OBJECTIVE: The role of heparin in acute ischemic stroke is controversial. We investigated the effect of heparin on ischemic lesion growth. METHODS: Data were analyzed on nonthrombolyzed ischemic stroke patients in whom diffusion-weighted imaging (DWI)/perfusion-weighted imaging (PWI) MRI was performed less than 12 hours of last known well and showed a PWI-DWI lesion mismatch, and who underwent follow-up neuroimaging at least 4 days after admission. Lesion growth was assessed by (1) absolute lesion growth and (2) percentage mismatch lost (PML). Univariate and multivariate regression analysis, and propensity score matching, were used to determine the effects of heparin on ischemic lesion growth. RESULTS: Of the 113 patients meeting study criteria, 59 received heparin within 24 hours. Heparin use was associated with ∼5-fold reductions in PML (3.5% versus 19.2%, P = .002) and absolute lesion growth (4.7 versus 20.5 mL, P = .009). In multivariate regression models, heparin independently predicted reduced PML (P = .04) and absolute lesion growth (P = .04) in the entire cohort, and in multiple subgroups (patients with and without proximal artery occlusion; DWI volume greater than 5 mL; cardio-embolic mechanism; DEFUSE-3 target mismatch). In propensity score matching analysis where patients were matched by admission NIHSS, DWI volume and proximal artery occlusion, heparin remained an independent predictor of PML (P = .048) and tended to predict absolute lesion growth (P = .06). Heparin treatment did not predict functional outcome at discharge or 90 days. CONCLUSION: Early heparin treatment in acute ischemic stroke patients with PWI-DWI mismatch attenuates ischemic lesion growth. Clinical trials with careful patient selection are warranted to investigate the potential ischemic protective effects of heparin.

Background and Purpose- We evaluated deep learning algorithms' segmentation of acute ischemic lesions on heterogeneous multi-center clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. Methods- Ischemic stroke data sets from the MRI-GENIE (MRI-Genetics Interface Exploration) repository consisting of 12 international genetic research centers were retrospectively analyzed using an automated deep learning segmentation algorithm consisting of an ensemble of 3-dimensional convolutional neural networks. Three ensembles were trained using data from the following: (1) 267 patients from an independent single-center cohort, (2) 267 patients from MRI-GENIE, and (3) mixture of (1) and (2). The algorithms' performances were compared against manual outlines from a separate 383 patient subset from MRI-GENIE. Univariable and multivariable logistic regression with respect to demographics, stroke subtypes, and vascular risk factors were performed to identify phenotypes associated with large acute diffusion-weighted MRI volumes and greater stroke severity in 2770 MRI-GENIE patients. Stroke topography was investigated. Results- The ensemble consisting of a mixture of MRI-GENIE and single-center convolutional neural networks performed best. Subset analysis comparing automated and manual lesion volumes in 383 patients found excellent correlation (ρ=0.92; P<0.0001). Median (interquartile range) diffusion-weighted MRI lesion volumes from 2770 patients were 3.7 cm (0.9-16.6 cm). Patients with small artery occlusion stroke subtype had smaller lesion volumes ( P<0.0001) and different topography compared with other stroke subtypes. Conclusions- Automated accurate clinical diffusion-weighted MRI lesion segmentation using deep learning algorithms trained with multi-center and diverse data is feasible. Both lesion volume and topography can provide insight into stroke subtypes with sufficient sample size from big heterogeneous multi-center clinical imaging phenotype data sets.

OBJECTIVES: To assess whether plasma biomarkers of oxidative stress predict diffusion-perfusion mismatch in patients with acute ischemic stroke (AIS). METHODS: We measured plasma levels of oxidative stress biomarkers such as F2-isoprostanes (F2-isoPs), total and perchloric acid Oxygen Radical Absorbance Capacity (ORAC and ORAC), urinary levels of 8-oxo-7,8-dihydro-2'-deoxyguoanosine, and inflammatory and tissue-damage biomarkers (high-sensitivity C-reactive protein, matrix metalloproteinase-2 and -9) in a prospective study of patients with AIS presenting within 9 hours of symptom onset. Diffusion-weighted (DWI) and perfusion-weighted (PWI) MRI sequences were analyzed with a semiautomated volumetric method. Mismatch was defined as baseline mean transit time volume minus DWI volume. A percent mismatch cutoff of >20% was considered clinically significant. A stricter definition of mismatch was also used. Mismatch salvage was the region free of overlap by final infarction. RESULTS: Mismatch >20% was present in 153 of 216 (70.8%) patients (mean [±SD] age 69.2 ± 14.3 years, 41.2% women). Patients with mismatch >20% were more likely to have higher baseline plasma levels of ORAC ( = 0.020) and F2-isoPs ( = 0.145). Multivariate binary logistic regression demonstrated that lnF2-isoP (odds ratio [OR] 2.44, 95% confidence interval [CI] 1.19-4.98, = 0.014) and lnORAC (OR 4.18, 95% CI 1.41-12.41, = 0.010) were independent predictors of >20% PWI-DWI mismatch and the stricter mismatch definition, respectively. lnORAC significantly predicted mismatch salvage volume (>20% mismatch = 0.010, stricter mismatch definition = 0.003). CONCLUSIONS: Elevated hyperacute plasma levels of F2-isoP and ORAC are associated with radiographic evidence of mismatch and mismatch salvage in patients with AIS. If validated, these findings may add to our understanding of the role of oxidative stress in cerebral tissue fate during acute ischemia.

