In patients with suspected acute stroke, CT perfusion-based cerebral blood flow maps cannot substitute for DWI in measuring the ischemic core

Combining Early Ischemic Change and Collateral Extent for Functional Outcomes After Endovascular Therapy: An Analysis From AcT Trial

BACKGROUND

Early ischemic change and collateral extent are colinear with ischemic core volume (ICV). We investigated the relationship between a combined score using the Alberta Stroke Program Early Computed Tomography Score and multiphase computed tomography angiography (mCTA) collateral extent, named mCTA-ACE score, on functional outcomes in endovascular therapy-treated patients.

METHODS

We performed a post hoc analysis of a subset of endovascular therapy-treated patients from the Alteplase Compared to Tenecteplase trial which was conducted between December 2019 and January 2022 at 22 centers across Canada. Ten-point mCTA collateral corresponding to M2 to M6 regions of the Alberta Stroke Program Early Computed Tomography Score grid was evaluated as 0 (poor), 1 (moderate), or 2 (normal) and additively combined with the 10-point Alberta Stroke Program Early Computed Tomography Score to produce a 20-point mCTA-ACE score. We investigated the association of mCTA-ACE score with modified Rankin Scale score ≤2 and return to prestroke level of function at 90 to 120 days using mixed-effects logistic regression. In the subset of patients who underwent baseline computed tomography perfusion imaging, we compared the mCTA-ACE score and ICV for outcome prediction.

RESULTS

Among 1577 intention-to-treat population in the trial, 368 (23%; 179 men; median age, 73 years) were included, with Alberta Stroke Program Early Computed Tomography Score, mCTA collateral, and combination of both (mCTA-ACE score: median [interquartile range], 8 [7-10], 9 [8-10], and 17 [16-19], respectively). The probability of modified Rankin Scale score ≤2 and return to prestroke level of function increased for each 1-point increase in mCTA-ACE score (odds ratio, 1.16 [95% CI, 1.06-1.28] and 1.22 [95% CI, 1.06-1.40], respectively). Among 173 patients in whom computed tomography perfusion data was assessable, the mCTA-ACE score was inversely correlated with ICV (ρ=-0.46; P<0.01). The mCTA-ACE score was comparable to ICV to predict a modified Rankin Scale score ≤2 and return to prestroke level of function (C statistics 0.71 versus 0.69 and 0.68 versus 0.64, respectively).

CONCLUSIONS

The mCTA-ACE score had a significant positive association with functional outcomes after endovascular therapy and had a similar predictive performance as ICV.

Estimating the volume of penumbra in rodents using DTI and stack-based ensemble machine learning framework

BACKGROUND

This study investigates the potential of diffusion tensor imaging (DTI) in identifying penumbral volume (PV) compared to the standard gadolinium-required perfusion-diffusion mismatch (PDM), utilizing a stack-based ensemble machine learning (ML) approach with enhanced explainability.

METHODS

Sixteen male rats were subjected to middle cerebral artery occlusion. The penumbra was identified using PDM at 30 and 90 min after occlusion. We used 11 DTI-derived metrics and 14 distance-based features to train five voxel-wise ML models. The model predictions were integrated using stack-based ensemble techniques. ML-estimated and PDM-defined PVs were compared to evaluate model performance through volume similarity assessment, the Pearson correlation analysis, and Bland-Altman analysis. Feature importance was determined for explainability.

RESULTS

In the test rats, the ML-estimated median PV was 106.4 mL (interquartile range 44.6-157.3 mL), whereas the PDM-defined median PV was 102.0 mL (52.1-144.9 mL). These PVs had a volume similarity of 0.88 (0.79-0.96), a Pearson correlation coefficient of 0.93 (p < 0.001), and a Bland-Altman bias of 2.5 mL (2.4% of the mean PDM-defined PV), with 95% limits of agreement ranging from -44.9 to 49.9 mL. Among the features used for PV prediction, the mean diffusivity was the most important feature.

