Nuclear Medicine at Complejo Hospitalario Universitario de Badajoz in Badajoz, Spain, and colleagues.

While recognizing that the predictive power of initial X-ray screening is not optimal for diagnosis of COVID-19, the study authors noted that the modality is inexpensive and widely available. These attributes could make the combination of the AI algorithm and initial X-rays beneficial in limited resource environments where RT-PCR testing results are delayed or unavailable, according to Madrid and colleagues.

“Applying this tool systematically to all (chest X-rays) performed in any health center would allow detection of the disease early without other tests,” noted the study authors.

Madrid and colleagues also suggested that the combination of chest X-rays with adjunctive AI may be more beneficial than RT-PCR testing for ongoing monitoring of lung involvement in patients diagnosed with emerging variants of COVID-19.

“We have experienced continuous outbreaks of new variants whose severity, infectious potential, and symptoms are initially uncertain and need to be intensively investigated for weeks or months. In this context, the RT-PCR test may be less sensitive,” said Madrid and colleagues.

In regard to study limitations, Madrid and coauthors noted the AI algorithm software was trained and tested in the same centers. They acknowledged a lack of information on symptoms, signs and additional tests ordered for patients suspected of having COVID-19. The study also did not examine the potential contribution of the AI algorithm software to radiologist assessment in clinical practice, according to Madrid and colleagues.

Preliminary research revealed an area under the curve (AUC) of 85 percent for an artificial intelligence (AI) algorithm in diagnosing COVID-19 on initial chest X-rays in comparison to a consensus 71 percent AUC for five radiologists.

Could AI Enhance Early Detection of COVID-19 on Chest X-Rays?

I have been struggling. I have previously been open about my experiences with professional burnout. While I have made significant changes that have helped me over the years, it is a continuous battle with a combination of work and life stresses as contributors. My burnout has not been cured, and I am constantly struggling to manage it.

I am not alone. We need to recognize that many of our colleagues are also struggling and may feel ashamed to openly acknowledge this. My personal story is only a single experience among countless others from health-care workers across all disciplines. What we have all endured over the past two and a half years has challenged us in ways we never previously considered. Our lives before 2020 will never return. We need to embrace (or at least come to terms with) our “new normal.” This is especially true for health-care workers. Burnout was a significant issue prior to the onset of COVID-19 but it has become more critical since 2020.

While we now have effective vaccines and some helpful treatments, the COVID-19 pandemic has created increased stress and anxiety related to the unknown long-term effects of infection and the uncertainty that each new disease wave brings. The dual impact of sick coworkers and people leaving the workforce has created critical shortages of staffing among physicians, technologists, nurses, and other staff members. The constant uncertainty about staffing has compounded the strains on our workforce, causing individuals to be pushed to the limits to provide optimal patient care. This is all in addition to the omnipresent background of national and global issues such as gun violence, political unrest, endangered protections of rights for females and LGBTQ+ communities, climate change and war. All of these issues, in addition to the constant health concerns, can be significant unseen burdens harming our colleagues.

Through my personal journey, I have explored ways to cope with my stress, anxiety, and ongoing struggles. I am still learning and adapting to what works best for me. One of the most helpful things is sharing my experiences. I have done this to help others recognize they are not alone, and to create awareness about these issues to mitigate associated stigma.

In medicine, there is an unsaid perception that those who provide patient care cannot falter. That level of perfectionism expected of health-care workers needs to be abolished. Physicians and other health-care workers are prone to stumbling (like all humans), especially in the current medical and societal climate. We need to acknowledge it is okay to not be okay, it is okay to not be perfect at everything, and it is okay to not be able to do everything. We need to acknowledge everyone has their own challenges (some of which may be invisible to others) and we need to be supportive of each other.

Life has become busy and hectic for most people. Therefore, supporting others or creating space to support oneself may not be prioritized. For me, it is easier to be there for others than myself (which many can relate to). My internal negative dialogue and “inner critic” hold me to a different standard than others and this can lead to negative spirals of self-depreciation, rumination, and guilt. My inner critic is an extrovert so muting this voice can be difficult. I have internal conversations with my critic, imagining my critic is a good friend who needs advice. I have realized I need to learn to take the advice that I can easily give others. Ongoing reflection and compassion for oneself can help break the hold of the aforementioned negative spirals on one’s outlook.

