Diagnostic Imaging brings you daily coverage of SIIM 2021 virtual annual meeting.

SIIM 2021 Virtual Annual Meeting: May 24 - 27, 2021

Whitney J. Palmer has been with Diagnostic Imaging since 2011, serving as the Senior Editor since November 2019. She has 20 years experience in healthcare and academic medicine reporting.

Diagnostic Imaging's Weekly Scan: The Top 5

Slideshows

Take a quick look at Diagnostic Imaging's Top 5 stories of the week.

Diagnostic Imaging's Weekly Scan: July 26, 2021, to July 30, 2021

Incorporating artificial intelligence (AI) as a second reader with chest X-ray can improve both radiology resident and board-certified radiologist performance in recommending chest CT exams for patients suspected of having lung cancer.

 a team of investigators from Massachusetts General Hospital revealed that AI could have a roughly 30-percent impact on both residents and providers, catching more findings to drive the most appropriate treatment.

“The use of AI could help to detect pulmonary nodules accurately with chest X-rays, as well as reduce the need for unnecessary chest CT exams in some patients,” said study co-investigator Mannudeep K. Kalra, M.D., a radiologist at Massachusetts General. "This finding can benefit patients by enabling them to avoid unneeded radiation exposure, and it can benefit the healthcare system by preventing certain medical costs.”

Even though chest X-ray is more affordable and more easily accessible, it is limited in its utility for detecting pulmonary nodules, the team said. Its projectional nature means anatomical structures can hide the nodules, effectively creating blind spots for radiologists. In fact, data from the National Lung Screening Trial (NLST) shows that chest X-ray does not lower mortality when compared to low-dose CT.

Consequently, the team set out to determine if it were possible to improve chest X-ray performance.

For their study, they examined 519 chest X-rays – 173 cancer-positive and 346 cancer-negative – collected from the NLST. Eight radiologists – three residents and five board-certified providers – interpreted the exams, and an AI algorithm provided image-level probability of a nodule or mass on chest X-ray, as well as a heatmap of detected lesions. The team assessed reader performance for area under the curve, sensitivity, specificity, false positives, and rate of chest CT recommendation.

According to their analysis, by using AI, residents were able to recommend chest CT in 28 percent more cases of people with suspected lung cancer than without the algorithm (70.2 percent versus 54.7 percent). The tool also reduced unnecessary chest CT recommendations by radiologists by 30 percent (11.7 percent versus 16.4 percent) in cancer-negative patients.

“Our results suggest that AI algorithm benefits less-experience readers in terms of sensitivity, and more-experienced readers in terms of specificity,” the team said. “In this study, the AI algorithm likely reduced such errors by locating such subtle lesions for the residents…AI helped improve the performance of the residents up to the level that was on par with radiologists, which led to a significant increase in the chest CT recommendation rate for patients with visible lung cancer.”

Ultimately, the team said, incorporating AI into computer-aided detection (CAD) can alleviate reliance on hand-crafted features and limit the impact of both high false-positive and high false-negative findings.

“We believe our results support the use of AI CAD systems for the detection of lung cancer in chest X-rays,” the team concluded. “Our AI CAD system can help avoid such issues while assisting readers in identifying subtle lesions that may otherwise be missed.”


For more coverage based on industry expert insights and research, subscribe to the Diagnostic Imaging e-Newsletter 

Articles


Both board-certified radiologists and radiology residents more appropriately suggested chest CT follow-up when using artificial intelligence with chest X-ray.

AI as a Second Chest X-ray Reader Better Informs Chest CT Recommendations for Lung Cancer

Filling defect disrupting the gastric contour and folds with suggestion of outpouching of the ulcer crater with a relatively smooth border consistent with an ulcer. No significant irregularity of the border.

Features that suggest a benign ulcer is an outpouching of the ulcer crater beyond the gastric border, smooth borders, smooth ulcer mound, smooth folds reaching the margin of the ulcer, “Hampton’s line”, and location along the lesser gastric curvature. Features that suggest malignant gastric ulcer includes irregularity of the borders, nodularity of the ulcer mound, and more commonly seen along the greater curvature of the gastric border.

What is the diagnosis in this patient with melena?

Image IQ Quiz: Patient Presents with Melena

Chest CT scans show thickening of lung tissue after six months in more than 70 percent of patients who are COVID-19-positive.

a team of investigators from Italy shared that it is possible to predict which patients are at higher risk for long-term effects by combining a baseline chest CT scan with clinical data.

Although chest CT findings have been used extensively during the pandemic, providers have mainly relied on them to reveal short- and mid-term progression of disease. As the pandemic enters this new phase, there is a need for a more long-term assessment.