Chronic white matter structural injury is a risk factor for poor long-term outcomes after acute ischemic stroke (AIS). However, it is unclear how white matter structural injury predisposes to poor outcomes after AIS. To explore this question, in 42 AIS patients with moderate to severe white matter hyperintensity (WMH) burden, we characterized WMH and normal-appearing white matter (NAWM) diffusivity anisotropy metrics in the hemisphere contralateral to acute ischemia in relation to ischemic tissue and early functional outcomes. All patients underwent brain MRI with dynamic susceptibility contrast perfusion and diffusion tensor imaging within 12 h and at day 3-5 post stroke. Early neurological outcomes were measured as the change in NIH Stroke Scale score from admission to day 3-5 post stroke. Target mismatch profile, percent mismatch lost, infarct growth, and rates of good perfusion were measured to assess ischemic tissue outcomes. NAWM mean diffusivity was significantly lower in the group with early neurological improvement (ENI, 0.79 vs. 0.82 × 10, mm/s; P = 0.02). In multivariable logistic regression, NAWM mean diffusivity was an independent radiographic predictor of ENI (β = - 17.6, P = 0.037). Median infarct growth was 118% (IQR 26.8-221.9%) despite good reperfusion being observed in 65.6% of the cohort. NAWM and WMH diffusivity metrics were not associated with target mismatch profile, percent mismatch lost, or infarct growth. Our results suggest that, in AIS patients, white matter structural integrity is associated with poor early neurological outcomes independent of ischemic tissue outcomes.

White matter hyperintensity (WMH) burden is a critically important cerebrovascular phenotype linked to prediction of diagnosis and prognosis of diseases, such as acute ischemic stroke (AIS). However, current approaches to its quantification on clinical MRI often rely on time intensive manual delineation of the disease on T2 fluid attenuated inverse recovery (FLAIR), which hinders high-throughput analyses such as genetic discovery. In this work, we present a fully automated pipeline for quantification of WMH in clinical large-scale studies of AIS. The pipeline incorporates automated brain extraction, intensity normalization and WMH segmentation using spatial priors. We first propose a brain extraction algorithm based on a fully convolutional deep learning architecture, specifically designed for clinical FLAIR images. We demonstrate that our method for brain extraction outperforms two commonly used and publicly available methods on clinical quality images in a set of 144 subject scans across 12 acquisition centers, based on dice coefficient (median 0.95; inter-quartile range 0.94-0.95; p < 0.01) and Pearson correlation of total brain volume (r = 0.90). Subsequently, we apply it to the large-scale clinical multi-site MRI-GENIE study (N = 2783) and identify a decrease in total brain volume of -2.4 cc/year. Additionally, we show that the resulting total brain volumes can successfully be used for quality control of image preprocessing. Finally, we obtain WMH volumes by building on an existing automatic WMH segmentation algorithm that delineates and distinguishes between different cerebrovascular pathologies. The learning method mimics expert knowledge of the spatial distribution of the WMH burden using a convolutional auto-encoder. This enables successful computation of WMH volumes of 2533 clinical AIS patients. We utilize these results to demonstrate the increase of WMH burden with age (0.950 cc/year) and show that single site estimates can be biased by the number of subjects recruited.

BACKGROUND AND PURPOSE: Accurate automated infarct segmentation is needed for acute ischemic stroke studies relying on infarct volumes as an imaging phenotype or biomarker that require large numbers of subjects. This study investigated whether an ensemble of convolutional neural networks trained on multiparametric DWI maps outperforms single networks trained on solo DWI parametric maps. MATERIALS AND METHODS: Convolutional neural networks were trained on combinations of DWI, ADC, and low b-value-weighted images from 116 subjects. The performances of the networks (measured by the Dice score, sensitivity, and precision) were compared with one another and with ensembles of 5 networks. To assess the generalizability of the approach, we applied the best-performing model to an independent Evaluation Cohort of 151 subjects. Agreement between manual and automated segmentations for identifying patients with large lesion volumes was calculated across multiple thresholds (21, 31, 51, and 70 cm). RESULTS: An ensemble of convolutional neural networks trained on DWI, ADC, and low b-value-weighted images produced the most accurate acute infarct segmentation over individual networks ( < .001). Automated volumes correlated with manually measured volumes (Spearman ρ = 0.91, < .001) for the independent cohort. For the task of identifying patients with large lesion volumes, agreement between manual outlines and automated outlines was high (Cohen κ, 0.86-0.90; < .001). CONCLUSIONS: Acute infarcts are more accurately segmented using ensembles of convolutional neural networks trained with multiparametric maps than by using a single model trained with a solo map. Automated lesion segmentation has high agreement with manual techniques for identifying patients with large lesion volumes.

