Automated identification of pulmonary arteries and veins depicted in non-contrast chest CT scans.

We present a novel integrative computerized solution to automatically identify and differentiate pulmonary arteries and veins depicted on chest computed tomography (CT) without iodinated contrast agents. We first identified the central extrapulmonary arteries and veins using a convolutional neural network (CNN) model. Then, a computational differential geometry method was used to automatically identify the tubular-like structures in the lungs with high densities, which we believe are the intrapulmonary vessels. Beginning with the extrapulmonary arteries and veins, we progressively traced the intrapulmonary vessels by following their skeletons and differentiated them into arteries and veins. Instead of manually labeling the numerous arteries and veins in the lungs for machine learning, this integrative strategy limits the manual effort only to the large extrapulmonary vessels. We used a dataset consisting of 120 chest CT scans acquired on different subjects using various protocols to develop, train, and test the algorithms. Our experiments on an independent test set (n = 15) showed promising performance. The computer algorithm achieved a sensitivity of ∼98% in labeling the pulmonary artery and vein branches when compared with a human expert's results, demonstrating the feasibility of our computerized solution in pulmonary artery/vein labeling.

Macrovasculature and positron emission tomography (PET) standardized uptake value in patients with lung cancer.

PURPOSE: To investigate the relationship between macrovasculature features and the standardized uptake value (SUV) of positron emission tomography (PET), which is a surrogate for the metabolic activity of a lung tumor. METHODS: We retrospectively analyzed a cohort of 90 lung cancer patients who had both chest CT and PET-CT examinations before receiving cancer treatment. The SUVs in the medical reports were used. We quantified three macrovasculature features depicted on CT images (i.e., vessel number, vessel volume, and vessel tortuosity) and several tumor features (i.e., volume, maximum diameter, mean diameter, surface area, and density). Tumor size (e.g., volume) was used as a covariate to adjust for possible confounding factors. Backward stepwise multiple regression analysis was performed to develop a model for predicting PET SUV from the relevant image features. The Bonferroni correction was used for multiple comparisons. RESULTS: PET SUV was positively correlated with vessel volume (R = 0.44, p < 0.001) and vessel number (R = 0.44, p < 0.001) but not with vessel tortuosity (R = 0.124, p > 0.05). After adjusting for tumor size, PET SUV was significantly correlated with vessel tortuosity (R = 0.299, p = 0.004) and vessel number (R = 0.224, p = 0.035), but only marginally correlated with vessel volume (R = 0.187, p = 0.079). The multiple regression model showed a performance with an R-Squared of 0.391 and an adjusted R-Squared of 0.355 (p < 0.001). CONCLUSIONS: Our investigations demonstrate the potential relationship between macrovasculature and PET SUV and suggest the possibility of inferring the metabolic activity of a lung tumor from chest CT images.

Randomized Trial Comparing Standard Versus Ultrasound-Assisted Thrombolysis for Submassive Pulmonary Embolism: The SUNSET sPE Trial.

OBJECTIVES: The aim of this trial was to determine whether ultrasound-assisted thrombolysis (USAT) is superior to standard catheter-directed thrombolysis (SCDT) in pulmonary arterial thrombus reduction for patients with submassive pulmonary embolism (sPE). BACKGROUND: Catheter-directed therapy has been increasingly used in sPE and massive pulmonary embolism as a decompensation prevention and potentially lifesaving procedure. It is unproved whether USAT is superior to SCDT using traditional multiple-side-hole catheters in the treatment of patients with pulmonary embolism. METHODS: Adults with sPE were enrolled. Participants were randomized 1:1 to USAT or SCDT. The primary outcome was 48-hour clearance of pulmonary thrombus assessed by pre- and postprocedural computed tomographic angiography using a refined Miller score. Secondary outcomes included improvement in right ventricular-to-left ventricular ratio, intensive care unit and hospital stay, bleeding, and adverse events up to 90 days. RESULTS: Eighty-one patients with acute sPE were randomized and were available for analysis. The mean total dose of alteplase for USAT was 19 ± 7 mg and for SCDT was 18 ± 7 mg (P = 0.53), infused over 14 ± 6 and 14 ± 5 hours, respectively (P = 0.99). In the USAT group, the mean raw pulmonary arterial thrombus score was reduced from 31 ± 4 at baseline to 22 ± 7 (P < 0.001). In the SCDT group, the score was reduced from 33 ± 4 to 23 ± 7 (P < 0.001). There was no significant difference in mean thrombus score reduction between the 2 groups (P = 0.76). The mean reduction in right ventricular/left ventricular ratio from baseline (1.54 ± 0.30 for USAT, 1.69 ± 0.44 for SCDT) to 48 hours was 0.37 ± 0.34 in the USAT group and 0.59 ± 0.42 in the SCDT group (P = 0.01). Major bleeding (1 stroke and 1 vaginal bleed requiring transfusion) occurred in 2 patients, both in the USAT group. CONCLUSIONS: In the SUNSET sPE (Standard vs. Ultrasound-Assisted Catheter Thrombolysis for Submassive Pulmonary Embolism) trial, patients undergoing USAT had similar pulmonary arterial thrombus reduction compared with those undergoing SCDT, using comparable mean lytic doses and durations of lysis.

