Diagnostic accuracy of automated occlusion detection in CT angiography using e-CTA.
Ontology highlight
ABSTRACT: Background and aim
The aim of this study was to assess the diagnostic accuracy of e-CTA (product name) (Brainomix) in the automatic detection of large vessel occlusions in anterior circulation stroke.Methods
Of 487 CT angiographies from patients with large vessel occlusions stroke, 327 were used to train the algorithm while the remaining cases together with 140 negative CT angiographies were used to validate its performance against ground truth. Of these 301 cases, 144 were randomly selected and used for an additional comparative analysis against 4 raters. Sensitivity, specificity, positive and negative predictive value (PPV and NPV), accuracy and level of agreement with ground truth (Cohen's Kappa) were determined and compared to the performance of a neuroradiologist, a radiology resident, and two neurology residents.Results
e-CTA had a sensitivity and specificity of 0.84 (0.77-0.89) and 0.96 (0.91-0.98) respectively for the detection of any large vessel occlusions on the correct side in the whole validation cohort. This performance was identical in the comparative analysis subgroup and was within the range of physicians at different levels of expertise: 0.86-0.97 and 0.91-1.00, respectively. For the detection of proximal occlusions, it was 0.92 (0.84-0.96) and 0.98 (0.94-1.00) for the whole cohort and 0.93 (0.80-0.98) and 1.00 (0.95-1.00) for the comparative analysis, respectively for e-CTA. The range was 0.8-0.97 for sensitivity and 0.97-1.00 for specificity for the four physicians.Conclusions
The performance of e-CTA in detecting any large vessel occlusions is comparable to less experienced physicians but is similar to experienced physicians for detecting proximal large vessel occlusions.
SUBMITTER: Seker F
PROVIDER: S-EPMC8739618 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The Automation Platform (AP) is a software platform to support the workflow of radiologists and includes a stroke CT package with integrated artificial intelligence (AI) based tools. The aim of this study was to evaluate the diagnostic performance of the AP for the detection of intracranial large vessel occlusions (LVO) on conventional CT angiography (CTA), and the duration of CT processing in a cohort of acute stroke patients. The diagnostic performance for intracranial LVO detection on CTA by the AP was evaluated in a retrospective cohort of 100 acute stroke patients and compared to the diagnostic performance of five radiologists with different levels of experience. The reference standard was set by an independent neuroradiologist, with access to the readings of the different radiologists, clinical data, and follow-up. The data processing time of the AP for ICH detection on non-contrast CT, LVO detection on CTA, and the processing of CTP maps was assessed in a subset 60 patients of the retrospective cohort. This was compared to 13 radiologists, who were prospectively timed for the processing and reading of 21 stroke CTs. The AP showed shorter processing time of CTA (mean 60 versus 395 s) and CTP (mean 196 versus 243-349 s) as compared to radiologists, but showed lower sensitivity for LVO detection (sensitivity 77% of the AP vs mean sensitivity 87% of radiologists). If the AP would have been used as a stand-alone system, 1 ICA occlusion, 2 M1 occlusions and 8 M2 occlusions would have been missed, which would be eligible for mechanical thrombectomy. In conclusion, the AP showed shorter processing time of CTA and CTP as compared with radiologists, which illustrates the potential of the AP to speed-up the diagnostic work-up. However, its performance for LVO detection was lower as compared with radiologists, especially for M2 vessel occlusions.
