Newer modalities for detection of pulmonary emboli

Deep Learning-Based Algorithm for Automatic Detection of Pulmonary Embolism in Chest CT Angiograms

Purpose: Since the prompt recognition of acute pulmonary embolism (PE) and the immediate initiation of treatment can significantly reduce the risk of death, we developed a deep learning (DL)-based application aimed to automatically detect PEs on chest computed tomography angiograms (CTAs) and alert radiologists for an urgent interpretation. Convolutional neural networks (CNNs) were used to design the application. The associated algorithm used a hybrid 3D/2D UNet topology. The training phase was performed on datasets adequately distributed in terms of vendors, patient age, slice thickness, and kVp. The objective of this study was to validate the performance of the algorithm in detecting suspected PEs on CTAs. Methods: The validation dataset included 387 anonymized real-world chest CTAs from multiple clinical sites (228 U.S. cities). The data were acquired on 41 different scanner models from five different scanner makers. The ground truth (presence or absence of PE on CTA images) was established by three independent U.S. board-certified radiologists. Results: The algorithm correctly identified 170 of 186 exams positive for PE (sensitivity 91.4% [95% CI: 86.4–95.0%]) and 184 of 201 exams negative for PE (specificity 91.5% [95% CI: 86.8–95.0%]), leading to an accuracy of 91.5%. False negative cases were either chronic PEs or PEs at the limit of subsegmental arteries and close to partial volume effect artifacts. Most of the false positive findings were due to contrast agent-related fluid artifacts, pulmonary veins, and lymph nodes. Conclusions: The DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.

A deep learning approach for automated diagnosis of pulmonary embolism on computed tomographic pulmonary angiography

Background

Computed tomographic pulmonary angiography (CTPA) is the diagnostic standard for confirming pulmonary embolism (PE). Since PE is a life-threatening condition, early diagnosis and treatment are critical to avoid PE-associated morbidity and mortality. However, PE remains subject to misdiagnosis.

Methods

We retrospectively identified 251 CTPAs performed at a tertiary care hospital between January 2018 to January 2021. The scans were classified as positive (n = 55) and negative (n = 196) for PE based on the annotations made by board-certified radiologists. A fully anonymized CT slice served as input for the detection of PE by the 2D segmentation model comprising U-Net architecture with Xception encoder. The diagnostic performance of the model was calculated at both the scan and the slice levels.

Results

The model correctly identified 44 out of 55 scans as positive for PE and 146 out of 196 scans as negative for PE with a sensitivity of 0.80 [95% CI 0.68, 0.89], a specificity of 0.74 [95% CI 0.68, 0.80], and an accuracy of 0.76 [95% CI 0.70, 0.81]. On slice level, 4817 out of 5183 slices were marked as positive for the presence of emboli with a specificity of 0.89 [95% CI 0.88, 0.89], a sensitivity of 0.93 [95% CI 0.92, 0.94], and an accuracy of 0.89 [95% CI 0.887, 0.890]. The model also achieved an AUROC of 0.85 [0.78, 0.90] and 0.94 [0.936, 0.941] at scan level and slice level, respectively for the detection of PE.

Conclusion

The development of an AI model and its use for the identification of pulmonary embolism will support healthcare workers by reducing the rate of missed findings and minimizing the time required to screen the scans.

Primary diagnosis of pulmonary embolism with unenhanced MRI for patients not eligible for CTPA: Clinical outcome

Purpose

To follow up the clinical outcome of patients with suspected pulmonary embolism (PE), in those only imaged using unenhanced, free-breathing magnetic resonance imaging (MRI).

Methods and materials

Fifty-seven patients aged 29-99 years (mean 70, SD 18) that could not undergo Computed Tomography Pulmonary Angiography (CTPA) were offered alternative imaging diagnostics in parallel with ongoing methodological studies validating MRI vs CTPA. Contraindications included renal failure (n = 44), severe iodine contrast allergy (n = 10), pregnancy (n = 2) and radioactive iodine therapy (n = 1). The unenhanced MRI protocol was based on free-breathing, steady-state free precession with no cardiac or respiratory gating. Retrospective review of the electronic medical record (EMR) was made of 0-12 months post-imaging and was collected during 2012-2018.