CONCLUSIONS

Our study confirmed that PV can be estimated using DTI metrics with a stack-based ensemble ML approach, yielding results comparable to the volume defined by the standard PDM. The model explainability enhanced its clinical relevance. Human studies are warranted to validate our findings.

RELEVANCE STATEMENT

The proposed DTI-based ML model can estimate PV without the need for contrast agent administration, offering a valuable option for patients with kidney dysfunction. It also can serve as an alternative if perfusion map interpretation fails in the clinical setting.

KEY POINTS

• Penumbral volume can be estimated by DTI combined with stack-based ensemble ML. • Mean diffusivity was the most important feature used for predicting penumbral volume. • The proposed approach can be beneficial for patients with kidney dysfunction.

Usefulness of Different Imaging Methods in the Diagnosis of Cerebral Vasculopathy

Assessment of cerebral vasculopathies is challenging and requires understanding the utility of different imaging methods. Various techniques are available to image the vessel lumen, each with unique advantages and disadvantages. Bolus-based CT and MR angiography requires careful timing of a contrast bolus to provide optimal luminal enhancement. Non-contrast MRA techniques do not require a contrast agent and can provide images with little venous contamination. Digital subtraction angiography remains the gold standard but is invasive, while VW-MRI provides a non-invasive way of assessing vessel wall pathology. Conventional brain MRI has high sensitivity in the diagnosis of vasculitis but findings are nonspecific.

Overestimation of ischemic core on baseline MRI in acute stroke

BACKGROUND AND PURPOSE

In patients with acute ischemic stroke (AIS), overestimation of ischemic core on MRI-DWI has been described primarily in regions with milder reduced diffusion. We aimed to assess the possibility of ischemic core overestimation on pretreatment MRI despite using more restricted reduced diffusion (apparent diffusion coefficient (ADC) ≤620 × 10-6 mm2/s) in AIS patients with successful reperfusion.

MATERIALS AND METHODS

In this retrospective single institutional study, AIS patients who had pretreatment MRI underwent successful reperfusion and had follow-up MRI to determine the final infarct volume were reviewed. Pretreatment ischemic core and final infarction volumes were calculated. Ghost core was defined as overestimation of final infarct volume by baseline MRI of >10 mL. Baseline clinical, demographic, and treatment-related factors in this cohort were reviewed.

RESULTS

A total of 6/156 (3.8%) patients had overestimated ischemic core volume on baseline MRI, with mean overestimation of 65.6 mL. Three out of six patients had pretreatment ischemic core estimation of >70 mL, while the final infarct volume was <70 mL. All six patients had last known well-to-imaging <120 min, median (IQR): 65 (53-81) minutes.

CONCLUSIONS

Overestimation of ischemic core, known as ghost core, is rare using severe ADC threshold (≤620 × 10-6 mm2/s), but it does occur in nearly 1 of every 25 patients, confined to hyperacute patients imaged within 120 min of symptom onset. Awareness of this phenomenon carries implications for treatment and trial enrollment.

A deep learning and radiomics based Alberta stroke program early CT score method on CTA to evaluate acute ischemic stroke

BACKGROUND

Alberta stroke program early CT score (ASPECTS) is a semi-quantitative evaluation method used to evaluate early ischemic changes in patients with acute ischemic stroke, which can guide physicians in treatment decisions and prognostic judgments.

OBJECTIVE

We propose a method combining deep learning and radiomics to alleviate the problem of large inter-observer variance in ASPECTS faced by physicians and assist them to improve the accuracy and comprehensiveness of the ASPECTS.

METHODS

Our study used a brain region segmentation method based on an improved encoding-decoding network. Through the deep convolutional neural network, 10 regions defined for ASPECTS will be obtained. Then, we used Pyradiomics to extract features associated with cerebral infarction and select those significantly associated with stroke to train machine learning classifiers to determine the presence of cerebral infarction in each scored brain region.