A pilot study: Quantify lung volume and emphysema extent directly from two-dimensional scout images.

PURPOSE: The potential to compute volume metrics of emphysema from planar scout images was investigated in this study. The successful implementation of this concept will have a wide impact in different fields, and specifically, maximize the diagnostic potential of the planar medical images. METHODS: We investigated our premise using a well-characterized chronic obstructive pulmonary disease (COPD) cohort. In this cohort, planar scout images from computed tomography (CT) scans were used to compute lung volume and percentage of emphysema. Lung volume and percentage of emphysema were quantified on the volumetric CT images and used as the "ground truth" for developing the models to compute the variables from the corresponding scout images. We trained two classical convolutional neural networks (CNNs), including VGG19 and InceptionV3, to compute lung volume and the percentage of emphysema from the scout images. The scout images (n = 1,446) were split into three subgroups: (1) training (n = 1,235), (2) internal validation (n = 99), and (3) independent test (n = 112) at the subject level in a ratio of 8:1:1. The mean absolute difference (MAD) and R-square (R2) were the performance metrics to evaluate the prediction performance of the developed models. RESULTS: The lung volumes and percentages of emphysema computed from a single planar scout image were significantly linear correlated with the measures quantified using volumetric CT images (VGG19: R2 = 0.934 for lung volume and R2 = 0.751 for emphysema percentage, and InceptionV3: R2 = 0.977 for lung volume and R2 = 0.775 for emphysema percentage). The mean absolute differences (MADs) for lung volume and percentage of emphysema were 0.302 ± 0.247L and 2.89 ± 2.58%, respectively, for VGG19, and 0.366 ± 0.287L and 3.19 ± 2.14, respectively, for InceptionV3. CONCLUSIONS: Our promising results demonstrated the feasibility of inferring volume metrics from planar images using CNNs.

The RSNA Pulmonary Embolism CT Dataset.

Supplemental material is available for this article.

CT Findings and Patterns of e-Cigarette or Vaping Product Use-Associated Lung Injury: A Multicenter Cohort of 160 Cases.

