While conventional imaging and procedures have limitations in the assessment of pelvic lymph node metastases in people with prostate cancer, new research suggests that an investigational positron emission tomography (PET) agent may be beneficial in this patient population.

In a Phase 3 trial of 296 patients recently diagnosed with prostate cancer, researchers assessed use of the prostate-specific membrane antigen (PSMA) agent 

F-rhPSMA-7.3 (Blue Earth Diagnostics/Bracco Imaging) with PET imaging prior to the patients undergoing radical prostatectomy and pelvic lymph node dissection (PLND).

According to a recent presentation of study results at the 22nd Annual Scientific Meeting of the Society of Urologic Oncology (SUO)

F-rhPSMA-7.3 had a 96 percent specificity rate (217 of 226 patients) for detecting pelvic lymph node metastases. The study authors also noted that 

F-rhPSMA-7.3 had a sensitivity rate of 24 percent, a finding that is “consistent” with sensitivity rates reported with other PSMA-targeted radiopharmaceuticals.

Brian F. Chapin, M.D., a co-author of the study, suggested that PET imaging with 

F-rhPSMA-7.3 may provide additional clarity with the staging of people with primary prostate cancer.

“Up to 25 percent of patients with primary prostate cancer may have detectable pelvic lymph node metastases, which are correlated with a risk for recurrence and associated overall survival,” noted Dr. Chapin, an Associate Professor in the Department of Urology in the Division of Surgery at the University of Texas MD Anderson Cancer Center. “Conventional imaging techniques, such as (magnetic resonance imaging and computed tomography), are limited in the information they may provide. Pelvic lymph node dissection (PLND), or pelvic lymphadenectomy, is considered the gold standard in assessing pelvic node lesions, but its use is limited to the planned surgical area. An ideal staging technique for detecting metastatic prostate cancer should include both pelvic nodes as well as more distant soft tissue and skeletal findings.”

The researchers also noted no serious adverse effects with the use of 

F-rhPSMA-7.3. While the study authors noted a 7.9 percent incidence of adverse events that may have been attributed to use of the imaging agent, they noted that injection site pain was the most reported adverse event in 0.8 percent of the study participants.

Could an Emerging PSMA/PET Imaging Agent Improve the Detection of Recurrent Prostate Cancer?

Can MRI-Based Prostate Cancer Screening be a Viable Alternative to PSA Testing?

Articles

For patients recently diagnosed with prostate cancer, the emerging positron emission tomography (PET) agent 18F-rhPSMA-7.3 reportedly has a 96 percent specificity rate for detecting pelvic lymph node metastases, according to Phase 3 study results recently presented at the 22nd Annual Scientific Meeting of the Society of Urologic Oncology (SUO).

Study Assesses PET Agent for Detecting Pelvic Lymph Node Metastases in Patients with Prostate Cancer

Artificial intelligence (AI) could play a key role in improving the consistency of breast density assessment, according to new research presented at the recent Radiological Society of North America (RSNA) conference.

In a retrospective, longitudinal study that examined the effectiveness of an emerging artificial intelligence (AI) model (WRDensity, Whiterabbit.ai ) for assessing breast density, researchers reviewed mammography data from over 61,000 patients who had three or more mammograms over a five-year period. They found the AI model was over 24 percent more consistent than radiologist assessment in assessing breast density via the Breast Imaging Reporting and Data System (BI-RADS).

In a recent video interview, Susan Holley, MD, a co-author of the study who presented the findings at the RSNA conference, said the AI model can help address the “innate variability” that can occur with radiologist assessment of breast density. As Dr. Holley pointed out, adjunctive AI assessment could be beneficial in risk stratification, determining the potential need for supplemental screening and documenting breast density assessment in radiology reports.

“With breast density being so important for risk assessment, supplemental screening and also now with a national breast density notification law that is looking like it’s going to actually happen in 2023, looking at variability and how we can get better at density assessment is even more timely and more critical,” noted Dr. Holley, a radiologist who is affiliated with UNC Health Care in Raleigh, N.C.

For more insights from Dr. Holley, watch the video below.