RESULTS

The experimental results show that the Dice coefficient for brain region segmentation reaches 0.79. Three radioactive features are selected to identify cerebral infarction in brain regions, and the 5-fold cross-validation experiment proves that these 3 features are reliable. The classifier trained based on 3 features reaches prediction performance of AUC = 0.95. Moreover, the intraclass correlation coefficient of ASPECTS between those obtained by the automated ASPECTS method and physicians is 0.86 (95% confidence interval, 0.56-0.96).

CONCLUSIONS

This study demonstrates advantages of using a deep learning network to replace the traditional template registration for brain region segmentation, which can determine the shape and location of each brain region more precisely. In addition, a new brain region classifier based on radiomics features has potential to assist physicians in clinical stroke detection and improve the consistency of ASPECTS.

Automated advanced imaging in acute ischemic stroke. Certainties and uncertainties

The purpose of this is study was to review pearls and pitfalls of advanced imaging, such as computed tomography perfusion and diffusion-weighed imaging and perfusion-weighted imaging in the selection of acute ischemic stroke (AIS) patients suitable for endovascular treatment (EVT) in the late time window (6-24 h from symptom onset). Advanced imaging can quantify infarct core and ischemic penumbra using specific threshold values and provides optimal selection parameters, collectively called target mismatch. More precisely, target mismatch criteria consist of core volume and/or penumbra volume and mismatch ratio (the ratio between total hypoperfusion and core volumes) with precise cut-off values. The parameters of target mismatch are automatically calculated with dedicated software packages that allow a quick and standardized interpretation of advanced imaging. However, this approach has several limitations leading to a misclassification of core and penumbra volumes. In fact, automatic software platforms are affected by technical artifacts and are not interchangeable due to a remarkable vendor-dependent variability, resulting in different estimate of target mismatch parameters. In addition, advanced imaging is not completely accurate in detecting infarct core, that can be under- or overestimated. Finally, the selection of candidates for EVT remains currently suboptimal due to the high rates of futile reperfusion and overselection caused by the use of very stringent inclusion criteria. For these reasons, some investigators recently proposed to replace advanced with conventional imaging in the selection for EVT, after the demonstration that non-contrast CT ASPECTS and computed tomography angiography collateral evaluation are not inferior to advanced images in predicting outcome in AIS patients treated with EVT. However, other authors confirmed that CTP and PWI/DWI postprocessed images are superior to conventional imaging in establishing the eligibility of patients for EVT. Therefore, the routine application of automatic assessment of advanced imaging remains a matter of debate. Recent findings suggest that the combination of conventional and advanced imaging might improving our selection criteria.

MRI-based parameter inference for cerebral perfusion modelling in health and ischaemic stroke

Cerebral perfusion modelling is a promising tool to predict the impact of acute ischaemic stroke treatments on the spatial distribution of cerebral blood flow (CBF) in the human brain. To estimate treatment efficacy based on CBF, perfusion simulations need to become suitable for group-level investigations and thus account for physiological variability between individuals. However, computational perfusion modelling to date has been restricted to a few patient-specific cases. This study set out to establish automated parameter inference for perfusion modelling based on neuroimaging data and thus enable CBF simulations of groups. Magnetic resonance imaging (MRI) data from 75 healthy senior adults were utilised. Brain geometries were computed from healthy reference subjects' T1-weighted MRI. Haemodynamic model parameters were determined from spatial CBF maps measured by arterial spin labelling (ASL) perfusion MRI. Thereafter, perfusion simulations were conducted in 75 healthy cases followed by 150 acute ischaemic stroke cases representing an occlusion and CBF cessation in the left and right middle cerebral arteries. The anatomical fitness of the brain geometries was evaluated by comparing the simulated grey (GM) and white matter (WM) volumes to measurements in healthy reference subjects. Strong positive correlations were found in both tissue types (GM: Pearson's r 0.74, P<0.001; WM: Pearson's r 0.84, P<0.001). Haemodynamic parameter tuning was verified by comparing the total volumetric blood flow rate to the brain in healthy reference subjects and simulations (Pearson's r 0.89, P<0.001). In acute ischaemic stroke cases, the simulated infarct volume using a perfusion-based estimate was 197±25 ml. Computational predictions were in agreement with anatomical and haemodynamic values from the literature concerning T1-weighted, T2-weighted, and phase-contrast MRI measurements in healthy scenarios and acute ischaemic stroke cases. The acute stroke simulations did not capture small infarcts (left tail of the distribution), which could be explained by neglected compensatory mechanisms, e.g. collaterals. The proposed parameter inference method provides a foundation for group-level CBF simulations and for in silico clinical stroke trials which could assist in medical device and drug development.