To determine whether the rich-club organization, essential for information transport in the human connectome, is an important biomarker of functional outcome after acute ischemic stroke (AIS). Consecutive AIS patients ( = 344) with acute brain magnetic resonance imaging (MRI) (<48 h) were eligible for this study. Each patient underwent a clinical MRI protocol, which included diffusion weighted imaging (DWI). All DWIs were registered to a template on which rich-club regions have been defined. Using manual outlines of stroke lesions, we automatically counted the number of affected rich-club regions and assessed its effect on the National Institute of Health Stroke Scale (NIHSS) and modified Rankin Scale (mRS; obtained at 90 days post-stroke) scores through ordinal regression. Of 344 patients (median age 65, inter-quartile range 54-76 years) with a median DWI lesion volume (DWIv) of 3cc, 64% were male. We established that an increase in number of rich-club regions affected by a stroke increases the odds of poor stroke outcome, measured by NIHSS (OR: 1.77, 95%CI 1.41-2.21) and mRS (OR: 1.38, 95%CI 1.11-1.73). Additionally, we demonstrated that the OR exceeds traditional markers, such as DWIv (OR 1.08, 95%CI 1.06-1.11; OR 1.05, 95%CI 1.03-1.07) and age (OR 1.03, 95%CI 1.01-1.05; OR 1.05, 95%CI 1.03-1.07). In this proof-of-concept study, the number of rich-club nodes affected by a stroke lesion presents a translational biomarker of stroke outcome, which can be readily assessed using standard clinical AIS imaging protocols and considered in functional outcome prediction models beyond traditional factors.

Prediction of functional outcome after stroke based on initial presentation remains an open challenge, suggesting that an important aspect is missing from these prediction models. There exists the notion of a protective mechanism called brain reserve, which may be utilized to understand variations in disease outcome. In this work, we expand the concept of brain reserve (effective reserve) to improve prediction models of functional outcome after acute ischemic stroke (AIS). Consecutive AIS patients with acute brain magnetic resonance imaging (<48 hours) were eligible for this study. White matter hyperintensity and acute infarct volume were determined on T2 fluid attenuated inversion recovery and diffusion weighted images, respectively. Modified Rankin Scale scores were obtained at 90days poststroke. Effective reserve was defined as a latent variable using structural equation modeling by including age, systolic blood pressure, and intracranial volume measurements. Of 453 AIS patients (mean age 66.6 ± 14.7 years), 36% were male and 311 hypertensive. There was inverse association between effective reserve and 90-day modified Rankin Scale scores (path coefficient -0.18 ± 0.01, P < .01). Compared to a model without effective reserve, correlation between predicted and observed modified Rankin Scale scores improved in the effective-reserve-based model (Spearman's ρ 0.29 ± 0.18 versus 0.15 ± 0.17, P < .001). Furthermore, hypertensive patients exhibited lower effective reserve (P < 10). Using effective reserve in prediction models of stroke outcome is feasible and leads to better model performance. Furthermore, higher effective reserve is associated with more favorable functional poststoke outcome and might correspond to an overall better vascular health.

Background and Purpose- The ability to model long-term functional outcomes after acute ischemic stroke represents a major clinical challenge. One approach to potentially improve prediction modeling involves the analysis of connectomics. The field of connectomics represents the brain's connectivity as a graph, whose topological properties have helped uncover underlying mechanisms of brain function in health and disease. Specifically, we assessed the impact of stroke lesions on rich club organization, a high capacity backbone system of brain function. Methods- In a hospital-based cohort of 41 acute ischemic stroke patients, we investigated the effect of acute infarcts on the brain's prestroke rich club backbone and poststroke functional connectomes with respect to poststroke outcome. Functional connectomes were created using 3 anatomic atlases, and characteristic path-length () was calculated for each connectome. The number of rich club regions affected were manually determined using each patient's diffusion weighted image. We investigated differences in with respect to outcome (modified Rankin Scale score; 90 days) and the National Institutes of Health Stroke Scale (NIHSS; early: 2-5 days; late: 90-day follow-up). Furthermore, we assessed the effect of including number of rich club regions and in outcome models, using linear regression and assessing the explained variance (R). Results- Of 41 patients (mean age [range]: 70 [45-89] years), 61% were male. Lower was generally associated with better outcome. Including number of rich club regions in the backward selection models of outcome, R increased between 1.3- and 2.6-fold beyond that of traditional markers (age and acute lesion volume) for NIHSS and modified Rankin Scale score. Conclusions- In this proof-of-concept study, we showed that information on network topology can be leveraged to improve modeling of poststroke functional outcome. Future studies are warranted to validate this approach in larger prospective studies of outcome prediction in stroke.