Abstract P544: Automated Artificial Intelligence Based Detection and Location Specification of Large Vessel Occlusion on CT Angiography in Stroke

Stroke ◽  
2021 ◽  
Vol 52 (Suppl_1) ◽  
Author(s):  
Agnetha A Bruggeman ◽  
Miou Koopman ◽  
Jazba Soomro ◽  
Albert J Yoo ◽  
Henk A Marquering ◽  
...  
Keyword(s):  
Predictive Value ◽  
Diagnostic Performance ◽  
Ground Truth ◽  
The United States ◽  
Detection Algorithm ◽  
Interrater Agreement ◽  
Large Vessel ◽  
Imaging Data ◽  
Anterior Circulation ◽  
Vessel Occlusion

Introduction: Fast and accurate detection of large vessel occlusions (LVOs) is crucial in selection of patients for endovascular treatment. We assessed the diagnostic performance and speed of an Artifical Intelligence algorithm for automated LVO detection with a novel feature that specifies the exact level of occlusion. Methods: All Computed Tomography Angiography (CTA) imaging data were analyzed by an automated algorithm for anterior circulation LVOs (internal carotid artery (ICA), M1 or M2 segments of the middle cerebral artery) (StrokeViewer, Nico.lab). Ground truth was established by consensus of two independent neuroradiologist readings. Diagnostic performance was assessed by calculating sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). Performance of the LVO localization feature was assessed by calculating interrater agreement (Cohen’s Kappa) between the algorithm and the expert panel. Results: CTAs from 297 patients referred or directly admitted to a comprehensive stroke center in the United States (mean age 67 years, SD 15; 145 males) were analyzed. One-hundred and fifty-six patients had an anterior circulation LVO. Location of the occlusions was ICA (n=43 [28%]), M1 (n=79 [51%]) and M2 (n=34 [22%]). Sensitivity and specificity for LVO detection were respectively 92% (95% CI, 86.2%-95.5%) and 85% (95% CI, 78.1%-90.5%). NPV and PPV were 90% and 87% respectively. Interrater agreement between the algorithm and the expert observers for LVO location was 0.92 (95% CI, 0.86-0.98). Median upload-to-notification time for all cases was 3 minutes, 34 seconds (minimal 2:28 minutes; maximal 5:03 minutes). Conclusions: Anterior circulation LVOs can be rapidly and accurately detected by an automated LVO detection algorithm with reliable localization of the involved vessel segment. Therefore, the algorithm presented in this study is a suitable screening tool to support diagnosis of LVOs.

scholarly journals Automating Stroke Data Extraction From Free-Text Radiology Reports Using Natural Language Processing: Instrument Validation Study (Preprint)

2020 ◽  
Author(s):  
Amy Y X Yu ◽  
Zhongyu A Liu ◽  
Chloe Pou-Prom ◽  
Kaitlyn Lopes ◽  
Moira K Kapral ◽  
...  
Keyword(s):  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Predictive Value ◽  
Large Vessel ◽  
Free Text ◽  
Imaging Data ◽  
Large Vessel Occlusion ◽  
Vessel Occlusion ◽  
Radiology Reports

BACKGROUND Diagnostic neurovascular imaging data are important in stroke research, but obtaining these data typically requires laborious manual chart reviews. OBJECTIVE We aimed to determine the accuracy of a natural language processing (NLP) approach to extract information on the presence and location of vascular occlusions as well as other stroke-related attributes based on free-text reports. METHODS From the full reports of 1320 consecutive computed tomography (CT), CT angiography, and CT perfusion scans of the head and neck performed at a tertiary stroke center between October 2017 and January 2019, we manually extracted data on the presence of proximal large vessel occlusion (primary outcome), as well as distal vessel occlusion, ischemia, hemorrhage, Alberta stroke program early CT score (ASPECTS), and collateral status (secondary outcomes). Reports were randomly split into training (n=921) and validation (n=399) sets, and attributes were extracted using rule-based NLP. We reported the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the overall accuracy of the NLP approach relative to the manually extracted data. RESULTS The overall prevalence of large vessel occlusion was 12.2%. In the training sample, the NLP approach identified this attribute with an overall accuracy of 97.3% (95.5% sensitivity, 98.1% specificity, 84.1% PPV, and 99.4% NPV). In the validation set, the overall accuracy was 95.2% (90.0% sensitivity, 97.4% specificity, 76.3% PPV, and 98.5% NPV). The accuracy of identifying distal or basilar occlusion as well as hemorrhage was also high, but there were limitations in identifying cerebral ischemia, ASPECTS, and collateral status. CONCLUSIONS NLP may improve the efficiency of large-scale imaging data collection for stroke surveillance and research.

scholarly journals Automating Stroke Data Extraction From Free-Text Radiology Reports Using Natural Language Processing: Instrument Validation Study

10.2196/24381 ◽  
2021 ◽  
Vol 9 (5) ◽  
pp. e24381
Author(s):  
Amy Y X Yu ◽  
Zhongyu A Liu ◽  
Chloe Pou-Prom ◽  
Kaitlyn Lopes ◽  
Moira K Kapral ◽  
...  
Keyword(s):  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Predictive Value ◽  
Large Vessel ◽  
Free Text ◽  
Imaging Data ◽  
Large Vessel Occlusion ◽  
Vessel Occlusion ◽  
Radiology Reports

Background Diagnostic neurovascular imaging data are important in stroke research, but obtaining these data typically requires laborious manual chart reviews. Objective We aimed to determine the accuracy of a natural language processing (NLP) approach to extract information on the presence and location of vascular occlusions as well as other stroke-related attributes based on free-text reports. Methods From the full reports of 1320 consecutive computed tomography (CT), CT angiography, and CT perfusion scans of the head and neck performed at a tertiary stroke center between October 2017 and January 2019, we manually extracted data on the presence of proximal large vessel occlusion (primary outcome), as well as distal vessel occlusion, ischemia, hemorrhage, Alberta stroke program early CT score (ASPECTS), and collateral status (secondary outcomes). Reports were randomly split into training (n=921) and validation (n=399) sets, and attributes were extracted using rule-based NLP. We reported the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the overall accuracy of the NLP approach relative to the manually extracted data. Results The overall prevalence of large vessel occlusion was 12.2%. In the training sample, the NLP approach identified this attribute with an overall accuracy of 97.3% (95.5% sensitivity, 98.1% specificity, 84.1% PPV, and 99.4% NPV). In the validation set, the overall accuracy was 95.2% (90.0% sensitivity, 97.4% specificity, 76.3% PPV, and 98.5% NPV). The accuracy of identifying distal or basilar occlusion as well as hemorrhage was also high, but there were limitations in identifying cerebral ischemia, ASPECTS, and collateral status. Conclusions NLP may improve the efficiency of large-scale imaging data collection for stroke surveillance and research.

scholarly journals Retrospective Validation of Prehospital Acute Stroke Severity Scale to Predict Large Vessel Occlusion in Acute Stroke Patients - Single Center study in China

2019 ◽  
Author(s):  
Xiaoli Si ◽  
Yuanjian Fang ◽  
Wenqing Xia ◽  
Tianwen Chen ◽  
Huan Huang ◽  
...  
Keyword(s):  
Acute Stroke ◽  
Predictive Value ◽  
Large Vessel ◽  
Motor Deficits ◽  
Stroke Severity ◽  
Large Vessel Occlusion ◽  
Anterior Circulation ◽  
Vessel Occlusion ◽  
Consciousness Disorder ◽  
Gaze Palsy

Abstract Background and Purpose - To date, identifying emergent large vessel occlusion (ELVO) patients in the prehospital stage is important but still challenging. We aimed to retrospectively validate a simple prehospital stroke scale——Prehospital Acute Stroke Severity (PASS) scale to identify ELVO. Methods - We retrospectively evaluated our consecutive cohort of acute ischemic stroke (AIS) who underwent CT angiography (CTA), MR angiography (MRA) or digital subtraction angiography (DSA). PASS scale was calculated based on National Institutes of Health Stroke Scale (NIHSS) items retrospectively. The comparison of diagnostic parameters between PASS scale and NIHSS scale were performed. Results - Finally, a total of 605 patients were enrolled. ELVO patients with PASS≥2 had a median NIHSS score of 14. The best predictive value of PASS≥2 showed a similar predictive value compared with NIHSS≥9. Cortical symptoms such as consciousness disorder and gaze palsy were more specific indicators for ELVO than motor deficits. Consciousness disorder was more serious in posterior circulation infarct (PIC) while gaze palsy was more common in anterior circulation infarct (AIC). Conclusions - PASS scale had both good discrimination and calibration in our retrospective cohort. It could reflect acute stroke severity well and predict ELVO in an effective and simple way. Moreover, cortical symptoms had high specificities to predict ELVO on their own.

scholarly journals Fast Automatic Detection of Large Vessel Occlusions on CT Angiography

Stroke ◽  
2019 ◽  
Vol 50 (12) ◽  
pp. 3431-3438 ◽  
Author(s):  
Shalini A. Amukotuwa ◽  
Matus Straka ◽  
Seena Dehkharghani ◽  
Roland Bammer
Keyword(s):  
Carotid Artery ◽  
Internal Carotid Artery ◽  
Diagnostic Performance ◽  
Internal Carotid ◽  
Roc Curves ◽  
Interquartile Range ◽  
Large Vessel ◽  
Therapeutic Window ◽  
Anterior Circulation ◽  
Vessel Occlusion

Background and Purpose— Accurate and rapid detection of anterior circulation large vessel occlusion (LVO) is of paramount importance in patients with acute stroke due to the potentially rapid infarction of at-risk tissue and the limited therapeutic window for endovascular clot retrieval. Hence, the optimal threshold of a new, fully automated software-based approach for LVO detection was determined, and its diagnostic performance evaluated in a large cohort study. Methods— For this retrospective study, data were pooled from: 2 stroke trials, DEFUSE 2 (n=62; 07/08–09/11) and DEFUSE 3 (n=213; 05/17–05/18); a cohort of endovascular clot retrieval candidates (n=82; August 2, 2014–August 30, 2015) and normals (n=111; June 6, 2017–January 28, 2019) from a single quaternary center; and code stroke patients (n=501; January 1, 2017–December 31, 2018) from a single regional hospital. All CTAs were assessed by the automated algorithm. Consensus reads by 2 neuroradiologists served as the reference standard. ROC analysis was used to assess diagnostic performance of the algorithm for detection of (1) anterior circulation LVOs involving the intracranial internal carotid artery or M1 segment middle cerebral artery (M1-MCA); (2) anterior circulation LVOs and proximal M2 segment MCA (M2-MCA) occlusions; and (3) individual segment occlusions. Results— CTAs from 926 patients (median age 70 years, interquartile range: 58-80; 422 females) were analyzed. Three hundred ninety-five patients had an anterior circulation LVO or M2-MCA occlusion (National Institutes of Health Stroke Scale 14 [median], interquartile range: 9–19). Sensitivity and specificity were 97% and 74%, respectively, for LVO detection, and 95% and 79%, respectively, when M2 occlusions were included. On analysis by occlusion site, sensitivities were 90% (M2-MCA), 97% (M1-MCA), and 97% (intracranial internal carotid artery) with corresponding area-under-the-ROC-curves of 0.874 (M2), 0.962 (M1), and 0.997 (intracranial internal carotid artery). Conclusions— Intracranial anterior circulation LVOs and proximal M2 occlusions can be rapidly and reliably detected by an automated detection tool, which may facilitate intra- and inter-instutional workflows and emergent imaging triage in the care of patients with stroke.

scholarly journals Diagnostic performance of an algorithm for automated large vessel occlusion detection on CT angiography

2021 ◽  
pp. neurintsurg-2021-017842
Author(s):  
Sven P R Luijten ◽  
Lennard Wolff ◽  
Martijne H C Duvekot ◽  
Pieter-Jan van Doormaal ◽  
Walid Moudrous ◽  
...  
Keyword(s):  
Ct Angiography ◽  
Diagnostic Performance ◽  
Area Under The Curve ◽  
Detection Algorithm ◽  
Machine Learning Algorithms ◽  
Large Vessel ◽  
Large Vessel Occlusion ◽  
Vessel Occlusion ◽  
High Detection Rate ◽  
Mr Clean

BackgroundMachine learning algorithms hold the potential to contribute to fast and accurate detection of large vessel occlusion (LVO) in patients with suspected acute ischemic stroke. We assessed the diagnostic performance of an automated LVO detection algorithm on CT angiography (CTA).MethodsData from the MR CLEAN Registry and PRESTO were used including patients with and without LVO. CTA data were analyzed by the algorithm for detection and localization of LVO (intracranial internal carotid artery (ICA)/ICA terminus (ICA-T), M1, or M2). Assessments done by expert neuroradiologists were used as reference. Diagnostic performance was assessed for detection of LVO and per occlusion location by means of sensitivity, specificity, and area under the curve (AUC).ResultsWe analyzed CTAs of 1110 patients from the MR CLEAN Registry (median age (IQR) 71 years (60–80); 584 men; 1110 with LVO) and of 646 patients from PRESTO (median age (IQR) 73 years (62–82); 358 men; 141 with and 505 without LVO). For detection of LVO, the algorithm yielded a sensitivity of 89% in the MR CLEAN Registry and a sensitivity of 72%, specificity of 78%, and AUC of 0.75 in PRESTO. Sensitivity per occlusion location was 88% for ICA/ICA-T, 94% for M1, and 72% for M2 occlusion in the MR CLEAN Registry, and 80% for ICA/ICA-T, 95% for M1, and 49% for M2 occlusion in PRESTO.ConclusionThe algorithm provided a high detection rate for proximal LVO, but performance varied significantly by occlusion location. Detection of M2 occlusion needs further improvement.

scholarly journals CT perfusion based ASPECTS improves the diagnostic performance of early ischemic changes in large vessel occlusion

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Tiegong Wang ◽  
Luguang Chen ◽  
Xianglan Jin ◽  
Yuan Yuan ◽  
Qianwen Zhang ◽  
...  
Keyword(s):  
Clinical Outcome ◽  
Diagnostic Performance ◽  
Ct Perfusion ◽  
Roc Curves ◽  
Large Vessel ◽  
Large Vessel Occlusion ◽  
Scoring Method ◽  
Anterior Circulation ◽  
Vessel Occlusion ◽  
Poor Outcomes

Abstract Background ASPECTS scoring method varies, but which one is most suitable for predicting the prognosis still unclear. We aimed to evaluate the diagnostic performance of Automated (Auto)-, noncontrast CT (NCCT)- and CT perfusion (CTP) -ASPECTS for early ischemic changes (EICs) in acute ischemic stroke patients with large vessel occlusion (LVO) and to explore which scoring method is most suitable for predicting the clinical outcome. Methods Eighty-one patients with anterior circulation LVO were retrospectively enrolled and grouped as having a good (0–2) or poor (3–6) clinical outcome using a 90-day modified Rankin Scale score. Clinical characteristics and perfusion parameters were compared between the patients with good and poor outcomes. Differences in scores obtained with the three scoring methods were assessed. Diagnosis performance and receiver operating characteristic (ROC) curves were used to evaluate the value of the three ordinal or dichotomized ASPECTS methods for predicting the clinical outcome. Results Sixty-three patients were finally included, with 36 (57.1%) patients having good clinical outcome. Significant differences were observed in the ordinal or dichotomized Auto-, NCCT- and CTP-ASPECTS between the patients with good and poor clinical outcomes (all p < 0.01). The areas under the curves (AUCs) of the ordinal and dichotomized CTP-ASPECTS were higher than that of the other two methods (all p < 0.01), but the AUCs of the Auto-ASPECTS was similar to that of the NCCT-ASPECTS (p > 0.05). Conclusions The CTP-ASPECTS is superior to the Auto- and NCCT-ASPECTS in detecting EICs in LVO. CTP-ASPECTS with a cutoff value of 6 is a good predictor of the clinical outcome at 90-day follow-up.

scholarly journals Factors associated with prominent vessel sign on susceptibility-weighted imaging in acute ischemic stroke

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Hai-fei Jiang ◽  
Yi-qun Zhang ◽  
Jiang-xia Pang ◽  
Pei-ning Shao ◽  
Han-cheng Qiu ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Large Vessel ◽  
Susceptibility Weighted Imaging ◽  
Regression Analyses ◽  
Large Vessel Occlusion ◽  
Stroke Patients ◽  
Anterior Circulation ◽  
Vessel Occlusion ◽  
Factors Associated ◽  
Univariate Analyses

AbstractThe prominent vessel sign (PVS) on susceptibility-weighted imaging (SWI) is not displayed in all cases of acute ischemia. We aimed to investigate the factors associated with the presence of PVS in stroke patients. Consecutive ischemic stroke patients admitted within 24 h from symptom onset underwent emergency multimodal MRI at admission. Associated factors for the presence of PVS were analyzed using univariate analyses and multivariable logistic regression analyses. A total of 218 patients were enrolled. The occurrence rate of PVS was 55.5%. Univariate analyses showed significant differences between PVS-positive group and PVS-negative group in age, history of coronary heart disease, baseline NIHSS scores, total cholesterol, hemoglobin, anterior circulation infarct, large vessel occlusion, and cardioembolism. Multivariable logistic regression analyses revealed that the independent factors associated with PVS were anterior circulation infarct (odds ratio [OR] 13.7; 95% confidence interval [CI] 3.5–53.3), large vessel occlusion (OR 123.3; 95% CI 33.7–451.5), and cardioembolism (OR 5.6; 95% CI 2.1–15.3). Anterior circulation infarct, large vessel occlusion, and cardioembolism are independently associated with the presence of PVS on SWI.

Feasibility and safety of the strategy of first stenting without retrieval using Solitaire FR as a treatment for emergent large-vessel occlusion due to underlying intracranial atherosclerosis

2021 ◽  
pp. 1-9
Author(s):  
Jong-Hoon Kim ◽  
Young-Jin Jung ◽  
Chul-Hoon Chang
Keyword(s):  
Functional Outcomes ◽  
Medical Center ◽  
Large Vessel ◽  
Intracranial Atherosclerosis ◽  
Large Vessel Occlusion ◽  
Anterior Circulation ◽  
Vessel Occlusion ◽  
Stent Retrieval ◽  
First Pass ◽  
Solitaire Fr

