Deep learning for pulmonary embolism detection on computed tomography pulmonary angiogram: a systematic review and meta-analysis.
Ontology highlight
ABSTRACT:
SUBMITTER: Soffer S
PROVIDER: S-EPMC8338977 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Differences between computed tomography pulmonary angiography and ventilation-perfusion lung scanning in pregnant patients with suspected acute pulmonary embolism are not well-known, leading to ongoing debate on which test to choose. We searched in PubMed, EMBASE, Web of Science and the Cochrane Library databases and identified all relevant articles and abstracts published up to October 1, 2017. We assessed diagnostic efficiency, frequency of non-diagnostic results and maternal and fetal exposure to radiation exposure. We included 13 studies for the diagnostic efficiency analysis, 30 for the analysis of non-diagnostic results and 22 for the radiation exposure analysis. The pooled rate of false negative test results was 0% for both imaging strategies with overlapping confidence intervals. The pooled rates of non-diagnostic results with computed tomography pulmonary angiography and ventilation-perfusion lung scans were 12% (95% confidence interval: 8-17) and 14% (95% confidence interval: 10-18), respectively. Reported maternal and fetal radiation exposure doses were well below the safety threshold, but could not be compared between the two diagnostic methods given the lack of high quality data. Both imaging tests seem equally safe to rule out pulmonary embolism in pregnancy. We found no significant differences in efficiency and radiation exposures between computed tomography pulmonary angiography and ventilation-perfusion lung scanning although direct comparisons were not possible.
| S-EPMC6312023 | biostudies-literature
Project description:The lack of publicly available datasets of computed-tomography angiography (CTA) images for pulmonary embolism (PE) is a problem felt by physicians and researchers. Although a number of computer-aided detection (CAD) systems have been developed for PE diagnosis, their performance is often evaluated using private datasets. In this paper, we introduce a new public dataset called FUMPE (standing for Ferdowsi University of Mashhad's PE dataset) which consists of three-dimensional PE-CTA images of 35 different subjects with 8792 slices in total. For each benchmark image, two expert radiologists provided the ground-truth with the assistance of a semi-automated image processing software tool. FUMPE is a challenging benchmark for CAD methods because of the large number (i.e., 3438) of PE regions and, more especially, because of the location of most of them (i.e., 67%) in lung peripheral arteries. Moreover, due to the reporting of the Qanadli score for each PE-CTA image, FUMPE is the first public dataset which can be used for the analysis of mortality and morbidity risks associated with PE. We also report some complementary prognosis information for each subject.
| S-EPMC6122162 | biostudies-literature
Project description:BackgroundRadiologic evidence of air trapping (AT) on expiratory computed tomography (CT) scans is associated with early pulmonary dysfunction in patients with cystic fibrosis (CF). However, standard techniques for quantitative assessment of AT are highly variable, resulting in limited efficacy for monitoring disease progression.ObjectiveTo investigate the effectiveness of a convolutional neural network (CNN) model for quantifying and monitoring AT, and to compare it with other quantitative AT measures obtained from threshold-based techniques.Materials and methodsPaired volumetric whole lung inspiratory and expiratory CT scans were obtained at four time points (0, 3, 12 and 24 months) on 36 subjects with mild CF lung disease. A densely connected CNN (DN) was trained using AT segmentation maps generated from a personalized threshold-based method (PTM). Quantitative AT (QAT) values, presented as the relative volume of AT over the lungs, from the DN approach were compared to QAT values from the PTM method. Radiographic assessment, spirometric measures, and clinical scores were correlated to the DN QAT values using a linear mixed effects model.ResultsQAT values from the DN were found to increase from 8.65% ± 1.38% to 21.38% ± 1.82%, respectively, over a two-year period. Comparison of CNN model results to intensity-based measures demonstrated a systematic drop in the Dice coefficient over time (decreased from 0.86 ± 0.03 to 0.45 ± 0.04). The trends observed in DN QAT values were consistent with clinical scores for AT, bronchiectasis, and mucus plugging. In addition, the DN approach was found to be less susceptible to variations in expiratory deflation levels than the threshold-based approach.ConclusionThe CNN model effectively delineated AT on expiratory CT scans, which provides an automated and objective approach for assessing and monitoring AT in CF patients.