Videos

In a recent video interview, Susan Holley, MD discussed key findings from a large retrospective longitudinal study, presented at the recent Radiological Society of North America (RSNA) conference, which found that an emerging artificial intelligence (AI) model was over 24 percent more consistent than radiologist assessment of breast density. 

Can AI Improve the Consistency of Breast Density Assessment by Radiologists?

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

Sometimes, one needs a smidge of comic relief. I have referenced one way for radiologists to get (or give) it in this blog: sharing crummy clinical histories we receive for imaging studies. After scrubbing identifying information, more than a few rads post the nonsense they encounter onto social media.

Most don't bother sharing the same-old, same-old. Commonplace not-reasons for exam like "R/O pain" and "eval" are the conversational equivalent of pocket lint. For good or ill, a typical day's work provides plenty of other, more exotic stuff from our "garbage in" pile.

Sometimes the postings are fertile ground for entertaining replies. For instance, someone recently groused about a scan he had gotten to "R/O befoer [sic] discharge." It was a worthy gripe. What was he supposed to be ruling out before the patient got to go home?

Any given rad's perspective on the situation relates to how long she or he has been in the field. A resident still in training might react in outraged disbelief. I went through all that schooling to do this, people's health care is on the line, and we've just thrown in the towel on getting reasonable clinical histories? Folks a decade or three out of fellowship, like me, tend to settle into a grim acceptance that this is just how things are if they don't opt for humor.

Those with sufficient experience know, however, that isn't entirely correct. This isn't just how things are. The quality of clinical history has deteriorated and will continue to do so. I am old enough to remember when a far greater percentage of clinical histories were reasonable ones because physicians expected better (of themselves and one another). From my perspective, things are on a long-term, gradual downhill slope.

It's not just me. Rads 10-20 years my senior, having seen medicine when it was even more professional than on my first day in the field, probably consider things further gone than I do. Those fresh out of training right now have today's state of affairs as their baseline, and no way to know that our field was ever any better unless they listen to (and believe) the complaints of older codgers.

This doesn't just go for radiology, or even medicine in particular. Everything slides towards entropy if sufficient mechanisms aren't in place to prevent it. If you live long enough, you probably regard certain aspects of civilization as completely broken compared to what you recall as the Golden Age of your youth.

Is there any more ground to be lost here? Sure, we still get some decent clinical histories but if trends continue, they will become vanishingly rare. Would that be rock bottom?

"Rock bottom" might not exist in this regard. Talking to the "everything's ruined" elder statesmen, one hears a common refrain: They never imagined things would deteriorate below point X. It seemed inevitable that common sense or collective societal sanity would have provided some sort of safety net but, somehow, that didn't turn out to be the case.

Let us assume I can't imagine just how bad clinical histories will ultimately get. The next best thing I can do is think up the most absurdly bad endpoint, continuing the trend from good clinical histories of yore through non-existent ones of the near future. What could come after that?

Could the day arrive when referrers have so little respect for the system and such absent fear of consequence that they start providing intentionally bad or wrong histories? Could it degenerate down to the equivalent of prank calls or the clerical equivalent of flipping off the rads?

My imagination ran wild with the idea, and I started getting chuckles from lousy clinical histories that haven't even happened yet. (I was on an airplane with nothing else to think about, okay?) I thought I might share some dystopian future clinical histories here.

"Scan this patient because I said so, and if you hassle me about it , I will make things unpleasant for you."

"Patient has radiation deficiency. Please correct this."

"Do the imaging study. All the clinical information is available in the patient's chart. I'm not spoon feeding it to you."

"My patient has too many vague complaints I can't do anything about. Perform whatever imaging you want so I can look like I'm still trying."

"Patient insists she has an allergy to iodine. I told her that's not possible but she insists. Give full dose of contrast without pre-medication so she understands who's the doctor here."

"I hear radiologists give kickbacks to their best referrers. Image these 100 patients sometime this month so I can be one of your VIPs."

"The doctor told me to order a scan, but I'm just a clerk and have no idea what to write here. Please don't get me in trouble."

"Patient has phantom limb pain below AKA. Scan area below the stump to humor him."

"Evacuation proctography needs to be performed this week. There is no clinical reason, I just don't like the rad who's on fluoro."