Results

All 57 MRIs were of diagnostic quality and 12 pulmonary embolisms were diagnosed. Of the 57 patients, 44 were already on, or had started anticoagulation therapy due to clinical suspicion of PE. Four of the patients were put on anticoagulation after the positive MRI and 13 were taken off anticoagulation after a negative MRI report. Other diagnoses reported (considering dyspnea) were pleural effusion (n = 24), consolidation (n = 12) and pericardial effusion (n = 2). One patient had a deep vein thrombosis (DVT) within three months of our negative MRI result and then had a stroke within one year. Another patient suffered a stroke within three months of being diagnosed (by MRI) with PE and given anticoagulation as treatment.

Conclusions

Our method supported or altered clinical decision-making and treatment in this cohort. A diagnostic tool for PE without intravenous contrast agent or radiation is of great benefit for certain patients.

Detection of pulmonary embolism using repeated MRI acquisitions without respiratory gating: a preliminary study

Background Pulmonary embolism (PE) is a severe medical condition with non-specific clinical findings. Computed tomography angiography (CTA) using iodinated contrast agents is the golden standard for diagnosis, but many patients have contraindications for CTA. Purpose To investigate the diagnostic accuracy of repeated acquisitions of magnetic resonance imaging (MRI), without respiratory gating or breath holding, in diagnosing PE using CTA as the reference standard. Material and Methods Thirty-three patients with clinically suspected PE underwent MRI within 48 h after diagnostic CTA. A control group of 37 healthy participants underwent MRI and was matched with an equal number of negative CTA exams. The MRI protocol was based on free-breathing steady-state free precession producing 4.5 mm slices in axial, sagittal, and coronal planes. Instead of respiratory or cardiac gating five repetitive slices were obtained in each anatomical position to compensate for movement and artifacts. Clinical assessment including d-dimer and Well's score was performed prior to imaging. One radiologist reviewed the CTA exams and two radiologists reviewed the MRI scans. Results All 70 MRI exams were of diagnostic quality and the total acquisition time for each MRI scan was 9 min 34 s. On CTA, 29 patients were diagnosed with PE and the MRI readers detected 26 and 27 of those, respectively. Specificity was 100% for both readers. Sensitivity was 90% and 93%, respectively. Inter-reader agreement using Cohen's kappa was 0.97. Conclusion Our unenhanced MRI protocol shows a high sensitivity and specificity for PE, but further studies are required before considering it as a safe diagnostic test.

Priority-driven plan optimization in locally advanced lung patients based on perfusion SPECT imaging

Purpose

Limits on mean lung dose (MLD) allow for individualization of radiation doses at safe levels for patients with lung tumors. However, MLD does not account for individual differences in the extent or spatial distribution of pulmonary dysfunction among patients, which leads to toxicity variability at the same MLD. We investigated dose rearrangement to minimize the radiation dose to the functional lung as assessed by perfusion single photon emission computed tomography (SPECT) and maximize the target coverage to maintain conventional normal tissue limits.

Methods and materials

Retrospective plans were optimized for 15 patients with locally advanced non-small cell lung cancer who were enrolled in a prospective imaging trial. A staged, priority-based optimization system was used. The baseline priorities were to meet physical MLD and other dose constraints for organs at risk, and to maximize the target generalized equivalent uniform dose (gEUD). To determine the benefit of dose rearrangement with perfusion SPECT, plans were reoptimized to minimize the generalized equivalent uniform functional dose (gEUfD) to the lung as the subsequent priority.

