Dataset Information

Use of biplane quantitative angiographic imaging with ensemble neural networks to assess reperfusion status during mechanical thrombectomy.

ABSTRACT: Digital subtraction angiography (DSA) is the main imaging modality used to assess reperfusion during mechanical thrombectomy (MT) when treating large vessel occlusion (LVO) ischemic strokes. To improve this visual and subjective assessment, hybrid models combining angiographic parametric imaging (API) with deep learning tools have been proposed. These models use convolutional neural networks (CNN) with single view individual API maps, thus restricting use of complementary information from multiple views and maps resulting in loss of relevant clinical information. This study investigates use of ensemble networks to combine hemodynamic information from multiple bi-plane API maps to assess level of reperfusion. Three-hundred-eighty-three anteroposterior (AP) and lateral view DSAs were retrospectively collected from patients who underwent MTs of anterior circulation LVOs. API peak height (PH) and area under time density curve (AUC) maps were generated. CNNs were developed to classify maps as adequate/inadequate reperfusion as labeled by two neuro-interventionalists. Outputs from individual networks were combined by weighting each output, using a grid search algorithm. Ensembled, AP-AUC, AP-PH, lateral-AUC, and lateral-PH networks achieved accuracies of 83.0% (95% confidence-interval: 81.2%-84.8%), 74.4% (72.0%-76.7%), 74.2% (72.8%-75.7%), 74.9% (72.2%-77.7%), and 76.9% (74.4%-79.5%); area under receiver operating characteristic curves of 0.86 (0.84-0.88), 0.81 (0.79-0.83), 0.83 (0.81-0.84), 0.82 (0.8-0.84), and 0.84 (0.82-0.87); and Matthews correlation coefficients of 0.66 (0.63-0.70), 0.48 (0.43-0.53), 0.49 (0.46-0.52), 0.51 (0.45-0.56), and 0.54 (0.49-0.59) respectively. Ensembled network performance was significantly better than individual networks (McNemar's p-value<0.05). This study proved feasibility of using ensemble networks to combine hemodynamic information from multiple bi-plane API maps to assess level of reperfusion during MTs.

SUBMITTER: Bhurwani MMS

PROVIDER: S-EPMC7946164 | biostudies-literature | 2021 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Use of biplane quantitative angiographic imaging with ensemble neural networks to assess reperfusion status during mechanical thrombectomy.

Bhurwani Mohammad Mahdi Shiraz MMS Snyder Kenneth V KV Waqas Muhammad M Mokin Maxim M Rava Ryan A RA Podgorsak Alexander R AR Sommer Kelsey N KN Davies Jason M JM Levy Elad I EI Siddiqui Adnan H AH Ionita Ciprian N CN

Proceedings of SPIE--the International Society for Optical Engineering 20210215

Digital subtraction angiography (DSA) is the main imaging modality used to assess reperfusion during mechanical thrombectomy (MT) when treating large vessel occlusion (LVO) ischemic strokes. To improve this visual and subjective assessment, hybrid models combining angiographic parametric imaging (API) with deep learning tools have been proposed. These models use convolutional neural networks (CNN) with single view individual API maps, thus restricting use of complementary information from multip ...[more]

PMID: 33707812

Similar Datasets

Project description:IntroductionSymptomatic intracranial hemorrhage represents one of the most feared complications of endovascular reperfusion. We aim to describe a series of patients that experienced immediate reperfusion injury with active intraprocedural extravasation within the territory of the deep penetrating arteries and provide real-time correlation with CT "spot sign."MethodsThis was a retrospective analysis of patients that suffered reperfusion injury with active arterial extravasation during endovascular stroke treatment in two tertiary care centers.ResultsFive patients were identified. Median age was 63 (58-71) years, 66% were male. Median NIHSS was 13.5 (9.5-23.0), platelet level 212,000 (142,000-235,000), baseline systolic blood pressure 152 (133-201) mm Hg, and non-contrast CT ASPECTS 7.0 (6.5-9.0). Two patients were taking aspirin and one had received intravenous thrombolysis. There were three middle cerebral artery M1, one internal carotid artery terminus, and one vertebrobasilar junction occlusion. Three patients had anterior circulation tandem occlusions. Stroke etiology was extracranial atherosclerosis (n = 2), intracranial atherosclerosis (n = 2), and cervical dissection (n = 1). The median time from onset to puncture was 5.5 (3.9-8.6) h. Intravenous heparin was administered in all patients (median dose of 4,750 [3,250-6,000] units) and intravenous abciximab in four. All tandem cases had the cervical lesion addressed first. Four lenticulostriates and one paramedian pontine artery were involved. Intraprocedural flat-panel CT was performed in four (80%) cases and provided real-time correlation between the active contrast extravasation and the "spot sign." The bailout included use of protamine, blood pressure control, and balloon guide catheter or intracranial compliant balloon inflation plus coiling of targeted vessel. All patients had angiographic cessation of bleeding at the end of the procedure with parenchymal hemorrhage type 1 in one case and type 2 in four. Three patients had modified Rankin score of 4 and two were dead at 90 days.ConclusionsActive reperfusion hemorrhage involving perforator arteries was observed to correlate with the CT "spot sign" and to be associated with poor outcomes.