"Pan scan for patient who says she has cancer. STAT read as is. Priors coming in out of state, but you can reinterpret the whole thing and make an addendum when those arrive. Who's your daddy?"

Lamenting the deteriorating quality of clinical histories that radiologists receive from referring clinicians, this author comically speculates about clinical histories one may see in the near future.

Clinical Histories in Radiology: Could They Get Worse?

'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: November 28-December 3

Slideshows

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

Subcategory indications for asymptomatic screening and diagnostic exams, expanded acquisition parameters, a new section for lymph node reporting and new Category 4 subdivisions are some of the proposed changes that may be coming down the pike for breast magnetic resonance imaging (MRI) updates to the Breast Imaging Reporting and Data System (BI-RADS).

In a recent lecture at the Radiological Society of North America (RSNA) conference, Wendy DeMartini, MD, professor and chief of the Breast Imaging Division in the Department of Radiology at the Stanford University School of Medicine, discussed a variety of proposed, preliminary updates to BI-RADS for breast MRI reporting. Here are a few takeaways.

1. Reporting of new cross-modality structured clinical indications and optional subcategory indications.

 When performing asymptomatic screening, subcategory indications could include elevated breast cancer risk, dense breasts or assessment of patients who have completed breast cancer treatment, according to Dr. DeMartini. For diagnostic workup with breast MRI, subcategory indications could include clinical findings, imaging findings, Category 3 follow-up, biopsy follow-up or implant assessment.

“There are many benefits to this across modalities. … It provides much greater clarity for the reason you are doing an examination for your patients, providers, and partners. (It also) allows separate audits by these different indications which have very different performance outcomes,” noted Dr. DeMartini. “Certainly, that is important for us (in performing) breast MRI where in the United States, we have only one code for breast MRI. There is no separate screening versus diagnostic (breast) MRI code.”

2. Revised acquisition parameters include descriptions of standard full protocol contrast-enhanced breast MRI with a least two post-contrast series, and abbreviated contrast-enhanced breast MRI, which is usually performed in less than 10 minutes and includes at least one post-contrast series.

 Dr. DeMartini also noted there is discussion of “faster” hybrid techniques with early high temporal series, also referred to as “ultra-fast imaging”.

3. The expanded acquisition parameters also include discussion of diffusion-weighted imaging (DWI) as a complement to dynamic contrast-enhanced (DCE) MRI for further characterization of findings seen with DCE.

 Dr. DeMartini noted there are no current plans to introduce BI-RADS reporting of DWI at this time but noted an international consensus statement from the European Society of Breast Imaging (EUSOBI) that provides some structure for reporting DWI.

4. Removal of the “focus” finding type from the BI-RADS lexicon. 

Dr. DeMartini noted these tiny marks of enhancement generally do not have clinical significance and are typically part of the background or other benign enhancement of the breast.

“With modern breast MRI techniques, we should be able to characterize findings of less than or equal to 5 mm as a small mass that meets criteria for a mass … or focal non-mass enhancement,” maintained Dr. Demartini.

5. Introduction of T2 signal intensity as a descriptor for masses.

 Dr. DeMartini noted that both benign and malignant enhancing masses can be T2 hyperintense. She added that a T2 hyperintense mass that is oval and circumscribed with dark internal septations or homogeneous internal enhancement has a very low probability of malignancy at less than 2 percent. Dr. DeMartini said radiologists should characterize T2 masses as hyperintense or not hyperintense. She said hyperintense masses are uniformly bright and as bright as a normal-appearing lymph node.

6. Addition of new section for lymph node reporting.

 Specifically, for breast MRI, one should note whether intramammary, axillary and internal mammary lymph nodes are normal or abnormal appearing, according to Dr. DeMartini. Based on the current evidence, Dr. DeMartini said abnormal-appearing axillary lymph nodes have subjectively asymmetric morphological features in comparison to ipsilateral or contralateral nodes, particularly when nodes are ipsilateral to current or prior breast cancer. She emphasized there is currently no quantitative threshold on breast MRI for asymmetric cortical thickening. Dr. DeMartini also added that asymmetric rounding or the absence of hila is not a sole criterion for abnormal axillary lymph nodes as this can be the case for many small normal lymph nodes.