“At six-month follow-up 72 percent of patients showed late sequelae, in particular fibrotic-like changes,” said the team led by Damiano Caruso, M.D., Ph.D., medical director with Sapienza University in Rome. “Baseline [lung severity score] and [quantitative chest CT] of well-aerated lung showed an excellent performance in predicting fibrotic-like changes at six-month chest CT.”

In order to learn more about any long-term impacts of the virus, Caruso’s team examined baseline CT scans captured between March 2020 and May 2020 from 118 patients who had moderate-to-severe COVID-19 infection. They followed up with additional CT scans after six months, looking at lung severity scores, quantitative chest CT measures, and qualitative findings.

In addition, the team evaluated three strategies that could predict long-term effects – radiological data alone, clinical data alone, or a combination of the two.

Based on their evaluation, the team determined that 72 percent of patients had fibrotic-like changes in their lungs, and 42 percent exhibited ground glass opacities. Their results indicated that quantitative chest CT measures and lung severity scores could effectively predict these fibrotic-like changes at follow-up.

However, there is a lack of consensus on whether Caruso’s team is taking the best approach in evaluating these lung changes in patients who have had COVID-19. In an accompanying 

, Athol Wells, consultant chest physician at Royal Brompton Hospital, and Anand Devaraj, professor of practice in thoracic radiologist at the National Lung and Heart Institute and consultant thoracic radiologist at the Royal Brompton Hospital, pointed out that all patients with residual disease at six months had a positive fibrotic-like score.

“The apparent increase in the prevalence of ‘fibrotic-like’ change at six months is highly deceptive as baseline scans were performed at admission, now when the disease was most severe, and before periods of prolonged ventilation in some of the study participants,” they said. “A more progressive lung fibrosis hypothesis post-COVID would instead require evaluating change from baseline CT performed at discharge from intensive care or at hospital discharge.”

Overall, they said, there is an urgent need to assess the clinical significance of residual disease on CT, as well as to create predictive models against this end point in future work.

More than 70 percent of patients exhibit lingering sequelae.



Chest CT Reveals Long-Term Abnormalities in COVID-19 Patients

Going through an MRI can be tough for some patients. Being in a small space and having to remain as still as possible can create fear and panic. And these problems can have a snowball effect – movement leads to blurry images that require repeated scans, prompting more discomfort.

“MRIs are frightening for everyone,” said Matthew David Hall, M.D., lead pediatric radiation oncologist at Miami Cancer Institute. “Even when my patients overcome the fear, they have no way to practice being still, so that the scans, or radiation treatments, can be as short and accurate as possible.”

Hall knew there had to be a way to alleviate those fears and help patients stay relatively motionless. To achieve that goal, his clinical team paired up with Reimagine Well, a company dedicated to improving the patient experience during the treatment process.

What they developed is a pre-MRI virtual reality (VR) game, call the MRI Stillness Game, that uses patient compliance as a way to override the uncomfortable stimuli that an MRI exam presents. Using a VR headset that detects movement, the team developed a three-level game that escalates from black-and-white to color. The game is timed to last as long as the MRI exam. Patients are tasked to hold the crosshairs at the center of a circle, increasing their score the longer they stay still.

Throughout the game, patients focus the crosshairs to colorize a rainforest, a hot air balloon scene, and a coral reef. Each level is longer and harder, and, as they play, the simulated MRI noises morph into part of the natural, surrounding environment, such as bird chirps or dolphin clicks. After each round, patients receive a score card on their performance.

The goal, the team said, is to help familiarize patients with the stillness requirement for an MRI exam with the hope that their scores will predict with how well they will successfully complete a scan without moving.

“Often virtual reality headsets are used in hospitals for distraction or entertainment purposes only. With this application, the patient is first acclimated to the sights and sounds of the MRI,” said Leonard Sender, M.D., co-founder of Reimagine Well. “We believe clinicians will be able to correlate the scores in the game to the accuracy of the imaging and the duration of the patient MRI sessions. Our hope is that it will become an important predictive tool, shortening the duration of procedures and getting more utilization out of their MRI suites.”

Three-level game helps patients overcome fears of the scan and learn how to remain as motionless as possible during imaging.

Virtual Reality Game Improves Patient Stillness for MRI

COVID-19 vaccine-related adenopathies are a known entity, but in many instances, they can still give providers pause about what they’re seeing on a screening breast exam.

a team of investigators from the Diagnostic Center for Women in Miami shared a pattern on screening that, when visualized, can bolster a radiologist's confidence in rendering a benign assessment.

“Knowing that the COVID-19 vaccine is administered intramuscularly in the deltoid muscle of the upper arm and understanding the lymphatic drainage pattern of the breast and upper extremity,” said the team led by Michael J. Plaza, M.D., a radiologist at the Center, “we suggest that a pattern of unilateral axillary lymphadenopathy involving the lateral, central, and/or apical groups with relative sparing of the anterior group is typical for vaccine-related lymphadenopathy.”