Results

When only physical MLD is minimized, lung gEUfD was 12.6 ± 4.9 Gy (6.3-21.7 Gy). When the dose is rearranged to minimize gEUfD directly in the optimization objective function, 10 of 15 cases showed a decrease in lung gEUfD of >20% (lung gEUfD mean 9.9 ± 4.3 Gy, range 2.1-16.2 Gy) while maintaining equivalent planning target volume coverage. Although all dose-limiting constraints remained unviolated, the dose rearrangement resulted in slight gEUD increases to the cord (5.4 ± 3.9 Gy), esophagus (3.0 ± 3.7 Gy), and heart (2.3 ± 2.6 Gy).

Conclusions

Priority-driven optimization in conjunction with perfusion SPECT permits image guided spatial dose redistribution within the lung and allows for a reduced dose to the functional lung without compromising target coverage or exceeding conventional limits for organs at risk.

Contrast enhanced pulmonary magnetic resonance angiography for pulmonary embolism: Building a successful program

The performance of contrast enhanced pulmonary magnetic resonance angiography (MRA) for the diagnosis of pulmonary embolism (PE) is an effective non-ionizing alternative to contrast enhanced computed tomography and nuclear medicine ventilation/perfusion scanning. However, the technical success of these exams is very dependent on careful attention to the details of the MRA acquisition protocol and requires reader familiarity with MRI and its artifacts. Most practicing radiologists are very comfortable with the performance and interpretation of computed tomographic angiography (CTA) performed to detect pulmonary embolism but not all are as comfortable with the use of MRA in this setting. The purpose of this review is to provide the general radiologist with the tools necessary to build a successful pulmonary embolism MRA program. This review will cover in detail image acquisition, image interpretation, and some key elements of outreach that help to frame the role of MRA to consulting clinicians and hospital administrators. It is our aim that this resource will help build successful clinical pulmonary embolism MRA programs that are well received by patients and physicians, reduce the burden of medical imaging radiation, and maintain good patient outcomes.

Myocardial infarction with angiographically normal coronary arteries

Myocardial Infarction with Normal Coronary Arteries (MINCA) is an important subgroup of myocardial infarction with a frequency of at least 3-4% of all myocardial infarctions. The interest and awareness of MINCA have increased recently due to the frequent use of coronary angiography, the description of Takotsubo stress cardiomyopathy and new sensitive troponin assays. Since myocarditis may mimic myocardial infarction it is essential to exclude this in patients with myocardial infarction with angiographically normal coronary arteries. Cardiac magnetic resonance imaging is a cornerstone not only to establish the diagnosis but also an important tool in the search for different causes of myocardial damage. In the future, atherosclerotic burden, hemostatic function, characterization of stressors and inflammation will be important targets for research in this group of patients.

Magnetic resonance imaging and computed tomography developments in imaging of venous thromboembolism

Venous thromboembolism (VTE) is a disease that causes high morbidity and mortality in the population. At present the first-line imaging test for a suspected pulmonary embolism (PE) is computed tomography (CT) pulmonary angiography, and ultrasonography is widely used for the diagnosis of deep-vein thrombosis (DVT). Although these modalities are proven to be safe and accurate, unresolved issues remain, such as whether CT scanning in patients with a suspected PE should be extended to the legs. Another issue is the diagnosis of recurrent DVT. Magnetic resonance imaging (MRI) offers a number of advantages in the imaging of VTE. Recent developments of scanning protocols with shorter acquisition times, sometimes complemented by navigator gating or making use of endogenous contrast, offer new perspectives for the use of MRI. This review provides an overview of state of the art MRI techniques for the diagnosis of PE and DVT. Furthermore, the use of new contrast agents such as fibrin labeling to detect thrombi are addressed.

Clinical prediction of pulmonary embolism in respiratory emergencies

INTRODUCTION

The initial management of suspected pulmonary embolism (PE) is commonly done in respiratory departments, but is based on clinical prediction rules developed in other settings.

OBJECTIVE

To determine the accuracy of established prediction rules for PE in patients with respiratory emergencies.