BACKGROUND: e-Cigarette or vaping-induced lung injury (EVALI) causes a spectrum of CT lung injury patterns. Relative frequencies and associations with vaping behavior are unknown. RESEARCH QUESTION: What are the frequencies of imaging findings and CT patterns in EVALI and what is the relationship to vaping behavior? STUDY DESIGN AND METHODS: CT scans of 160 subjects with EVALI from 15 institutions were retrospectively reviewed. CT findings and patterns were defined and agreed on via consensus. The parenchymal organizing pneumonia (OP) pattern was defined as regional or diffuse ground-glass opacity (GGO) ± consolidation without centrilobular nodules (CNs). An airway-centered OP pattern was defined as diffuse CNs with little or no GGO, whereas a mixed OP pattern was a combination of the two. Other patterns included diffuse alveolar damage (DAD), acute eosinophilic-like pneumonia, and pulmonary hemorrhage. Cases were classified as atypical if they did not fit into a pattern. Imaging findings, pattern frequencies, and injury severity were correlated with substance vaped (marijuana derives [tetrahydrocannabinol] [THC] only, nicotine derivates only, and both), vaping frequency, regional geography, and state recreational THC legality. One-way analysis of variance, χ2 test, and multivariable analyses were used for statistical analysis. RESULTS: A total of 160 patients (79.4% men) with a mean age of 28.2 years (range, 15-68 years) with EVALI underwent CT scan. Seventy-seven (48.1%), 15 (9.4%), and 68 (42.5%) patients admitted to vaping THC, nicotine, or both, respectively. Common findings included diffuse or lower lobe GGO with subpleural (78.1%), lobular (59.4%), or peribronchovascular (PBV) sparing (40%). Septal thickening (50.6%), lymphadenopathy (63.1%), and CNs (36.3%) were common. PBV sparing was associated with younger age (P = .02). Of 160 subjects, 156 (97.5%) had one of six defined patterns. Parenchymal, airway-centered, and mixed OP patterns were seen in 89 (55.6%), 14 (8.8%), and 32 (20%) patients, respectively. Acute eosinophilic-like pneumonia (six of 160, 3.8%), DAD (nine of 160, 5.6%), pulmonary hemorrhage (six of 160, 3.8%), and atypical (four of 160, 2.5%) patterns were less common. Increased vaping frequency was associated with more severe injury (P = .008). Multivariable analysis showed a negative association between vaping for > 6 months and DAD pattern (P = .03). Two subjects (1.25%) with DAD pattern died. There was no relation between pattern and injury severity, geographic location, and state legality of recreational use of THC. INTERPRETATION: EVALI typically causes an OP pattern but exists on a spectrum of acute lung injury. Vaping habits do not correlate with CT patterns except for negative correlation between vaping > 6 months and DAD pattern. PBV sparing, not previously described in acute lung injury, is a common finding.

Subclavian artery aneurysmal rupture and left internal mammary artery extravasation secondary to advanced Marfan syndrome.

This case highlights the unusual life-threatening findings found in a patient with Marfan syndrome (MFS) in the emergency department setting. MFS is a rare autosomal dominant disease that affects 1 in 3000-5000 individuals and has a highly variable range of clinical severity. This case is a 63-year-old male with COPD, scoliosis, aortic and mitral valve replacements on warfarin, and MFS who presented with acute onset hemoptysis, tachypnea, and oxygen saturation of 77% on 4 l nasal cannula. Emergent chest computed tomography angiography (CTA) revealed both a contained rupture of a left subclavian artery aneurysm and active extravasation from his left internal mammary artery (LIMA) into his left chest. The patient was on warfarin and reversed with IV vitamin K and prothrombin complex concentrate. Vascular surgery emergently took the patient to the operating room for embolization of his LIMA and stenting of the contained ruptured left subclavian artery aneurysm. The patient was discharged home one month after admission. This case report illustrates the potential severe sequelae of MFS and the importance of rapid recognition by emergency physicians. An expanded understanding of the pathophysiology of MFS has resulted in great advancement in medical therapies and lifestyle modification and thus has significantly prolonged life expectancy in these patients. Increased awareness and familiarity will facilitate continued high-quality management and treatment by emergency physicians.

Interobserver Reliability of the Coronary Artery Disease Reporting and Data System in Clinical Practice.