| S-EPMC7990199 | biostudies-literature
Project description:Background Hypodense filling defects within the pulmonary veins on computed tomography described as pulmonary vein sign (PVS) have been noted in acute pulmonary embolism and shown to be associated with poor prognosis. We evaluated venous flow abnormalities in chronic thromboembolic pulmonary hypertension (CTEPH) to determine its usefulness in the computed tomography assessment of CTEPH. Methods and Results Blinded retrospective computed tomography analysis of 50 proximal CTEPH cases and 3 control groups-50 acute pulmonary embolism, 50 nonthromboembolic cohort, and 50 pulmonary arterial hypertension. Venous flow reduction was assessed by the following: (1) presence of a filling defect of at least 2 cm in a pulmonary vein draining into the left atrium, and (2) left atrium attenuation (>160 Hounsfield units). PVS was most prevalent in CTEPH. Compared with all controls, sensitivity and specificity of PVS for CTEPH is 78.0% and 85.3% (95% CI, 64.0-88.5 and 78.6-90.6, respectively) versus 34.0% and 70.7% (95% CI, 21.2-48.8 and 62.7-77.8) in acute pulmonary embolism, 8.0% and 62% (95% CI, 2.2-19.2 and 53.7-69.8) in nonthromboembolic and 2.0% and 60% (95% CI, 0.1-10.7 and 51.7-67.9) in pulmonary arterial hypertension. In CTEPH, lobar and segmental arterial occlusive disease was most commonly associated with corresponding absent venous flow. PVS detection was highly reproducible (Kappa=0.96, 95% CI, 0.90-1.01, P<0.001). Conclusions PVS is easy to detect with higher sensitivity and specificity in CTEPH compared with acute pulmonary embolism and is not a feature of pulmonary arterial hypertension. Asymmetric enhancement of pulmonary veins may serve as an additional parameter in the computed tomography assessment of CTEPH and can be used to differentiate CTEPH from pulmonary arterial hypertension.
| S-EPMC7763423 | biostudies-literature
Project description:(1) Background: Computed tomography pulmonary angiography (CTPA) is the standard imaging test for the evaluation of acute pulmonary embolism (PE), but it is associated with patients' exposure to radiation. Studies have suggested that radiation exposure can be reduced without compromising PE detection by limiting the scan range (the z-axis, going from up to down); (2) Methods: A literature search was conducted in MEDLINE and EMBASE on 17 July 2021. Studies were included if they enrolled patients who had undergone a CTPA and described the yield of PE diagnoses, number of missed filling defects and/or other diagnoses using a reduced z-axis in comparison to a full-length scan. To assess risk of bias, we modified an existing risk of bias tools for observational studies, the Newcastle-Ottawa Scale. Results were synthesized in a narrative review. Primary outcomes were the number of missed PE diagnoses (based on at least one filling defect) and filling defects; the secondary outcome was the number of other missed findings; (3) Results: Eleven cohort studies and one case-control study were included reporting on a total of 3955 scans including 1025 scans with a diagnosis of PE. Six different reduced scan ranges were assessed; the most studied was from the top of the aortic arch to below the heart, in which no PEs were missed (seven studies). One sub-segmental PE was missed when the scan coverage was 10 cm starting from the bottom of the aortic arch and 14.7 cm starting from the top of the arch. Five studies that reported on other findings all found that other diagnoses were missed with a reduced z-axis. Most of the included studies had a high risk of bias; (4) Conclusions: CTPA scan coverage reduction from the top of aortic arch to below the heart reduced radiation exposure without affecting PE diagnoses, but studies were generally at high risk of bias.