7. Clarifying the use of BI-RADS Category 3 for breast MRI.

 While acknowledging multiple studies that show BI-RADS Category 3 “can be employed with malignancy rates less than or equal to 2 percent,” Dr. DeMartini said there is a lack of evidence on BI-RADS Category 3 for breast MRI in comparison to the evidence for mammography and ultrasound. Suggesting that Category 3 may be overutilized for breast MRI findings, Dr. DeMartini noted a goal of reserving Category 3 for less than 5 percent of examinations and urged caution in regard to the use of Category 3 for non-baseline examinations.

“There are some data indicating that frequencies of malignancies are higher for things assessed as Category 3 if breast MRI is not the baseline examination,” explained Dr. Demartini.

Dr. DeMartini noted there is sufficient data to use Category 3 to describe masses that are oval, circumscribed, have homogeneous internal enhancement or dark internal septations, are T2 hyperintense, and not new or increasing in size.

8. Addition of BI-RADS Category 4 subdivisions 4A-4C.

 The proposed preliminary changes would include: Category 4A lesions with a 2.5 percent likelihood of malignancy; Category 4B lesions with a 27.6 percent likelihood of malignancy, and Category 4C lesions with an 83.3 percent likelihood of malignancy. Dr. DeMartini added that this risk stratification falls in line with similar BI-RADS Category 4 subcategories for mammography and ultrasound imaging. She noted the subcategories offer "potential benefits for more meaningful practice audits for rad-path correlation and for the setting of patient/provider expectations."

1. Baltzer P, Mann RM, Iima M, et al. Diffusion-weighted imaging of the breast - a consensus and mission statement from the EUSOBI International Breast Diffusion-Weighted Imaging working group. 

2. Edmonds CE, Lamb LR, Mercaldo SF, Sippo DA, Burk KS, Lehman CD. Frequency and cancer yield of BI-RADS Category 3 lesions detected at high-risk screening breast MRI. 

In a recent lecture at the Radiological Society of North America (RSNA) conference, Wendy DeMartini, MD, discussed a variety of preliminary proposed changes to the Breast Imaging Reporting and Data System (BI-RADS) for breast magnetic resonance imaging (MRI) examinations.

What BI-RADS Changes Are on the Way for Breast MRI Reporting?

Compared with prostate-specific antigen (PSA)-based screening for prostate cancer, the use of biparametric prostate magnetic resonance imaging (MRI) could potentially reduce unnecessary biopsies without increasing the risk of underdiagnosis, according to a study presented at the 

Radiological Society of North America (RSNA) 2022 Annual Meeting

“MRI-based screening showed lower proportions of screening-positive cases, lower false positive rates, and higher PPVs [positive predictive values] than those of PSA-based screening,” noted Taek Min Kim, M.D., an assistant professor in the Department of Radiology at Seoul National University Hospital in Seoul, Korea, and colleagues.

While the PSA test is utilized for prostate cancer screening, Kim and colleagues maintained that questions linger about the evidence of the test’s benefits and risks. Accordingly, the researchers compared the performance of biparametric prostate MRI and PSA testing in a large paired-cohort study.

In their retrospective review, the study authors reviewed data from 1,793 men (mean age 59.7) who had biparametric prostate MRI and PSA testing for prostate cancer screening from January 2018 to June 2021. A urology consultation at a tertiary medical center was requested if the PSA level was 3 ng/mL or greater, or if prostate cancer was suspected based on MRI. The urologist in each center determined whether the patient needed follow-up or biopsy. 

The researchers compared the proportion of men with a positive MRI, based on a Prostate Imaging Reporting & Data System [PI-RADS] score of 3-5 or 4-5, with those with a PSA level of 3 ng/mL or greater. The cancer detection rates, false positive rates, and PPVs were compared between the two screening tests. The PPVs were calculated from all screening-positive cases (PPV1) or biopsied cases (PPV2).

Overall, 150 men (8.4 percent) had a positive PSA test, a finding that was significantly greater than the proportion of men with a positive MRI result with 62 men (3.5 percent) with PI-RADS 3-5 and 41 men (2.3 percent) with PI-RADS 4-5.