Assessing the diagnostic accuracy of CT perfusion: a systematic review

Background and purpose

Computed tomography perfusion (CTP) has successfully extended the time window for reperfusion therapies in ischemic stroke. However, the published perfusion parameters and thresholds vary between studies. Using Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy Studies (PRISMA-DTA) guidelines, we conducted a systematic review to investigate the accuracy of parameters and thresholds for identifying core and penumbra in adult stroke patients.

Methods

We searched Medline, Embase, the Cochrane Library, and reference lists of manuscripts up to April 2022 using the following terms "computed tomography perfusion," "stroke," "infarct," and "penumbra." Studies were included if they reported perfusion thresholds and undertook co-registration of CTP to reference standards. The quality of studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool and Standards for Reporting of Diagnostic Accuracy (STARD) guidelines.

Results

A total of 24 studies were included. A meta-analysis could not be performed due to insufficient data and significant heterogeneity in the study design. When reported, the mean age was 70.2 years (SD+/-3.69), and the median NIHSS on admission was 15 (IQR 13-17). The perfusion parameter identified for the core was relative cerebral blood flow (rCBF), with a median threshold of <30% (IQR 30, 40%). However, later studies reported lower thresholds in the early time window with rapid reperfusion (median 25%, IQR 20, 30%). A total of 15 studies defined a single threshold for all brain regions irrespective of collaterals and the gray and white matter.

Conclusion

A single threshold and parameter may not always accurately differentiate penumbra from core and oligemia. Further refinement of parameters is needed in the current era of reperfusion therapy.

FDA-approved machine learning algorithms in neuroradiology: A systematic review of the current evidence for approval

Over the past decade, machine learning (ML) and artificial intelligence (AI) have become increasingly prevalent in the medical field. In the United States, the Food and Drug Administration (FDA) is responsible for regulating AI algorithms as "medical devices" to ensure patient safety. However, recent work has shown that the FDA approval process may be deficient. In this study, we evaluate the evidence supporting FDA-approved neuroalgorithms, the subset of machine learning algorithms with applications in the central nervous system (CNS), through a systematic review of the primary literature. Articles covering the 53 FDA-approved algorithms with applications in the CNS published in PubMed, EMBASE, Google Scholar and Scopus between database inception and January 25, 2022 were queried. Initial searches identified 1505 studies, of which 92 articles met the criteria for extraction and inclusion. Studies were identified for 26 of the 53 neuroalgorithms, of which 10 algorithms had only a single peer-reviewed publication. Performance metrics were available for 15 algorithms, external validation studies were available for 24 algorithms, and studies exploring the use of algorithms in clinical practice were available for 7 algorithms. Papers studying the clinical utility of these algorithms focused on three domains: workflow efficiency, cost savings, and clinical outcomes. Our analysis suggests that there is a meaningful gap between the FDA approval of machine learning algorithms and their clinical utilization. There appears to be room for process improvement by implementation of the following recommendations: the provision of compelling evidence that algorithms perform as intended, mandating minimum sample sizes, reporting of a predefined set of performance metrics for all algorithms and clinical application of algorithms prior to widespread use. This work will serve as a baseline for future research into the ideal regulatory framework for AI applications worldwide.

Predictors of ghost infarct core on baseline computed tomography perfusion in stroke patients with successful recanalization after mechanical thrombectomy

OBJECTIVES

To assess the predictors of ghost infarct core (GIC) in stroke patients achieving successful recanalization after mechanical thrombectomy (MT), based on final infarct volume (FIV) calculated from follow-up diffusion-weighted imaging (DWI).