| S-EPMC8700432 | biostudies-literature
Project description:Background The association of pulmonary embolism (PE) with deep vein thrombosis (DVT) in patients with coronavirus disease 2019 (COVID-19) remains unclear, and the diagnostic accuracy of D-dimer tests for PE is unknown. Purpose To conduct meta-analysis of the study-level incidence of PE and DVT and to evaluate the diagnostic accuracy of D-dimer tests for PE from multicenter individual patient data. Materials and Methods A systematic literature search identified studies evaluating the incidence of PE or DVT in patients with COVID-19 from January 1, 2020, to June 15, 2020. These outcomes were pooled using a random-effects model and were further evaluated using metaregression analysis. The diagnostic accuracy of D-dimer tests for PE was estimated on the basis of individual patient data using the summary receiver operating characteristic curve. Results Twenty-seven studies with 3342 patients with COVID-19 were included in the analysis. The pooled incidence rates of PE and DVT were 16.5% (95% CI: 11.6, 22.9; I2 = 0.93) and 14.8% (95% CI: 8.5, 24.5; I2 = 0.94), respectively. PE was more frequently found in patients who were admitted to the intensive care unit (ICU) (24.7% [95% CI: 18.6, 32.1] vs 10.5% [95% CI: 5.1, 20.2] in those not admitted to the ICU) and in studies with universal screening using CT pulmonary angiography. DVT was present in 42.4% of patients with PE. D-dimer tests had an area under the receiver operating characteristic curve of 0.737 for PE, and D-dimer levels of 500 and 1000 ?g/L showed high sensitivity (96% and 91%, respectively) but low specificity (10% and 24%, respectively). Conclusion Pulmonary embolism (PE) and deep vein thrombosis (DVT) occurred in 16.5% and 14.8% of patients with coronavirus disease 2019 (COVID-19), respectively, and more than half of patients with PE lacked DVT. The cutoffs of D-dimer levels used to exclude PE in preexisting guidelines seem applicable to patients with COVID-19. © RSNA, 2020 Supplemental material is available for this article. See also the editorial by Woodard in this issue.
| S-EPMC7745997 | biostudies-literature
Project description:ObjectivesDeep learning (DL) has been increasingly employed for automated landmark detection, e.g., for cephalometric purposes. We performed a systematic review and meta-analysis to assess the accuracy and underlying evidence for DL for cephalometric landmark detection on 2-D and 3-D radiographs.MethodsDiagnostic accuracy studies published in 2015-2020 in Medline/Embase/IEEE/arXiv and employing DL for cephalometric landmark detection were identified and extracted by two independent reviewers. Random-effects meta-analysis, subgroup, and meta-regression were performed, and study quality was assessed using QUADAS-2. The review was registered (PROSPERO no. 227498).DataFrom 321 identified records, 19 studies (published 2017-2020), all employing convolutional neural networks, mainly on 2-D lateral radiographs (n=15), using data from publicly available datasets (n=12) and testing the detection of a mean of 30 (SD: 25; range.: 7-93) landmarks, were included. The reference test was established by two experts (n=11), 1 expert (n=4), 3 experts (n=3), and a set of annotators (n=1). Risk of bias was high, and applicability concerns were detected for most studies, mainly regarding the data selection and reference test conduct. Landmark prediction error centered around a 2-mm error threshold (mean; 95% confidence interval: (-0.581; 95 CI: -1.264 to 0.102 mm)). The proportion of landmarks detected within this 2-mm threshold was 0.799 (0.770 to 0.824).ConclusionsDL shows relatively high accuracy for detecting landmarks on cephalometric imagery. The overall body of evidence is consistent but suffers from high risk of bias. Demonstrating robustness and generalizability of DL for landmark detection is needed.Clinical significanceExisting DL models show consistent and largely high accuracy for automated detection of cephalometric landmarks. The majority of studies so far focused on 2-D imagery; data on 3-D imagery are sparse, but promising. Future studies should focus on demonstrating generalizability, robustness, and clinical usefulness of DL for this objective.
| S-EPMC8310492 | biostudies-literature
Project description:This study aimed to determine the diagnostic accuracy of computed tomography imaging for the diagnosis of chronic thromboembolic pulmonary hypertension (CTEPH). Additionally, the effect of test and study characteristics was explored. Studies published between 1990 and 2015 identified by PubMed, OVID search and citation tracking were examined. Of the 613 citations, 11 articles (n=712) met the inclusion criteria. The patient-based analysis demonstrated a pooled sensitivity of 76% (95% confidence interval [CI]: 69% to 82%), and a pooled specificity of 96% (95%CI: 93% to 98%). This resulted in a pooled diagnostic odds ratio (DOR) of 191 (95%CI: 75 to 486). The vessel-based analyses were divided into 3 levels: total arteries?main+ lobar arteries and segmental arteries. The pooled sensitivity were 88% (95%CI: 87% to 90%)?95% (95%CI: 92% to 97%) and 88% (95%CI: 87% to 90%), respectively, with a pooled specificity of 90% (95%CI: 88% to 91%)?96% (95%CI: 94% to 97%) and 89% (95% CI: 87% to 91%). This resulted in a pooled diagnostic odds ratio of 76 (95%CI: 23 to 254),751 (95%CI: 57 to 9905) and 189 (95%CI: 21 to 1072), respectively. In conclusion, CT is a favorable method to rule in CTEPH and to rule out pulmonary endarterectomy (PEA) patients for proximal branches. Furthermore, dual-energy and 320-slices CT can increase the sensitivity for subsegmental arterials, which are promising imaging techniques for balloon pulmonary angioplasty (BPA) approach. In the near future, CT could position itself as the key for screening consideration and for surgical and interventional operability.