| S-EPMC10397283 | biostudies-literature
Project description:ObjectiveTo evaluate whether the use of multiphase CT angiography (CTA) improves interrater agreement for intracranial occlusion detection between stroke neurology trainees and an expert neuroradiologist.MethodsA neuroradiologist and 2 stroke neurology fellows independently reviewed 100 prospectively collected single-phase and multiphase CTA scans from acute ischemic stroke patients with mild symptoms (NIH Stroke Scale score ≤5). The presence and location of a vascular occlusion(s) were documented. Interrater agreement single- and multiphase CTA was quantified using unweighted κ statistics. We assessed for any occlusions, anterior vs posterior occlusions, and pial vessel asymmetry.ResultsUsing multiphase CTA, the neuroradiologist detected 50 scans with anterior circulation occlusions and 15 scans with posterior circulation occlusions. Median reading time was 2 minutes per scan. Median reading time for the neurologists was 3 minutes per multiphase CTA scan. Interrater agreement was fair between the 2 neurologists and neuroradiologist when using single-phase CTA (κ = 0.45 and 0.32). Agreement improved minimally when stratified by anterior vs posterior circulation. When using multiphase CTA, agreement was high for detection of occlusion or asymmetry of pial vessels in the anterior circulation (κ = 0.80 and 0.84).ConclusionsMultiphase CTA improves diagnostic accuracy in minor ischemic stroke for detection of anterior circulation intracranial occlusion.Classification of evidenceThis study provides Class II evidence that multiphase CTA, compared to single-phase CTA, improves the interrater agreement between stroke neurology trainees and an expert neuroradiologist for detecting anterior circulation intracranial vascular occlusion in patients with minor acute ischemic strokes.
| S-EPMC4977373 | biostudies-literature
Project description:ObjectivesComputed tomography (CT) pulmonary angiography is widely used in patients with suspected pulmonary hypertension (PH). However, the diagnostic and prognostic significance remains unclear. The aim of this study was to (a) build a diagnostic CT model and (b) test its prognostic significance.MethodsConsecutive patients with suspected PH undergoing routine CT pulmonary angiography and right heart catheterisation (RHC) were identified. Axial and reconstructed images were used to derive CT metrics. Multivariate regression analysis was performed in the derivation cohort to identify a diagnostic CT model to predict mPAP ≥ 25 mmHg (the existing ESC guideline definition of PH) and > 20 mmHg (the new threshold proposed at the 6th World Symposium on PH). In the validation cohort, sensitivity, specificity and compromise CT thresholds were identified with receiver operating characteristic (ROC) analysis. The prognostic value of the CT model was assessed using Kaplan-Meier analysis.ResultsBetween 2012 and 2016, 491 patients were identified. In the derivation cohort (n = 247), a CT model was identified including pulmonary artery diameter, right ventricular outflow tract thickness, septal angle and left ventricular area. In the validation cohort (n = 244), the model was diagnostic, with an area under the ROC curve of 0.94/0.91 for mPAP ≥ 25/> 20 mmHg respectively. In the validation cohort, 93 patients died; mean follow-up was 42 months. The diagnostic thresholds for the CT model were prognostic, log rank, all p < 0.01.DiscussionIn suspected PH, a diagnostic CT model had diagnostic and prognostic utility.Key points• Diagnostic CT models have high diagnostic accuracy in a tertiary referral population of with suspected PH. • Diagnostic CT models stratify patients by mortality in suspected PH.
| S-EPMC7431437 | biostudies-literature
Project description:PurposeComputed tomography angiography (CTA) is routinely used to detect large-vessel occlusion (LVO) in patients with suspected acute ischemic stroke; however, visual analysis is time consuming and prone to error. To evaluate solutions to support imaging triage, we tested performance of automated analysis of CTA source images (CTASI) at identifying patients with LVO.MethodsStroke patients with LVO were selected from a prospectively acquired cohort. A control group was selected from consecutive patients with clinically suspected stroke without signs of ischemia on CT perfusion (CTP) or infarct on follow-up. Software-based automated segmentation and Hounsfield unit (HU) measurements were performed on CTASI for all regions of the Alberta Stroke Program Early CT score (ASPECTS). We derived different parameters from raw measurements and analyzed their performance to identify patients with LVO using receiver operating characteristic curve analysis.ResultsThe retrospective analysis included 145 patients, 79 patients with LVO stroke and 66 patients without stroke. The parameters hemispheric asymmetry ratio (AR), ratio between highest and lowest regional AR and M2-territory AR produced area under the curve (AUC) values from 0.95-0.97 (all p < 0.001) for detecting presence of LVO in the total population. Resulting sensitivity (sens)/specificity (spec) defined by the Youden index were 0.87/0.97-0.99. Maximum sens/spec defined by the specificity threshold ≥0.70 were 0.91-0.96/0.77-0.83. Performance in a small number of patients with isolated M2 occlusion was lower (AUC: 0.72-0.85).ConclusionAutomated attenuation measurements on CTASI identify proximal LVO stroke patients with high sensitivity and specificity. This technique can aid in accurate and timely patient selection for thrombectomy, especially in primary stroke centers without CTP capacity.