DESIGN

A prospective study

MATERIALS AND METHODS

Patients presenting to respiratory emergency department with acute symptoms and signs suggestive of PE (n=183) and subsequently admitted to hospital were prospectively enrolled. Wells' rule, original and revised Geneva scores, their components separately, and other common clinical parameters were recorded during admission. PE was diagnosed by perfusion lung scanning, computed tomographic pulmonary angiography, lower limb venous ultrasonography, magnetic resonance pulmonary angiography, and/or pulmonary angiography.

RESULTS

PE was confirmed in 52 and ruled out in 131 patients. Tachycardia, atelectasis, elevated hemidiaphragm, clinical signs of deep-venous thrombosis, physician perception that PE is the likeliest diagnosis, previous thromboembolism, chest pain, and absence of chronic obstructive pulmonary disease or cough were associated with the presence of PE. These significant parameters could be combined for accurate pre-test PE prediction, with a newly devised combinatorial tool exhibiting the highest area under curve [0.92 (95% CI: 0.87-0.97)], followed by Wells' rule [0.86 (95% CI 0.79-0.92)], the revised Geneva score [0.83 (95% CI 0.77-0.90)], and the original Geneva score [0.75 (95% CI 0.68-0.83)].

CONCLUSION

Wells' rule and the revised Geneva score are more useful in diagnosing PE in respiratory emergencies. A newly devised prediction tool can be of even greater accuracy in this patient population.

Cost-effective diagnostic strategies in patients with a high, intermediate, or low clinical probability of pulmonary embolism

Rapid quantitative D-dimer assays (DD), lower extremity venous duplex ultrasonography (US), and multislice computed tomographic (CT) angiography have been shown to have adequate sensitivities and specificities for diagnostic purpose. The purpose of this study was to evaluate cost-effectiveness of diagnostic strategies for pulmonary embolism (PE) in patients with a high, intermediate, or low clinical probability of PE. A formal cost-effectiveness analysis for the diagnosis of PE was performed. The main outcome measure for effectiveness was 3-month expected survival. The strategy of DD followed by CT was cost-effective and had the lowest cost per life saved for all patients suspected with PE. The conventional strategy including ventilation and perfusion lung scanning followed by pulmonary angiography (PA) or CT was not cost-effective. The leg US after CT was not also cost-effective. In clinical practice, the individual patient's condition should be considered when choosing appropriate diagnostic tests.

Impact and clinical significance of recurrent venous thromboembolism

Background

The purpose of this review was to analyse current knowledge and controversies associated with the diagnosis, treatment and prevention of recurrent venous thromboembolism (VTE).

Methods

MEDLINE and manual searches were performed to select prospective papers on the diagnosis, treatment and prevention of recurrent VTE for their relevance and quality.

Results

The cumulative incidence of recurrent VTE increases from 11 per cent at 1 year to 40 per cent at 10 years. The incidence of recurrence is higher in unprovoked thrombosis compared with provoked VTE. Patients with unprovoked deep vein thrombosis also have a greater number of multiple recurrences. Ultrasonography or D-dimer monitoring may have an impact on the duration of anticoagulation but further refinements are needed. The incidence of skin damage is higher in ipsilateral recurrence compared with contralateral or no recurrence. Legs with ipsilateral recurrence more often have both reflux and obstruction.

Conclusion

The role and weight of the predictive factors for recurrent VTE and its sequelae, and the type and optimal duration of anticoagulation have not been studied adequately. Fatality associated with pulmonary embolism and rates of recurrent VTE remain unacceptably high.

Imaging presentation of venous thrombosis in patients with cancer

OBJECTIVE

The purpose of this article is to review the imaging of venous thrombosis in patients with cancer.

CONCLUSION

Multiple imaging techniques have the capacity to display thrombosis accurately. The optimal choice is dictated by the location and duration of symptoms and by the availability of imaging techniques. Peripheral and superficial thrombi are best managed with ultrasound, whereas central thrombi require CT or MRI. If CT and MRI are contraindicated, flow studies are appropriate. FDG PET/CT appropriately shows venous thrombosis and might play a prominent role in the future.