PURPOSE: This study aimed to evaluate interobserver reproducibility between cardiothoracic radiologists applying the Coronary Artery Disease Reporting and Data System (CAD-RADS) to describe atherosclerotic burden on coronary computed tomography angiography. METHODS: Forty clinical computed tomography angiography cases were retrospectively and independently evaluated by 3 attending and 2 fellowship-trained cardiothoracic radiologists using the CAD-RADS lexicon. Radiologists were blinded to patient history and underwent initial training using a practice set of 10 subjects. Interobserver reproducibility was assessed using an intraclass correlation (ICC) on the basis of single-observer scores, absolute agreement, and a 2-way random-effects model. Nondiagnostic studies were excluded. ICC was also performed for CAD-RADS scores grouped by management recommendations for absent (0), nonobstructive (1 to 2), and potentially obstructive (3 to 5) CAD. RESULTS: Interobserver reproducibility was moderate to good (ICC: 0.748, 95% confidence interval [CI]: 0.639-0.842, P<0.0001), with higher agreement among cardiothoracic radiology fellows (ICC: 0.853, 95% CI: 0.730-0.922, P<0.0001) than attending radiologists (ICC: 0.711, 95% CI: 0.568-0.824, P<0.0001). Interobserver reproducibility for clinical management categories was marginally decreased (ICC: 0.692, 95% CI: 0.570-0.802, P<0.0001). The average percent agreement between pairs of radiologists was 84.74%. Percent observer agreement was significantly reduced in the presence (M=62.22%, SD=15.17%) versus the absence (M=80.91%, SD=17.97%) of modifiers, t(37.95)=3.566, P=0.001. CONCLUSIONS: Interobserver reliability and agreement with the CAD-RADS terminology are moderate to good in clinical practice. However, further investigations are needed to characterize the causes of interobserver disagreement that may lead to differences in management recommendations.

Imaging Findings of Vaping-Associated Lung Injury.

To listen to the podcast associated with this article, please select one of the following: iTunes or Google Play. OBJECTIVE. E-cigarettes are devices that aerosolize nicotine- or cannabis-based concentrates mixed with other solvents and have been marketed as an alternative to cigarettes. E-cigarette use, or vaping, is increasingly popular but has not been proven to be an innocuous substitute for traditional smoking. Several patterns of vaping-associated inhalational lung injuries have been reported in the past few years. This article reviews many of the imaging patterns that have been encountered in association with e-cigarette use. CONCLUSION. E-cigarette use is associated with a range of lung injury patterns that have only recently been recognized as use of these products continues to rise. When the radiologist sees one of these patterns of lung injury, it is important to raise the possibility of vaping-induced lung injury because cessation of vaping is an important step in treatment.

Interventional Therapies for Central Airways.

Central airway obstruction (CAO) is a dangerous and increasingly common problem. CAO refers to lesions causing narrowing of the trachea or mainstem bronchi and is generally divided into malignant and nonmalignant categories. These 2 entities may be caused by a variety of thoracic and extrathoracic diseases. Imaging is critical during the initial assessment of CAO and may help thoracic physicians focus the differential diagnosis and plan the safest and most appropriate diagnostic and therapeutic interventions. However, direct visualization via flexible or rigid bronchoscopy is often necessary for diagnostic and treatment purposes. A large number of procedures can be performed through bronchoscopy, with the goal of relieving the obstruction and improving patency of the airway. Deciding which procedure to perform is based both upon the type of lesion and whether the lesion is due to a malignant or nonmalignant process. Possible interventions include mechanical debridement, laser therapy, argon plasma coagulation, electrocautery, brachytherapy, and stent placement. Immediate postoperative and follow-up imaging is crucial to monitor for immediate, subacute, and chronic complications as well as disease progression and recurrence.

Radiologist Performance in the Detection of Pulmonary Embolism: Features that Favor Correct Interpretation and Risk Factors for Errors.

PURPOSE: This study aimed to assess the factors contributing toward accurate detection and erroneous interpretation of pulmonary embolism (PE). MATERIALS AND METHODS: Over 13 months, all computed tomography pulmonary angiography studies were retrospectively rereviewed by a chest radiologist. Two additional chest radiologists assessed cases with disagreement between the first interpretation and rereview. The number, extent, and location of PE and specialty training, experience, time of study, kV, resident prelim, use of iterative reconstruction, signal to noise ratio (SNR), and reports describing the study as "limited" were recorded. Parametric and nonparametric statistical testing was performed (significance P<0.05). RESULTS: Of 2555 computed tomography pulmonary angiography cases assessed, there were 230 true positive (170 multiple, 60 single PE), 2271 true negative, 35 false-negative (15 multiple and 20 single PE), and 19 false-positive studies. The overall sensitivity, specificity, positive predictive value, negative predictive value and accuracy of radiologists was 86.8%, 99.2%, 92.4%, 98.5%, and 97.9%. Sensitivity for the detection of multiple and central PE was significantly higher than the detection of single and peripheral PE, respectively (P<0.01 for both). The sensitivity of thoracic radiologists (91.7%) was higher than nonthoracic (82.8%) and reached significance for single PE (89.2% vs. 61.4%, P<0.02). Errors were more likely in cases with lower SNR (P=0.04) and those described as limited (P<0.001). Misses occurred more frequently in the upper lobe posterior and lower lobe lateral segments and subsegments (P=0.038). CONCLUSIONS: The accuracy for PE detection is high, but errors are more likely in studies with single PE interpreted by nonthoracic radiologists, especially when located in certain segments and in cases with low SNR or described as limited.