Learning how to take time for myself and give myself the grace and empathy I can easily provide others has been difficult. I have been working on incorporating breathing exercises, meditation, and other mindfulness strategies into my daily routine. This can vary for individuals. For example, meditation is difficult for me to do but long walks, cooking, or spending time talking to friends and colleagues provide easy daily mindfulness times. The ability to disconnect from devices, email, and social media to create space and time to concentrate on oneself, even for a short period of time, can be especially helpful.

I have been working on burnout research and work for years, and the issue is more critical than ever. We all went into health care for different reasons but one common goal has been providing excellent patient care. We have all heard the phrase “You need to take care of yourself before taking care of others.” However, how many of us truly practice this? I ask everyone to reflect on how they are doing this. Is there anything we can be doing better for ourselves? Is there anything we can do better to support our colleagues?

We owe ourselves the same standard of care we provide for our patients. This isn’t self-indulgence. This is allowing ourselves to experience the grace and humanity that we would readily grant to others. We and our patients deserve nothing less.

Amid a sea of radiology staffing shortages, a global pandemic, and national issues, taking time out for ongoing self-care is more critical than ever to mitigate potential burnout.

Addressing Burnout, Anxiety and Self-Care in the Age of COVID-19

Viz Subdural Hematoma (SDH), a new artificial intelligence (AI)-powered algorithm for diagnosing subdural hemorrhages, has garnered 510(k) clearance from the Food and Drug Administration (FDA), according to Viz.ai, the manufacturer of the module.

Noting a rising incidence of subdural hematomas, Viz.ai said the 

SDH algorithm demonstrated sensitivity and specificity rates of 94 and 92 percent respectively in a multicenter trial

“The algorithm is very sensitive and specific, significantly increasing the number of subdural hemorrhages detected and ensuring patients receive the necessary follow-up from this potentially life-threatening disease,” maintained Jayme Strauss, the chief clinical officer at Viz.ai.

The company emphasizes that Viz SDH is currently the only AI-powered platform specifically geared to identifying and differentiating between acute and chronic subdural hemorrhages. Jason Davies, M.D., Ph.D., said this is a key benefit given the different treatment pathways for these conditions.

“Acute (subdural hemorrhages) require urgent intervention. Therefore, prompt notification will allow us to improve outcomes in emergent cases. Chronic (subdural hemorrhages) have a very different pathway and having an algorithm that identifies both can allow us to take better care of our patients,” noted Dr. Davies, an associate professor of neurosurgery and biomedical informatics at the State University of New York (SUNY) at Buffalo. “The Viz.ai algorithm ensures that patients are quickly identified and routed to the appropriate therapy.”

1. Neifert SN, Chaman EK, Hardigan T, et al. Increases in subdural hematoma with an aging population – the future of Amerivcan cerebrovascular disease. 

2. Viz.ai. Viz.ai receives FDA 510(k) clearance for Viz SDH. Available at: 

https://www.viz.ai/press-release/viz-ai-receives-fda-510-k-clearance-for-viz-sdh

 . Published July 27, 2022. Accessed August 1, 2022.

Viz.ai said the Viz Subdural Hematoma (SDH) artificial intelligence (AI) algorithm provides automatic detection of acute and chronic subdural hemorrhages, facilitating timely triage and treatment of patients.

FDA Clears New AI Tool for Diagnosing Subdural Hemorrhage

Eric Postal, MD, is a diagnostic radiologist with the University of Pittsburgh Medical Center. 

This morning, my Facebook feed included a lamentation/venting from a rad: “According to overnight telerad prelims, 87 percent of the planet has ‘possible mild non-specific bowel findings,’ ‘enteritis/colitis cannot be excluded’ or ‘fecal loading’ needs to be mentioned … .”

As a nearly 11-year telerad veteran, I have to plead guilty but not quite as charged.

Some of the respondents to that thread (not all telerads coming to their own defense) reasonably pointed out that you can’t completely blame the telerad. One person summed it up: “A real miss at night is what matters. A little perspective goes a long way … lack of history, prior images/reports … depends on local QA culture and infrastructure.”