| S-EPMC8463515 | biostudies-literature
Project description:Study questionWhat are the diagnostic yield and accuracy of early computed tomography (CT) angiography followed by magnetic resonance imaging/angiography (MRI/MRA) and digital subtraction angiography (DSA) in patients with non-traumatic intracerebral haemorrhage?MethodsThis prospective diagnostic study enrolled 298 adults (18-70 years) treated in 22 hospitals in the Netherlands over six years. CT angiography was performed within seven days of haemorrhage. If the result was negative, MRI/MRA was performed four to eight weeks later. DSA was performed when the CT angiography or MRI/MRA results were inconclusive or negative. The main outcome was a macrovascular cause, including arteriovenous malformation, aneurysm, dural arteriovenous fistula, and cavernoma. Three blinded neuroradiologists independently evaluated the images for macrovascular causes of haemorrhage. The reference standard was the best available evidence from all findings during one year's follow-up.Study answer and limitationsA macrovascular cause was identified in 69 patients (23%). 291 patients (98%) underwent CT angiography; 214 with a negative result underwent additional MRI/MRA and 97 with a negative result for both CT angiography and MRI/MRA underwent DSA. Early CT angiography detected 51 macrovascular causes (yield 17%, 95% confidence interval 13% to 22%). CT angiography with MRI/MRA identified two additional macrovascular causes (18%, 14% to 23%) and these modalities combined with DSA another 15 (23%, 18% to 28%). This last extensive strategy failed to detect a cavernoma, which was identified on MRI during follow-up (reference strategy). The positive predictive value of CT angiography was 72% (60% to 82%), of additional MRI/MRA was 35% (14% to 62%), and of additional DSA was 100% (75% to 100%). None of the patients experienced complications with CT angiography or MRI/MRA; 0.6% of patients who underwent DSA experienced permanent sequelae. Not all patients with negative CT angiography and MRI/MRA results underwent DSA. Although the previous probability of finding a macrovascular cause was lower in patients who did not undergo DSA, some small arteriovenous malformations or dural arteriovenous fistulas may have been missed.What this study addsCT angiography is an appropriate initial investigation to detect macrovascular causes of non-traumatic intracerebral haemorrhage, but accuracy is modest. Additional MRI/MRA may find cavernomas or alternative diagnoses, but DSA is needed to diagnose macrovascular causes undetected by CT angiography or MRI/MRA.Funding, competing interests, data sharingDutch Heart Foundation and The Netherlands Organisation for Health Research and Development, ZonMw. The authors have no competing interests. Direct requests for additional data to the corresponding author.
| S-EPMC4637845 | biostudies-literature
Project description:ObjectivesTo compare diagnostic performance and time efficiency between 3D multipath curved planar reformations (mpCPRs) and axial images of CT angiography for the pre-interventional assessment of peripheral arterial disease (PAD), with digital subtraction angiography as the standard of reference.MethodsForty patients (10 females, mean age 72 years), referred to CTA prior to endovascular treatment of PAD, were prospectively included and underwent peripheral CT angiography. A semiautomated toolbox was used to render mpCPRs. Twenty-one arterial segments were defined in each leg; for each segment, the presence of stenosis > 70% was assessed on mpCPRs and axial images by two readers, independently, with digital subtraction angiography as gold standard.ResultsBoth readers reached lower sensitivity (Reader 1: 91 vs. 94%, p = 0.08; Reader 2: 89 vs. 93%, p = 0.03) but significantly higher specificity (Reader 1: 94 vs. 89%, p < 0.01; Reader 2: 96 vs. 95%, p = 0.01) with mpCPRs than with axial images. Reader 1 achieved significantly higher accuracy with mpCPRs (93 vs. 91%, p = 0.02), and Reader 2 had similar overall accuracy in both evaluations (94 vs. 94%, p = 0.96). Both readers read mpCPRs significantly faster than axial images (Reader 1: 5'45? based on mpCPRs vs. 7'40? based on axial images; Reader 2: 4'41? based on mpCPRs vs. 6'57? based on axial images; p < 0.01).ConclusionsmpCPRs are a promising 3D reformation technique that facilitates a fast assessment of PAD with high diagnostic accuracy.
