When a test is too good: how CT pulmonary angiograms find pulmonary emboli that do not need to be found.
Ontology highlight
ABSTRACT:
SUBMITTER: Wiener RS
PROVIDER: S-EPMC4688549 | biostudies-literature | 2013 Jul
REPOSITORIES: biostudies-literature
Publications
When a test is too good: how CT pulmonary angiograms find pulmonary emboli that do not need to be found.
Wiener Renda Soylemez RS Schwartz Lisa M LM Woloshin Steven S
BMJ (Clinical research ed.) 20130702
Similar Datasets
Project description:BackgroundThe aim of this study was to develop and evaluate a deep neural network model in the automated detection of pulmonary embolism (PE) from computed tomography pulmonary angiograms (CTPAs) using only weakly labelled training data.MethodsWe developed a deep neural network model consisting of two parts: a convolutional neural network architecture called InceptionResNet V2 and a long-short term memory network to process whole CTPA stacks as sequences of slices. Two versions of the model were created using either chest X-rays (Model A) or natural images (Model B) as pre-training data. We retrospectively collected 600 CTPAs to use in training and validation and 200 CTPAs to use in testing. CTPAs were annotated only with binary labels on both stack- and slice-based levels. Performance of the models was evaluated with ROC and precision-recall curves, specificity, sensitivity, accuracy, as well as positive and negative predictive values.ResultsBoth models performed well on both stack- and slice-based levels. On the stack-based level, Model A reached specificity and sensitivity of 93.5% and 86.6%, respectively, outperforming Model B slightly (specificity 90.7% and sensitivity 83.5%). However, the difference between their ROC AUC scores was not statistically significant (0.94 vs 0.91, p = 0.07).ConclusionsWe show that a deep learning model trained with a relatively small, weakly annotated dataset can achieve excellent performance results in detecting PE from CTPAs.
| S-EPMC8919639 | biostudies-literature
Project description:PurposeSince 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.MethodsThe 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.ResultsThe 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.ConclusionsThe DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.
| S-EPMC10093638 | biostudies-literature
Project description:ObjectivesTo evaluate the interobserver agreement (IOA) between the initial radiology resident and the final staff radiologist reports of combined computed tomographic pulmonary angiograms (CTPA) and computed tomographic venograms (CTV) performed during on-call hours.Materials and methodsApproval by the institutional review board was obtained. Six-hundred and ninety-six consecutive studies (CTPA or CTPA with CTV) performed during on-call hours and interpreted by 30 residents were identified. Radiology residents' reports were compared to the final staff reports. Three tests outcomes were considered (positive, P; negative, N; indeterminate, I). Discordant cases were reviews by a chest radiologist.ResultsCTPAs were reported by staff radiologists as positive for pulmonary embolism (PE) in 18% (126/694), with a kappa of 0.81 (95% CI 0.77-0.86) with 3 outcomes (P, N, I), and a kappa of 0.89 (95% CI 0.85-0.94) with 2 outcomes (P, N). Regarding PE location, good concordance was observed for positive studies, with a kappa of 0.86 (95% CI 0.78 - 0.95). CTVs were reported as positive by staff radiologists in 8.5% (33/388), with a kappa of 0.66 (95% CI 0.55-0.77) with 3 outcomes (P, N, I), and a kappa of 0.89 (95% CI 0.8-1.0) with 2 outcomes (P, N). The IOA between residents and staff radiologists increased with increasing residency year level for CTPAs, but did not for CTVs.ConclusionsVery good and good IOA were observed between resident and staff radiologist interpretations for CTPA and CTV, respectively, with tendency towards improved IOA as residency level of training increased for CTPA, but not for CTV.
| S-EPMC4418836 | biostudies-literature
Project description:In patients who present with pulmonary embolism, right-heart thrombus is a rare condition that is associated with increased mortality rates, compared with pulmonary embolism alone. Thrombolytic therapy has been associated with a survival benefit in previous studies of pulmonary embolism arising from right-heart thrombus. However, older patients have been excluded from such studies because thrombolysis places them at excessively high risk of bleeding. We present a case, in a 92-year-old woman, of pulmonary embolism arising from right-heart thrombi that we successfully treated with heparin.
| S-EPMC3783120 | biostudies-literature
Project description:Interactions between individuals and the structure of their environment play a crucial role in shaping self-organized collective behaviors. Recent studies have shown that ants crossing asymmetrical bifurcations in a network of galleries tend to follow the branch that deviates the least from their incoming direction. At the collective level, the combination of this tendency and the pheromone-based recruitment results in a greater likelihood of selecting the shortest path between the colony's nest and a food source in a network containing asymmetrical bifurcations. It was not clear however what the origin of this behavioral bias is. Here we propose that it results from a simple interaction between the behavior of the ants and the geometry of the network, and that it does not require the ability to measure the angle of the bifurcation. We tested this hypothesis using groups of ant-like robots whose perceptual and cognitive abilities can be fully specified. We programmed them only to lay down and follow light trails, avoid obstacles and move according to a correlated random walk, but not to use more sophisticated orientation methods. We recorded the behavior of the robots in networks of galleries presenting either only symmetrical bifurcations or a combination of symmetrical and asymmetrical bifurcations. Individual robots displayed the same pattern of branch choice as individual ants when crossing a bifurcation, suggesting that ants do not actually measure the geometry of the bifurcations when travelling along a pheromone trail. Finally at the collective level, the group of robots was more likely to select one of the possible shorter paths between two designated areas when moving in an asymmetrical network, as observed in ants. This study reveals the importance of the shape of trail networks for foraging in ants and emphasizes the underestimated role of the geometrical properties of transportation networks in general.