The cancer detection rates were similar at 1.1 percent for PSA, PI-RADS 3-5, and PI-RADS 4-5. The false positive rate was significantly higher with the PSA test (7.4 percent) in comparison with that of MRI PI-RADS 3-5 scores (2.4 percent) and MRI PI-RADS 4-5 scores (1.2 percent). The PPVs were significantly lower with the PSA test (PPV1, 12.7 percent; PPV2, 36.5 percent) than that for PI-RADS 3-5 (PPV1 32.3 percent, PPV2 58.8 percent) or PI-RADS 4-5 (PPV1, 46.3 percent; PPV2, 87.5 percent).

For patients who underwent a biopsy, the sensitivity for detecting castration-sensitive prostate cancer was 95.2 percent for PI-RADS 3-5 and 90.4 percent for PI-RADS 4-5, according to the study. However, the researchers noted that a low proportion of the patients underwent a biopsy after positive screening results, at 54.8 percent to 58.5 percent with MRI and 35.3 percent with PSA, a factor that may contribute to overestimation of false positive rates for both tests.

A simplified cost-effectiveness analysis found that “despite the high cost, MRI was expected to reduce unnecessary costs for prostate cancer diagnosis,” according to the study authors.

In what may be the first study to assess the effectiveness of biparametric magnetic resonance imaging (MRI) for prostate cancer screening in a large cohort, researchers found that MRI screening had a significantly lower false positive rate and significantly higher positive predictive values (PPVs) than prostate-specific antigen (PSA)-based screening.

For Raymond Y. Kwong, MD, MPH, FACC, FSCMR, long scan times with cardiac magnetic resonance imaging (MRI) can have a significant downstream effect on patient access to this technology. He noted the increased demand for cardiac MRI has led to wait times of three to five weeks for outpatients and four to five days for in-patients to be scheduled for a cardiac MRI.

“These numbers are really unacceptable for patient care from our standpoint,” noted Dr. Kwong, the director of Cardiac Magnetic Resonance Imaging in the Cardiovascular Division of the Department of Medicine at Brigham and Women’s Hospital in Boston, and a professor medicine at Harvard Medical School.

With this in mind, Dr. Kwong recently started using Vista.ai (formerly HeartVista) One-Click MRI, an artificial intelligence (AI)-enabled software that helps improve image quality as well as the efficiency of cardiac MRI procedures. At the recent Radiological Society of North America (RSNA) conference, Dr. Kwong shared his six-month experience with Vista.ai’s One-Click MRI and noted a 31 percent reduction in scan times based on 1,100 consecutive cardiac MRI studies performed between April 2022 and September 2022 for patients with cardiomyopathy or structural heart disease.

When employing full use of the Vista.ai One-Click MRI software, Dr. Kwong found that 90 percent of cardiac MRI studies could be completed in less than 45 minutes in comparison to 25 percent of non-AI-aided studies.

(Editor’s note: This interview was recorded prior to HeartVista’s announcement about the company name change to Vista.ai.)

For more insights from Dr. Kwong, watch the video below.

In a recent video interview, Raymond Y. Kwong, MD, discussed his clinical experience with the Vista.ai (formerly HeartVista) One Click MRI software and recent research, presented at the Radiological Society of North America (RSNA) conference, that revealed a 31 percent decrease in cardiac MRI scan times for patients with cardiomyopathy or structural heart disease.

Can Emerging AI-Guided Software Rein in Scan Times for Cardiac MRI?

New research with ultrasound shear wave elastography (SWE) suggests that liver stiffness, a marker of fibrosis, increases in patients with a history of COVID-19 infection, according to a study presented at the 

“Our study is part of emerging evidence that COVID-19 infection may lead to liver injury that lasts well after the acute illness,” said Firouzeh Heidari, M.D., a post-doctorate research fellow at Massachusetts General Hospital in Boston.

Elevated transaminases in patients with COVID-19 suggest the presence of liver injury during acute infection, but whether this liver injury leads to lasting liver damage remains unknown. To identify lasting hepatic injury, Dr. Heidari and colleagues compared liver stiffness on ultrasound SWE in patients with a history of COVID-19 infection to that of healthy controls.