For their study, they used findings from a screening breast MRI conducted on a 63-year-old asymptomatic female patient. She received her second vaccine dose in her left arm six days prior to the scan.

No suspicious enhancement present in the left breast.

Morphologically abnormal high level II/III axillary lymph nodes (central and apical group) in the presence of normal low level I axillary lymph nodes (anterior group)

“Identifying this pattern of unilateral axillary adenopathy in the setting of ipsilateral COVID-19 vaccine administration and a negative screening breast MRI allows one to confidently give a benign assessment in keeping with a pragmatic approach,” the team said.

Consequently, they added, radiologists may be able to reduce the number of unneeded follow-ups and/or biopsies. This knowledge also pushes back against the practice of delaying breast MRI screening due to recent vaccinations.

Still, the team cautioned, providers should treat every instance with patients who have a history of cancer as an individual experience. Determine the course of action and treatment with a multi-disciplinary team based on cancer type, lymph node drainage pathway, and metastasis risk.

Looking for these features could reduce biopsies and facilitate more timely screening.

Vaccine Adenopathy Pattern on Breast MRI Supports Benign Assessment

 Filling defect diffusely seen broadly within the small bowel consistent with a parasitic infection consistent with ascariasis.

 GI ascariasis appears as potentially single to multiple worms (nematodes) seen as a longitudinal linear filling defect on barium study. Often times, the intestinal tract of the worm itself can be visualized or if its GI tract is empty, a pure filling defect is seen.

What is the diagnosis in this patient with weight loss?


Image IQ Quiz: Patient Presents with Weight Loss

Volpara Health announced Friday it has received 510(k) clearance from the U.S. Food & Drug Administration for the latest version of Volpara Imaging Software.

This version – VIS 3.2 – incorporates learnings from artificial intelligence to augment the robustness of the breast density assessment algorithm. In addition, this clearance approves VIS 3.2 for use on other mammography machines, including those from Giotto and Siemens.

The new version offers improved imaging processing security through the Open Virtual Appliance architecture, and it also makes it easier for Volpara to monitor, services, and update software, company officials said.

“These new innovations improve the overall security, scalability, robustness, and breadth of our breast health offering,” said Ralph Highnam, Ph.D., Volpara’s Group chief executive officer. “Our objective with each enhancement is the pursuit of our mission – to eliminate advanced-stage breast cancer and save more families from cancer.”

With this approval, Volpara marks its fifth clearance since launching Volpara Imaging Software in 2010. It is a key science algorithm that supports the Volpara Breast Health Platform, a product suite designed to improve early breast cancer detection through augmented mammography quality and workflow, volumetric breast density assessment, and personalized breast care.

Software improves on breast density assessment algorithm with artificial intelligence.

Volpara Receives FDA Clearance for Breast Density Algorithm

Warming iodinated CT contrast media to body temperature prior to injection may not be necessary, according to newly published research.

a team of investigators from Valleywise Health in Phoenix revealed that warming contrast agents to body temperature does not necessarily lower the number of adverse reactions in patients.

“The data suggest that maintaining iohexol 350 at room temperature is non-inferior to warming the agent to body temperature before injection,” said first investigator Noor Basharat from Valleywise Health’s radiology department.

Typically, iodinated CT contrast media is warmed to body temperature as a way to reduce viscosity and injection pressure. But, without data that support the overall benefit of warming, hospitals could change their protocols around iohexol 350 prior to injection, they said.

“If warming contrast media were shown to not be advantageous, then institutions would potentially no longer need to expend resources on warming, as well as no longer need to maintain a daily temperature log or provide evidence of device maintenance,” the team said.

Currently, the American College of Radiology has not provided conclusive guidance on warming iodine-based contrast media. But, they do suggest it under certain circumstances, including high-rate power injections, injections with viscous contrast material, arterial injections through small-caliber catheters, or arterial studies where timing and peak enhancement are critical.

To date, though, there has been little investigation into the relationship between contrast media temperature and adverse events. In order to get a better idea, the team conducted a retrospective study of adult patients who underwent CT exams using intravenous iohexol 350 that was either warmed or at room temperature.

Adverse Event Categories

Overall, they enrolled 3,939 patients who had scans with warmed contrast media and 3,933 patients who receive it at room temperature. The average patient age was 48, and the group was roughly half male and half female. Based on the team’s analysis, they saw both study groups showed no significant difference in the overall adverse event rate, extravasation rate, or allergic/allergic-like reaction rate.

“Our findings question the utility of pre-warming contrast media, which can be a resource-intensive process,” they said.

Specifically, they determined that the body temperature group experienced 11 adverse events – all extravasations. The group saw a 0-percent allergic or allergic-like reaction rate. Among the room temperature group, there were 17 adverse events – 13 extravasations and four allergic/allergic-like reactions. Their analysis identified an excess of 0.27 adverse events per 100 patients with the room temperature group – a rate below the 0.6-percent non-inferiority margin.