| S-EPMC3610605 | biostudies-literature
Project description:Capturing a paradoxical embolism in real-time has been a challenge in recent literature. We present the unique case of a 33-year-old, G3P2 female at 8 weeks gestation presenting with dyspnea. An active thrombus through an undiagnosed patent foramen ovale was found requiring emergent surgical intervention with a positive outcome. The presence of a deep vein thrombosis, inferior vena caval thrombus, patent foramen ovale, and pulmonary artery thrombi was contemporarily documented. To our knowledge, there is minimal literature with this presentation.
| S-EPMC9732962 | biostudies-literature
Project description:This case report discusses a previously healthy young male who suffered an in hospital cardiac arrest, found to be secondary to massive pulmonary emboli (PE). This patient was not a candidate for PE dosing of tPA, but continued to have worsening cardiogenic shock and acute hypoxic respiratory failure. The decision was made to initiate venoarterial extracorporeal membrane oxygenation (VA ECMO) in combination with catheter assisted thrombolysis. With time, the sequelae of his disease process changed. Although his cardiogenic shock improved, the hypoxemia worsened, and the patient was transitioned from VA to venovenous (VV) ECMO, while managing further complications due to multiple episodes of cardiac arrest and severe coagulopathy with multiple, acute, life-threatening bleeding events. Albeit a growing field, VA and VV ECMO are less than common, individually, however this case reports outlines the even rarer event of transition from VA to VV ECMO in combination with catheter assisted thrombolysis treatment via the EKOS EndoWave Infusion Catheter System. Providing care for critically ill patients often demands a collaborative effort between multiple specialties. This case report exemplifies this collaboration, leading to the combination and unification of alternative treatments: ECMO and catheter assisted thrombolysis.
| S-EPMC5986164 | biostudies-literature
Project description:Pulmonary embolism occurs when blood flow to a part of the lungs is blocked by a venous thrombus that has traveled from the lower limbs. Little is known about the mechanical behavior of emboli under compressive forces from the surrounding musculature and blood pressure. We measured the stress-strain responses of human pulmonary emboli under cyclic compression, and showed that emboli exhibit a hysteretic stress-strain curve. The fibrin fibers and red blood cells (RBCs) are damaged during the compression process, causing irreversible changes in the structure of the emboli. We showed using electron and confocal microscopy that bundling of fibrin fibers occurs due to compression, and damage is accumulated as more cycles are applied. The stress-strain curves depend on embolus structure, such that variations in composition give quantitatively different responses. Emboli with a high fibrin component demonstrate higher normal stress compared to emboli that have a high RBC component. We compared the compression response of emboli to that of whole blood clots containing various volume fractions of RBCs, and found that RBCs rupture at a certain critical stress. We describe the hysteretic response characteristic of foams, using a model of phase transitions in which the compressed foam is segregated into coexisting rarefied and densified phases whose fractions change during compression. Our model takes account of the rupture of RBCs in the compressed emboli and stresses due to fluid flow through their small pores. Our results can help in classifying emboli as rich in fibrin or rich in red blood cells, and can help in understanding what responses to expect when stresses are applied to thrombi in vivo.
| S-EPMC7241098 | biostudies-literature
Project description:Blood clots and thrombi undergo platelet-driven contraction/retraction followed by structural rearrangements. We have established quantitative relationships between the composition of blood clots and extent of contraction to determine intravital contraction of thrombi and emboli based on their content. The composition of human blood clots and thrombi was quantified using histology and scanning electron microscopy. Contracting blood clots were segregated into the gradually shrinking outer layer that contains a fibrin-platelet mesh and the expanding inner portion with compacted red blood cells (RBCs). At 10% contraction, biconcave RBCs were partially compressed into polyhedral RBCs, which became dominant at 20% contraction and higher. The polyhedral/biconcave RBC ratio and the extent of contraction displayed an exponential relationship, which was used to determine the extent of intravital contraction of ex vivo thrombi, ranging from 30% to 50%. In venous thrombi, the extent of contraction decreased gradually from the older (head) to the younger (body, tail) parts. In pulmonary emboli, the extent of contraction was significantly lower than in the venous head but was similar to the body and tail, suggesting that the emboli originate from the younger portion(s) of venous thrombi. The extent of contraction in arterial cerebral thrombi was significantly higher than in the younger parts of venous thrombi (body, tail) and pulmonary emboli but was indistinguishable from the older part (head). A novel tool, named the "contraction ruler," has been developed to use the composition of ex vivo thrombi to assess the extent of their intravital contraction, which contributes to the pathophysiology of thromboembolism.
| S-EPMC8941457 | biostudies-literature
Project description:Finding an antibody that works for a specific application can be a difficult task. Hundreds of vendors offer millions of antibodies, but the quality of these products and available validation information varies greatly. In addition, several studies have called into question the reliability of published data as the primary metric for assessing antibody quality. We briefly discuss the antibody quality problem and provide best practice guidelines for selecting and validating an antibody, as well as for publishing data generated using antibodies.
| S-EPMC5499787 | biostudies-other