The retrospective cohort study included 131 patients (67 women) who underwent ultrasound SWE exams between January 2019 and January 2022 at Massachusetts General Hospital. Researchers assessed three groups of patients based on when they had an ultrasound SWE exam and whether they tested positive for COVID-19 on a polymerase chain reaction (PCR) test.

The COVID-19 positive group included 31 patients (mean age 53.1) who had a positive PCR test result at least 12 weeks prior to having ultrasound SWE. Elastography was performed at an average of 44 weeks (12 to 81 weeks) after a positive PCR test result. The pandemic control group was comprised of 50 patients (mean age 55.2) who had ultrasound SWE during the COVID-19 pandemic and had a history of only negative PCR test results. The pre-pandemic control group included 50 patients (mean age 58.2) who had ultrasound SWE before the pandemic. The primary endpoint was the average difference in median Young’s modulus between post-COVID-19 patients and controls after accounting for age, sex and time period.

COVID-19 infection was associated with an average increase in the median Young’s modulus of 1.71 kPa. The COVID-19 positive group had a higher median liver stiffness (7.68 kPa) in comparison to the pandemic control group (5.99 kPa). The pre-pandemic control group had a median liver stiffness of 7.01 kPa.

“The trend of declining liver stiffness among controls is not completely understood but it may be a consequence of changing referral patterns” during the COVID-19 pandemic, said Heidari and colleagues. They also noted that patients referred for ultrasound SWE before the pandemic were older than those referred after the start of the pandemic.

“We don’t yet know if elevated liver stiffness observed after COVID-19 infection will lead to adverse patient outcomes,” said Dr. Heidari. “We are currently investigating whether the severity of acute COVID-related symptoms is predictive of long-term liver injury severity. We hope to enrich our existing database with additional patient data and a broader scope of covariates to better understand the post-acute effects of COVID-19 within the liver.”

The retrospective study involving the use of ultrasound shear wave elastography showed a significant increase in liver stiffness 44 weeks after the diagnosis of COVID-19 in comparison to pre-pandemic and pandemic controls. 

New Ultrasound Study Shows Increased Liver Stiffness 10 Months After COVID-19 Infection

In patients with fibrosing interstitial lung disease (ILD), the presence of associated pulmonary hypertension has a significant impact on the appearance of thorax computed tomography (CT). In a recently presented study at the 

, researchers found that main pulmonary artery and right ventricular diameter findings were predictors of pulmonary hypertension.

“In patients with fibrosing ILD, a combination of radiologic parameters may be used to predict the presence and severity of pulmonary hypertension,” and to guide clinical decisions on right heart catheterization, lung transplant assessment and prognosis, said Lubna Siddiqi, M.D. a resident at the Department of Diagnostic Radiology at the Schulich School of Medicine & Dentistry at Western University in London, Ontario, Canada, and colleagues.

The development of pulmonary hypertension in patients with fibrosing ILD is associated with considerably shorter survival. While earlier detection is desirable, the current gold standard for diagnosis is right heart catheterization with risk stratification being based on clinical and echocardiographic findings. However, pulmonary hypertension can occur in asymptomatic patients and echocardiography may not detect mild or early pulmonary hypertension. Dr. Siddiqi and colleagues aimed to use objective values from CT imaging to predict the presence of pulmonary hypertension in patients with fibrosing ILD.

An initial 128 patients with fibrosing ILD followed at the London Health Sciences Centre in Ontario underwent right heart catheterization. Of these patients, 94 patients (average age 66.2; 66 male) with thorax CT within six months of right heart catheterization were assessed for main pulmonary artery to aorta ratio, right to left ventricular ratio, caliber of the right and left pulmonary arteries, interventricular septal thickness, and right ventricular wall thickness, presence of pulmonary thromboembolic disease and degree of mosaic attenuation. 

Pulmonary hypertension was defined as a mean pulmonary arterial pressure of 25 mmHg or greater. Radiologic data were analyzed to determine the best predictors of associated pulmonary hypertension, with the results adjusted for sex, body mass index and age.