I couldn’t tell you the whys and wherefores, but there is one experience that has been common to all of the places I have covered by telerad. Somehow, I never get all of the history that referring clinicians and even some on-site rads seem to have at their fingertips. It’s not at all rare for me to find out that my “reason for exam” has left out the most important information about the patient or is outright false.

Some referrers have expressed frustration to me about this. They stood on their head to convey the relevant info, which somehow never got conveyed to us. Maybe a lazy or inattentive clerk fumbled the ball. Maybe the order entry system was the medical equivalent of HAL-9000 (“I’m sorry, Doctor, I’m afraid I can’t allow that CT with the history you have input”). Then they see our reports replace the valuable history paragraph they tried to give us with something totally irrelevant and/or untrue.

Another piece of the puzzle, one that the griping on-site/over-reading rads do have some control over, is the QA machine and their own routine input into it.

Rewinding the clock to the days when I was first doing telerad, I can tell you it was an unpleasant shock to my system just how nitpicky and fault-finding some of those on-site rads could be. It felt like some of them were constantly trying to “ding” their telerads’ stats by claiming diagnostic misses and occasionally overcalls for some of the pettiest, most subjective stuff.

Some of the hypercritical over-readers surely thought they were superior to the telerads. “Well, here I am, reading out the overnight cases again and it is time to clean up all of their amateur hour mistakes.” But sometimes I wondered whether there might be a baser motivation at work, like taking the telerads down a few pegs.

Why? Maybe because the on-site rads didn’t like the idea that somebody else got to work from home, and/or resented having to do over-reads for them every morning. Accordingly, maybe they leveraged QA to punish them a little. Perhaps some on-site rads considered it a form of job security: “Hey, if we pepper those telerads with a bunch of QA misses, maybe the hospital won’t think about firing us and contracting with the telerad company for full-time coverage.”

The mate to that puzzle piece was that at least some telerad companies bent over backwards to accommodate those hypercritical on-site rads. Whenever there was a claimed error, the default assumption was that the teleradiologist had indeed done wrong. I know at least some of that was in the name of customer satisfaction. The on-site rads’ opinions mattered, and if the telerad company didn’t agree with their complaints often enough, it might take a toll on goodwill.

A telerad reading multiple hundreds of cases each week, getting disproportionately accused of petty misses and then subjected to a less than impartial “QA committee” soon learns that he or she has a choice: Get a checkered history on performance evaluations or learn to defend him- or herself against the excessive criticisms. Unfortunately, one way to do that is to hedge in reports, offering a lot of “can’t rule out” statements and, yes, overcall enteritis or colitis.

Consider this case in point. I once got accused of “missing” a lesion that I thought was normal lingual tonsillar tissue. I was pretty sure it looked like a bunch of other neck CTs I’d seen before and said so in my defense to the QA committee. They ruled against me without any explanation, such as an annotated image showing where a lesion should be measured, or clinical follow-up proving that something had been found there.

As if this made it better, the QA committee told me it was “only a minor miss.”

Well, I certainly didn’t want my record to suffer. Who knows? Maybe the QA committee knew something I did not. Accordingly, in the probable hundreds of subsequent necks I read for that telerad company (all of whose lingual tonsils looked similar to the one I “missed”), I never again failed to mention that a subtle tonsillar lesion could not be excluded. It should be noted the QA committee never told me I erred by overcalling or hedging.

Not every on-site rad is guilty of this, of course, but some of the folks liable to complain about other rads’ reporting are precisely the hypercritical colleagues who helped form that defensive behavior. Just like with any other self-manufactured problem, the first thing to do after realizing you might have created the hole you are standing in is to stop digging. In this case: Any time you’re thinking about calling something a “miss,” try to step back and think about whether it is really something that needs to be flagged. Remember that your feedback will shape the future behavior of the rad you judge.

For folks running or simply sitting on QA committees, whether you’re with a big telerad company or just a local radiology group: If you don’t ever find yourself functioning as a filter to keep your rads from getting unnecessarily dinged in their performance metrics, don’t be surprised when they find ways to do it themselves. At that point, fixing the methods they have found will be an awful lot harder than if you had averted any need for them in the first place.

Has an overly critical approach from quality assurance committees led to an overly cautious approach in reports from teleradiologists?