Computed Tomography Aortic Valve Calcium Scoring in Patients With Aortic Stenosis.

BACKGROUND: Computed tomography aortic valve calcium scoring (CT-AVC) holds promise for the assessment of patients with aortic stenosis (AS). We sought to establish the clinical utility of CT-AVC in an international multicenter cohort of patients. METHODS AND RESULTS: Patients with AS who underwent ECG-gated CT-AVC within 3 months of echocardiography were entered into an international, multicenter, observational registry. Optimal CT-AVC thresholds for diagnosing severe AS were determined in patients with concordant echocardiographic assessments, before being used to arbitrate disease severity in those with discordant measurements. In patients with long-term follow-up, we assessed whether CT-AVC thresholds predicted aortic valve replacement and death. In 918 patients from 8 centers (age, 77±10 years; 60% men; peak velocity, 3.88±0.90 m/s), 708 (77%) patients had concordant echocardiographic assessments, in whom CT-AVC provided excellent discrimination for severe AS (C statistic: women 0.92, men 0.89). Our optimal sex-specific CT-AVC thresholds (women 1377 Agatston unit and men 2062 Agatston unit) were nearly identical to those previously reported (women 1274 Agatston unit and men 2065 Agatston unit). Clinical outcomes were available in 215 patients (follow-up 1029 [126-2251] days). Sex-specific CT-AVC thresholds independently predicted aortic valve replacement and death (hazard ratio, 3.90 [95% confidence interval, 2.19-6.78]; P<0.001) after adjustment for age, sex, peak velocity, and aortic valve area. Among 210 (23%) patients with discordant echocardiographic assessments, there was considerable heterogeneity in CT-AVC scores, which again were an independent predictor of clinical outcomes (hazard ratio, 3.67 [95% confidence interval, 1.39-9.73]; P=0.010). CONCLUSIONS: Sex-specific CT-AVC thresholds accurately identify severe AS and provide powerful prognostic information. These findings support their integration into routine clinical practice. CLINICAL TRIAL REGISTRATION: URL: http://www.clinicaltrials.gov. Unique identifiers: NCT01358513, NCT02132026, NCT00338676, NCT00647088, NCT01679431.

Spectrum of Coronary Artery Aneurysms: From the Radiologic Pathology Archives.

Advances in medical diagnosis reveal that coronary artery aneurysms (CAAs) may develop in several clinical scenarios and manifest variable symptoms, imaging appearances, and outcomes. Aneurysms are pathologically classified into three groups: atherosclerotic, inflammatory, and noninflammatory. The last category is associated with congenital, inherited, and connective tissue disorders. Overlap exists among the groups, because secondary atherosclerotic change may be present in an aneurysm of any cause. Atherosclerosis is the most common cause of CAAs in adults, and inflammation is considered the underlying mechanism. In children, Kawasaki disease is the most likely cause of CAAs. In both conditions, the aneurysms are usually multiple and affect more than one coronary artery. Mycotic (infectious), iatrogenic, and cocaine-induced CAAs are also well documented. Most CAAs are discovered incidentally, but potential cardiovascular complications include thrombosis, occlusion, fistula formation, rupture, myocardial infarction, and cardiac tamponade. Imaging modalities to evaluate a suspected CAA include transthoracic echocardiography, angiographic cardiac catheterization, electrocardiographically gated computed tomographic angiography, cardiac magnetic resonance (MR) imaging, and MR angiography. Management is usually individualized, and options include surveillance, anticoagulant therapy, percutaneous stent or coil placement, surgical resection, and coronary artery bypass grafting.