| S-EPMC4414539 | biostudies-literature
Project description:Background COVID-19 appears to be associated with a high risk of venous thromboembolism (VTE). We aimed to systematically review and meta-analyze the risk of clinically relevant VTE in patients hospitalized for COVID-19. Methods This meta-analysis included original articles in English published from January 1st, 2020 to June 15th, 2020 in Pubmed/MEDLINE, Embase, Web of science, and Cochrane. Outcomes were major VTE, defined as any objectively diagnosed pulmonary embolism (PE) and/or proximal deep vein thrombosis (DVT). Primary analysis estimated the risk of VTE, stratified by acutely and critically ill inpatients. Secondary analyses explored the separate risk of proximal DVT and of PE; the risk of major VTE stratified by screening and by type of anticoagulation. Results In 33 studies (n?=?4009 inpatients) with heterogeneous thrombotic risk factors, VTE incidence was 9% (95%CI 5–13%, I2 =?92.5) overall, and 21% (95%CI 14–28%, I2 =?87.6%) for patients hospitalized in the ICU. Proximal lower limb DVT incidence was 3% (95%CI 1–5%, I2 =?87.0%) and 8% (95%CI 3–14%, I2 =?87.6%), respectively. PE incidence was 8% (95%CI 4–13%, I2 =?92.1%) and 17% (95%CI 11–25%, I2 =?89.3%), respectively. Screening and absence of anticoagulation were associated with a higher VTE incidence. When restricting to medically ill inpatients, the VTE incidence was 2% (95%CI 0–6%). Conclusions The risk of major VTE among COVID-19 inpatients is high but varies greatly with severity of the disease. These findings reinforce the need for the use of thromboprophylaxis in all COVID-19 inpatients and for clinical trials testing different thromboprophylaxis regimens in subgroups of COVID-19 inpatients. Trial registration The review protocol was registered in PROSPERO International Prospective Register of Systematic Reviews (CRD42020193369). Supplementary Information The online version contains supplementary material available at 10.1186/s12959-021-00266-x.
| S-EPMC7942819 | biostudies-literature
Project description:ObjectiveTo assess the diagnostic performance of a deep learning-based algorithm for automated detection of acute and chronic rib fractures on whole-body trauma CT.Materials and methodsWe retrospectively identified all whole-body trauma CT scans referred from the emergency department of our hospital from January to December 2018 (n = 511). Scans were categorized as positive (n = 159) or negative (n = 352) for rib fractures according to the clinically approved written CT reports, which served as the index test. The bone kernel series (1.5-mm slice thickness) served as an input for a detection prototype algorithm trained to detect both acute and chronic rib fractures based on a deep convolutional neural network. It had previously been trained on an independent sample from eight other institutions (n = 11455).ResultsAll CTs except one were successfully processed (510/511). The algorithm achieved a sensitivity of 87.4% and specificity of 91.5% on a per-examination level [per CT scan: rib fracture(s): yes/no]. There were 0.16 false-positives per examination (= 81/510). On a per-finding level, there were 587 true-positive findings (sensitivity: 65.7%) and 307 false-negatives. Furthermore, 97 true rib fractures were detected that were not mentioned in the written CT reports. A major factor associated with correct detection was displacement.ConclusionWe found good performance of a deep learning-based prototype algorithm detecting rib fractures on trauma CT on a per-examination level at a low rate of false-positives per case. A potential area for clinical application is its use as a screening tool to avoid false-negative radiology reports.
| S-EPMC7289702 | biostudies-literature