Project description:Background and purposeAlthough current guidelines advocate pretreatment with intravenous thrombolysis (IVT) in all eligible patients with acute ischemic stroke with large-vessel occlusion before mechanical thrombectomy, there are observational data questioning the efficacy of this approach. One of the main arguments in favor of IVT pretreatment is the potential for tissue-type plasminogen activator-induced successful reperfusion (SR) before the onset of endovascular procedure.MethodsWe performed a systematic review and meta-analysis of randomized controlled clinical trials and observational cohorts providing rates of SR with IVT in patients with large-vessel occlusion before the initiation of mechanical thrombectomy. We also performed subgroup analyses according to study type (randomized controlled clinical trials versus observational) and according to the inclusion per protocol of patients with tandem (intracranial/extracranial) occlusions.ResultsWe identified 13 eligible studies (7 randomized controlled clinical trials and 6 observational cohorts), including 1561 patients with acute ischemic stroke (median National Institutes of Health Stroke Scale score, 17) with large-vessel occlusion. SR following IVT and before mechanical thrombectomy was documented in 11% (95% confidence interval, 7%-16%), with no difference among cohorts derived from randomized controlled clinical trials and observational studies. There was significant heterogeneity across included studies both in the overall analysis and among subgroups (I2>84%; P for Cochran Q, <0.001). Higher tissue-type plasminogen activator-induced SR rates were documented in studies reporting the exclusion of tandem occlusions (17%; 95% confidence interval, 11%-23%) compared with the rest (7%; 95% confidence interval, 4%-11%; P for subgroup differences, 0.003).ConclusionsPretreatment with systemic thrombolysis in patients with large-vessel occlusion eligible for mechanical thrombectomy results in SR in 1 of 10 cases, negating the need for additional endovascular reperfusion. Tandem occlusions seem to be the least responsive to IVT pretreatment.

Project description:Background and purposePost hoc analyses of randomized controlled clinical trials evaluating mechanical thrombectomy have suggested that admission-to-groin-puncture (ATG) delays are associated with reduced reperfusion rates. Purpose of this analysis was to validate this association in a real-world cohort and to find associated factors and confounders for prolonged ATG intervals.MethodsPatients included into the BEYOND-SWIFT cohort (Bernese-European Registry for Ischemic Stroke Patients Treated Outside Current Guidelines With Neurothrombectomy Devices Using the Solitaire FR With the Intention for Thrombectomy; https://www.clinicaltrials.gov; Unique identifier: NCT03496064) were analyzed (n=2386). Association between baseline characteristics and ATG was evaluated using mixed linear regression analysis. The effect of increasing symptom-onset-to-admission and ATG intervals on successful reperfusion (defined as Thrombolysis in Cerebral Infarction [TICI] 2b-3) was evaluated using logistic regression analysis adjusting for potential confounders.ResultsMedian ATG was 73 minutes. Prolonged ATG intervals were associated with the use of magnetic resonance imaging (+19.1 [95% CI, +9.1 to +29.1] minutes), general anesthesia (+12.1 [95% CI, +3.7 to +20.4] minutes), and borderline indication criteria, such as lower National Institutes of Health Stroke Scale, late presentations, or not meeting top-tier early time window eligibility criteria (+13.8 [95% CI, +6.1 to +21.6] minutes). There was a 13% relative odds reduction for TICI 2b-3 (adjusted odds ratio [aOR], 0.87 [95% CI, 0.79-0.96]) and TICI 2c/3 (aOR, 0.87 [95% CI, 0.79-0.95]) per hour ATG delay, while the reduction of TICI 2b-3 per hour increase symptom-onset-to-admission was minor (aOR, 0.97 [95% CI, 0.94-0.99]) and inconsistent regarding TICI 2c/3 (aOR, 0.99 [95% CI, 0.97-1.02]). After adjusting for identified factors associated with prolonged ATG intervals, the association of ATG delay and lower rates of TICI 2b-3 remained tangible (aOR, 0.87 [95% CI, 0.76-0.99]).ConclusionsThere is a great potential to reduce ATG, and potential targets for improvement can be deduced from observational data. The association between in-hospital delay and reduced reperfusion rates is evident in real-world clinical data, underscoring the need to optimize in-hospital workflows. Given the only minor association between symptom-onset-to-admission intervals and reperfusion rates, the causal relationship of this association warrants further research. Registration: URL: https://www.clinicaltrials.gov. Unique identifier: NCT03496064.

Dataset Information

Use of biplane quantitative angiographic imaging with ensemble neural networks to assess reperfusion status during mechanical thrombectomy.

Publications

Use of biplane quantitative angiographic imaging with ensemble neural networks to assess reperfusion status during mechanical thrombectomy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets