Hypoperfusion Intensity Ratio Predicts Malignant Edema and Functional Outcome in Large-Vessel Occlusive Stroke with Poor Revascularization.

BACKGROUND AND OBJECTIVE: Malignant cerebral edema (MCE) is a well-known complication in patients with acute ischemic stroke with core infarcts ≥ 80 mL caused by large-vessel occlusions. MCE can also develop in patients with smaller infarcts with moderate -to-large volume of tissue at risk who do not achieve successful revascularization with endovascular thrombectomy (ET). Features that predict the development of MCE in this population are not well-described. We aim to identify predictors of MCE and 90-day functional outcome in stroke patients with an anterior circulation large vessel occlusion (LVO) and a < 80 mL ischemic core who do not achieve complete reperfusion. METHODS: We reviewed our institutional stroke registry and included patients who achieved unsuccessful revascularization, mTICI 0-2a, after ET and whose baseline imaging was notable for a core infarct < 80 mL, a Tmax > 6 s volume ≥ 80 mL, and a mismatch ratio ≥ 1.8. MCE was defined as ≥ 5 mm of midline shift on follow-up imaging, obtained 6-48 h after the pre-ET perfusion scan. RESULTS: Thirty-six patients met inclusion criteria. Unadjusted analysis demonstrated that younger age, higher systolic blood pressure, larger core volume, and higher hypoperfusion intensity ratio (HIR) were associated with MCE (all p < 0.02). In multivariate logistic regression analysis, age, HIR, and core infarct volume were independent predictors of MCE. The optimal HIR threshold to predict MCE was ≥ 0.54 (OR 14.7, 95% CI 2.4-78.0, p = 0.003). HIR was also associated with 3-month mRS (HIR ≥ 0.54 for mRS of 3-6: OR 10.8, 95% CI 1.9-44.0, p = 0.02). CONCLUSIONS: Younger age, larger core infarct volume, and higher HIR are predictive of MCE in patients with anterior circulation LVO, moderate-to-large tissue at risk, and suboptimal revascularization. HIR is correlated with three-month functional outcomes.

Artificial intelligence to diagnose ischemic stroke and identify large vessel occlusions: a systematic review.

BACKGROUND AND PURPOSE: Acute stroke caused by large vessel occlusions (LVOs) requires emergent detection and treatment by endovascular thrombectomy. However, radiologic LVO detection and treatment is subject to variable delays and human expertise, resulting in morbidity. Imaging software using artificial intelligence (AI) and machine learning (ML), a branch of AI, may improve rapid frontline detection of LVO strokes. This report is a systematic review of AI in acute LVO stroke identification and triage, and characterizes LVO detection software. METHODS: A systematic review of acute stroke diagnostic-focused AI studies from January 2014 to February 2019 in PubMed, Medline, and Embase using terms: 'artificial intelligence' or 'machine learning or deep learning' and 'ischemic stroke' or 'large vessel occlusion' was performed. RESULTS: Variations of AI, including ML methods of random forest learning (RFL) and convolutional neural networks (CNNs), are used to detect LVO strokes. Twenty studies were identified that use ML. Alberta Stroke Program Early CT Score (ASPECTS) commonly used RFL, while LVO detection typically used CNNs. Image feature detection had greater sensitivity with CNN than with RFL, 85% versus 68%. However, AI algorithm performance metrics use different standards, precluding ideal objective comparison. Four current software platforms incorporate ML: Brainomix (greatest validation of AI for ASPECTS, uses CNNs to automatically detect LVOs), General Electric, iSchemaView (largest number of perfusion study validations for thrombectomy), and Viz.ai (uses CNNs to automatically detect LVOs, then automatically activates emergency stroke treatment systems). CONCLUSIONS: AI may improve LVO stroke detection and rapid triage necessary for expedited treatment. Standardization of performance assessment is needed in future studies.