| S-EPMC5876266 | biostudies-literature
Project description:IntroductionEarly detection of large vessel occlusion (LVO) is essential to facilitate fast endovascular treatment. CT angiography (CTA) is used to detect LVO in suspected stroke patients. We aimed to assess the accuracy of CTA evaluations in daily clinical practice in a large cohort of suspected stroke patients.Patients and methodsWe used data from the PRESTO study, a multicenter prospective observational cohort study that included suspected stroke patients between August 2018 and September 2019. Baseline CTAs were re-evaluated by an imaging core laboratory and compared to the local assessment. LVO was defined as an occlusion of the intracranial internal carotid artery, M1 segment, or basilar artery. Medium vessel occlusion (MeVO) was defined as an A1, A2, or M2 occlusion. We calculated the accuracy, sensitivity, and specificity to detect LVO and LVO+MeVO, using the core laboratory evaluation as reference standard.ResultsWe included 656 patients. The core laboratory detected 89 LVOs and 74 MeVOs in 155 patients. Local observers missed 6 LVOs (7%) and 28 MeVOs (38%), of which 23 M2 occlusions. Accuracy of LVO detection was 99% (95% CI: 98-100%), sensitivity 93% (95% CI: 86-97%), and specificity 100% (95% CI: 99-100%). Accuracy of LVO+MeVO detection was 95% (95% CI: 93-96%), sensitivity 79% (95% CI: 72-85%), and specificity 99% (95% CI: 98-100%).Discussion and conclusionCTA evaluations in daily clinical practice are highly accurate and LVOs are adequately recognized. The detection of MeVOs seems more challenging. The evolving EVT possibilities emphasize the need to improve CTA evaluations in the acute setting.
| S-EPMC8948506 | biostudies-literature
Project description:ObjectivesTo assess diagnostic accuracy of automated 3D volumetry of cardiac chambers based on computed tomography pulmonary angiography (CTPA) for the differentiation of pulmonary hypertension due to left heart disease (group 2 PH) from non-group 2 PH compared to manual diameter measurements.MethodsPatients with confirmed PH undergoing right heart catheterisation and CTPA within 100 days for diagnostic workup of PH between August 2013 and February 2016 were included in this retrospective, single-centre study. Automated 3D segmentation of left atrium, left ventricle, right atrium and right ventricle (LA/LV/RA/RV) was performed by two independent and blinded radiologists using commercial software. For comparison, axial diameters were manually measured. The ability to differentiate group 2 PH from non-group 2 PH was assessed by means of logistic regression.ResultsNinety-one patients (median 67.5 years, 44 women) were included, thereof 19 patients (20.9%) classified as group 2 PH. After adjustment for age, sex and mean pulmonary arterial pressure, group 2 PH was significantly associated with larger LA volume (p < 0.001), larger LV volume (p = 0.001), lower RV/LV volume ratio (p = 0.04) and lower RV/LA volume ratio (p = 0.003). LA volume demonstrated the highest discriminatory ability to identify group 2 PH (AUC, 0.908; 95% confidence interval, 0.835-0.981) and was significantly superior to LA diameter (p = 0.009). Intraobserver and interobserver agreements were excellent for all volume measurements (intraclass correlation coefficients 0.926-0.999, all p < 0.001).ConclusionsLA volume quantified by automated, CTPA-based 3D volumetry can differentiate group 2 PH from other PH groups with good diagnostic accuracy and yields significantly higher diagnostic accuracy than left atrial diameter.Key points• Automated cardiac chamber volumetry using non-gated CT pulmonary angiography can differentiate pulmonary hypertension due to left heart disease from other causes with good diagnostic accuracy. • Left atrial volume yields significantly higher diagnostic accuracy than left atrial axial diameter for identification of pulmonary hypertension due to left heart disease without time-consuming manual processing.