Of the patients, 38 (40 percent) underwent right heart catheterization electively and 56 (60 percent) as part of a lung transplant assessment. The average mean pulmonary arterial pressure was 25 mmHg, wedge pressure was 9 mmHg, cardiac output was 5 L/min and pulmonary vascular resistance was 3.3 Wood units. Overall, 66 patients (70 percent) had pulmonary hypertension with an average mean pulmonary arterial pressure of 38 mmHg and a pulmonary vascular resistance of 5.7 Wood units.

Main pulmonary artery and right ventricular diameters were independent and significant predictors of pulmonary vascular resistance (F ratio 6.64, 6.32; respectively). However, the correlation between main pulmonary artery and right ventricular diameters was weak and not significant. The main pulmonary artery diameter was also a significant predictor of mean pulmonary arterial pressure (F ratio 11.5).

While the authors noted the small sample size and said they “will seek validation of these data in a separate cohort of patients,” they suggested that “these parameters could be especially helpful in populations for whom right heart catheterization is not readily available or is a contraindication.”

Main pulmonary artery and right ventricular diameters on computed tomography (CT) scans of the thorax were predictors of pulmonary hypertension.

Radiologic Parameters of Pulmonary Hypertension on Thorax CT May Help Risk Stratify Patients with Fibrosing Interstitial Lung Disease

An artificial intelligence (AI) algorithm developed to estimate volumetric breast density (VBD) from three-dimensional (3D)-reconstructed digital breast tomosynthesis (DBT) images showed potential in a study presented at the 

. Moreover, researchers found that the algorithm’s VBD estimation (which does not require raw image data) and body mass index (BMI) were associated with breast cancer risk.

“Such a tool can enable large retrospective epidemiological and personalized risk assessments of breast density with DBT,” said Vinayak Ahluwalia, B.S.E., an M.D. candidate at the Perelman School of Medicine at the University of Pennsylvania, Philadelphia, and colleagues.

While the use of DBT is increasing in breast cancer screening, current methods estimate VBD from two-dimensional (2D) images and often require raw DBT data, which is rarely stored by clinical centers. In this study, Ahluwalia and colleagues developed a computational tool based on deep learning for the assessment of VBD using only 3D-reconstructed DBT images. They also analyzed the ability of the tool to perform individualized breast cancer risk assessment.

The study included data from 1,080 negative DBT screening exams obtained between September 2011 and November 2016 from the Hospital of the University of Pennsylvania. Both 2D raw and 3D-reconstructed DBT images were available in craniocaudal and mediolateral oblique views, with 7,850 views in total. The mean age of the participants was 57 years, the mean BMI was 28.7 kg/m2 and the racial makeup of the cohort was 41.2 percent White, 54.2 percent Black and 4.6 percent other individuals. The researchers generated corresponding 3D reference-standard tissue segmentations from previously validated software that employed both 3D reconstructed slices and raw 2D DBT data.

Ahluwalia and colleagues trained a 3D deep learning model that used U-Net architecture to create a three-label image segmentation task for background, dense tissue, and fatty tissue. The dataset was randomly divided into training (70 percent), validation (15 percent) and test (15 percent) sets. The labels were evaluated overall and separately with the weighted Dice score (DSC), with 0 signifying no overlap and 1 signifying perfect overlap.

When the AI algorithm was evaluated on an independent testing set, it achieved unweighted DSC of 78 percent and weighted DSC of 56 percent. The algorithm accurately segmented the three labels of background, fatty tissue, and dense tissue with respective DSCs of 94 percent, 89 percent, and 49 percent. The AI algorithm performed similarly among racial groups, according to the study.

The researchers also evaluated whether the VBD estimated from the algorithm and BMI were associated with breast cancer risk in an independent case-control set of 193 patients who developed breast cancer and 714 control patients from the same hospital system. The results showed that VBD (odds ratio of 1.27) and BMI (odds ratio of 1.37) were statistically significant factors in individualized breast cancer prediction.

“We have created an AI algorithm for estimating VBD from 3D-reconstructed DBT images that does not require the use of raw image data and could aid in predicting an individual’s risk of breast cancer,” noted the study authors.

An emerging artificial intelligence algorithm, developed to estimate volumetric breast density from 3D-reconstructed digital breast tomosynthesis images, could potentially facilitate individual risk assessments for breast cancer.