Based on their findings, the team concluded that expending the resources needed to pre-warm iohexol 350 prior to injection may not be warranted.

Warming media to body temperature does not decrease rate of adverse events. 

Warming CT Contrast Media May Not Be Necessary

When it comes to treating patients suspected of stroke, CT angiography (CTA) and perfusion CT have become integral parts of assessment and triage. But, should they be done with every patient?

Despite the benefits of these scans, the answer isn’t clear cut. In a point-counterpoint series published in the 

experts from the University of Wisconsin School of Medicine and Public Health and University of Pennsylvania Perelman School of Medicine debated the merits of routine use of both scans.

“CTA and perfusion CT have become crucial imaging techniques for making an appropriate diagnosis and triaging these patients to the best therapy,” said Howard A. Rowley, M.D., professor of radiology, neurology, and neurosurgery, as well as chief of neurological MRI, at Wisconsin. “Although the value of the information from these studies has been established, the clinical implementation and application of these studies remain varied.”

To support more consistent CTA use, Rowley and his colleague Anthony D. Kuner, M.D., a radiologist at Wisconsin, pointed to its reliability in identifying large-vessel occlusion for faster triage to mechanical thrombectomy or aspiration and its ability to pinpoint a potential cause of hemorrhage and assess active bleeding.

In fact, they said, Wisconsin providers perform CTA for all emergent stroke evaluations.

“This uninterrupted approach that does not rely on interpretation of head CT yields rapid image acquisition that facilitates faster treatment decisions,” they said.

They also advocated for using perfusion CT at any time during the stroke window because it offers several benefits. Alongside identifying and characterizing stroke mimics, including migraine, tumor, infection, and seizures, it may also be the only imaging technique to capture abnormal findings in as many as 25 percent of cases.

“Recognizing early parenchymal changes on CT and localizing thrombi may be difficult in the charged setting of acute stroke triage,” they said. “Even normal perfusion imaging findings remain helpful because they provide greater confidence in interpretation of negative examination findings, as well as confident disposition.”

Routine use can also streamline workflow and create a predictable, coordinated stroke team approach, reducing delays, indecisions, wrong decisions, and incomplete examinations. Doing these exams more frequently can also make providers more proficient with interpretation.

But, there are other considerations with CTA and perfusion CT, said Linda J. Bagley, M.D., associate program director of the Perelman neuroradiology fellowship, and her colleague Laurie A. Loevner, M.D., division chief of neuroradiology. Using these techniques incorrectly is associated with risk.

It is possible, when perfusion CT is conducted with CTA of the cervical and intracranial vasculature, that patients can experience radiation doses between two and 10 times greater than with a heat CT alone, they said.

In addition, employing these strategies prompts a stroke alert that can divert needed resources from other patients – those with sepsis, pulmonary embolism, or trauma – potentially leading to worse outcomes. It is better, they said, if CTA and perfusion CT are reserved for patients who are actual candidates for mechanical thrombectomy. In fact, they said, implementing an automated post-processing platform has resulted in a 1,500-percent increase in their use of CTA and perfusion CT, corresponding to a 150-to-200-percent increase in thrombectomies – half of which were for patients with stroke mimicking conditions.

Providers should also keep in mind that perfusion CT is inherently limited by a poor signal-to-noise ratio and can be subject to multiple artifacts, such as patient motion and sub-optimal bolus timing. Existing data also shows, they said, that the benefit of perfusion CT patient selection is questionable for patients who present within six hours of symptom onset or those whose symptoms are localized to posterior circulation.

“[Perfusion] CT studies are of questionable reliability, and CTA/perfusion CT may subject patients to an increased radiation dose and possibly divert resources from other critically ill patients, thereby delaying their care,” they said. “More importantly, [perfusion] CT studies often serve as a means to exclude patients from thrombectomy, a relatively low-risk procedure that is cost effective and of benefit to most patients who present within 24 hours of symptom onset with a confirmed large-vessel occlusion.”

Still, said Kuner and Rowley, routine incorporation of CTA and perfusion CT is a best practice for both early and late windows of triage because it give providers the most opportunity for assessment that can guide treatment decisions.

“Stroke triage is ‘like a box of chocolates,” they said, quoting the movie 

 “’You never know what you’re gonna get.’ CTA and [perfusion] CT help reveal the hidden.”

Implementing routine use of these techniques could offer the most opportunity for assessment, but there are risks to doing so.

Volpara Receives FDA Clearance for Breast Density Algorithm

Warming CT Contrast Media May Not Be Necessary

To CTA and Perfusion CT or Not – What’s Best for Stroke Triage?