| S-EPMC9279230 | biostudies-literature
Project description:BackgroundPrehospital automated large vessel occlusion (LVO) detection in Mobile Stroke Units (MSUs) could accelerate identification and treatment of patients with LVO acute ischemic stroke. Here, we evaluate the performance of a machine learning (ML) model on CT angiograms (CTAs) obtained from 2 MSUs to detect LVO.MethodsPatients evaluated on MSUs in Houston and Los Angeles with out-of-hospital CTAs were identified. Anterior circulation LVO was defined as an occlusion of the intracranial internal carotid artery, middle cerebral artery (M1 or M2), or anterior cerebral artery vessels and determined by an expert human reader. A ML model to detect LVO was trained and tested on independent data sets consisting of in-hospital CTAs and then tested on MSU CTA images. Model performance was determined using area under the receiver-operator curve statistics.ResultsAmong 68 patients with out-of-hospital MSU CTAs, 40% had an LVO. The most common occlusion location was the middle cerebral artery M1 segment (59%), followed by the internal carotid artery (30%), and middle cerebral artery M2 (11%). Median time from last known well to CTA imaging was 88.0 (interquartile range, 59.5-196.0) minutes. After training on 870 in-hospital CTAs, the ML model performed well in identifying LVO in a separate in-hospital data set of 441 images with area under receiver-operator curve of 0.84 (95% CI, 0.80-0.87). ML algorithm analysis time was under 1 minute. The performance of the ML model on the MSU CTA images was comparable with area under receiver-operator curve 0.80 (95% CI, 0.71-0.89). There was no significant difference in performance between the Houston and Los Angeles MSU CTA cohorts.ConclusionsIn this study of patients evaluated on MSUs in 2 cities, a ML algorithm was able to accurately and rapidly detect LVO using prehospital CTA acquisitions.
| S-EPMC9038611 | biostudies-literature
Project description:We aimed to assess the ability of deep learning (DL) and support vector machine (SVM) to detect a nonperfusion area (NPA) caused by retinal vein occlusion (RVO) with optical coherence tomography angiography (OCTA) images. The study included 322 OCTA images (normal: 148; NPA owing to RVO: 174 [128 branch RVO images and 46 central RVO images]). Training to construct the DL model using deep convolutional neural network (DNN) algorithms was provided using OCTA images. The SVM used a scikit-learn library with a radial basis function kernel. The area under the curve (AUC), sensitivity and specificity for detecting an NPA were examined. We compared the diagnostic ability (sensitivity, specificity and average required time) between the DNN, SVM and seven ophthalmologists. Heat maps were generated. With regard to the DNN, the mean AUC, sensitivity, specificity and average required time for distinguishing RVO OCTA images with an NPA from normal OCTA images were 0.986, 93.7%, 97.3% and 176.9 s, respectively. With regard to SVM, the mean AUC, sensitivity, and specificity were 0.880, 79.3%, and 81.1%, respectively. With regard to the seven ophthalmologists, the mean AUC, sensitivity, specificity and average required time were 0.962, 90.8%, 89.2%, and 700.6 s, respectively. The DNN focused on the foveal avascular zone and NPA in heat maps. The performance of the DNN was significantly better than that of SVM in all parameters (p < 0.01, all) and that of the ophthalmologists in AUC and specificity (p < 0.01, all). The combination of DL and OCTA images had high accuracy for the detection of an NPA, and it might be useful in clinical practice and retinal screening.
| S-EPMC6837754 | biostudies-literature