OBJECTIVEThe optimal treatment for underlying intracranial atherosclerosis (ICAS) in patients with emergent large-vessel occlusion (ELVO) remains unclear. Reocclusion during endovascular treatment (EVT) occurs frequently (57.1%–77.3%) after initial recanalization with stent retriever (SR) thrombectomy in ICAS-related ELVO. This study aimed to compare treatment outcomes of the strategy of first stenting without retrieval (FRESH) using the Solitaire FR versus SR thrombectomy in patients with ICAS-related ELVO.METHODSThe authors retrospectively reviewed consecutive patients with acute ischemic stroke and intracranial ELVO of the anterior circulation who underwent EVT between January 2017 and December 2019 at Yeungnam University Medical Center. Large-vessel occlusion (LVO) of the anterior circulation was classified by etiology as follows: 1) no significant stenosis after recanalization (embolic group) and 2) remnant stenosis > 70% or lesser degree of stenosis with a tendency toward reocclusion and/or flow impairment during EVT (ICAS group). The ICAS group was divided into the SR thrombectomy group (SR thrombectomy) and the FRESH group.RESULTSA total of 105 patients (62 men and 43 women; median age 71 years, IQR 62.5–79 years) were included. The embolic, SR thrombectomy, and FRESH groups comprised 66 (62.9%), 26 (24.7%), and 13 (12.4%) patients, respectively. There were no significant differences between the SR thrombectomy and FRESH groups in symptom onset–to-door time, but puncture-to-recanalization time was significantly shorter in the latter group (39 vs 54 minutes, p = 0.032). There were fewer stent retrieval passes but more first-pass recanalizations in the FRESH group (p < 0.001). Favorable functional outcomes were significantly more frequent in the FRESH group (84.6% vs 42.3%, p = 0.017).CONCLUSIONSThis study’s findings suggest that FRESH, rather than rescue stenting, could be a treatment option for ICAS-related ELVO.

Abstract 31: Penumbra Consumption Rates Based on T Max Delay and Reperfusion Status: A Post-Hoc Analysis of the Defuse-3 Trial

Stroke ◽  
2021 ◽  
Vol 52 (Suppl_1) ◽  
Author(s):  
Shadi Yaghi ◽  
Eytan Raz ◽  
Seena Dehkharghani ◽  
Howard Riina ◽  
Ryan McTaggart ◽  
...  
Keyword(s):  
Infarct Volume ◽  
Secondary Analysis ◽  
Large Vessel ◽  
Large Vessel Occlusion ◽  
Anterior Circulation ◽  
Vessel Occlusion ◽  
Consumption Rates ◽  
Definition Of ◽  
Post Hoc ◽  
New Generation

Introduction: In patients with acute large vessel occlusion, the definition of penumbral tissue based on T max delay perfusion imaging is not well established in relation to late-window endovascular thrombectomy (EVT). In this study, we sought to evaluate penumbra consumption rates for T max delays in patients treated between 6 and 16 hours from last known normal. Methods: This is a secondary analysis of the DEFUSE-3 trial, which included patients with an acute ischemic stroke due to anterior circulation occlusion within 6-16 hours of last known normal. The primary outcome is percentage penumbra consumption defined as (24 hour infarct volume-core infarct volume)/(Tmax volume-baseline core volume). We stratified the cohort into 4 categories (untreated, TICI 0-2a, TICI 2b, and TICI3) and calculated penumbral consumption rates. Results: We included 143 patients, of which 66 were untreated, 16 had TICI 0-2a, 46 had TICI 2b, and 15 had TICI 3. In untreated patients, a median (IQR) of 48% (21% - 85%) of penumbral tissue was consumed based on Tmax6 as opposed to 160.6% (51% - 455.2%) of penumbral tissue based on Tmax10. On the contrary, in patients achieving TICI 3 reperfusion, a median (IQR) of 5.3% (1.1% - 14.6%) of penumbral tissue was consumed based on Tmax6 and 25.7% (3.2% - 72.1%) of penumbral tissue based on Tmax10. Conclusion: Contrary to prior studies, we show that at least 75% of penumbral tissue with Tmax > 10 sec delay can be salvaged with successful reperfusion and new generation devices. In untreated patients, since infarct expansion can occur beyond 24 hours, future studies with delayed brain imaging are needed to determine the optimal T max delay threshold that defines penumbral tissue in patients with proximal anterior circulation large vessel occlusion.

Sign in / Sign up

Export Citation Format

Share Document