METHODS

A total of 115 consecutive stroke patients who had undergone baseline computed tomography perfusion (CTP) scan, achieved successful recanalization after MT, and finished follow-up DWI evaluation were retrospectively enrolled. Ischemic core volume was automatically generated from baseline CTP, and FIV was determined manually based on follow-up DWI. Stroke-related risk factors and demographic, clinical, imaging, and procedural data were collected and assessed. Univariate and multivariate analyses were applied to identify the predictors of GIC.

RESULTS

Of the 115 included patients (31 women and 84 men; median age, 66 years), 18 patients (15.7%) showed a GIC. The GIC group showed significantly shorter time interval from stroke onset to CTP scan and that from stroke onset to recanalization (both p < 0.001), but higher ischemic core volume (p < 0.001), hypoperfused area volume (p < 0.001), mismatch area volume (p = 0.006), and hypoperfusion ratio (p = 0.001) than the no-GIC group. In multivariate analysis, time interval from stroke onset to CTP scan (odds ratio [OR], 0.983; p = 0.005) and ischemic core volume (OR, 1.073; p < 0.001) were independently associated with the occurrence of GIC.

CONCLUSIONS

In stroke patients achieving successful recanalization after MT, time interval from stroke onset to CTP and ischemic core volume are associated with the occurrence of GIC. Patients cannot be excluded from MT solely based on baseline CTP-derived ischemic core volume, especially for patients with a shorter onset time.

KEY POINTS

• Ghost infarct core (GIC) was found in 15.7% of patients with acute ischemic stroke (AIS) in our study cohort. • GIC was associated with stroke onset time, volumetric parameters derived from CTP, and collateral status indicated by HIR. • Time interval from stroke onset to CTP scan and ischemic core volume were independent predictors of GIC.

Radiomics Analysis of Diffusion-Weighted Imaging and Long-Term Unfavorable Outcomes Risk for Acute Stroke

BACKGROUND

Diffusion-weighted imaging radiomics could be used as prognostic biomarkers in acute ischemic stroke. We aimed to identify a clinical and diffusion-weighted imaging radiomics model for individual unfavorable outcomes risk assessment in acute ischemic stroke.

METHODS

A total of 1716 patients with acute ischemic stroke from 2 centers were divided into a training cohort and a validation cohort. Patient outcomes were measured with the modified Rankin Scale score. An unfavorable outcome was defined as a modified Rankin Scale score greater than 2. The primary end point was all-cause mortality or outcomes 1 year after stroke. The MRI-DRAGON score was calculated based on previous publications. We extracted and selected the infarct features on diffusion-weighted imaging to construct a radiomic signature. The clinic-radiomics signature was built by measuring the Cox proportional risk regression score (CrrScore) and compared with the MRI-DRAGON score and the ClinicScore. CrrScore model performance was estimated by 1-year unfavorable outcomes prediction.

RESULTS

A high radiomic signature predicted a higher probability of unfavorable outcomes than a low radiomic signature in the training (hazard ratio, 3.19 [95% CI, 2.51-4.05]; P<0.0001) and validation (hazard ratio, 3.25 [95% CI, 2.20-4.80]; P<0.0001) cohorts. The diffusion-weighted imaging Alberta Stroke Program Early CT Score, age, glucose level before therapy, National Institutes of Health Stroke Scale score on admission, glycated hemoglobin' radiomic signature, hemorrhagic infarction, and malignant cerebral edema were associated with an unfavorable outcomes risk after multivariable adjustment. A CrrScore nomogram was developed to predict outcomes and had the best performance in the training (area under the curve, 0.862) and validation cohorts (area under the curve, 0.858). The CrrScore model time-dependent areas under the curve of the probability of unfavorable outcomes at 1 year in the training and validation cohorts were 0.811 and 0.801, respectively.

CONCLUSIONS

The CrrScore model allows the accurate prediction of patients with acute ischemic stroke outcomes and can potentially guide rehabilitation therapies for patients with different risks of unfavorable outcomes.

Optimizing the Definition of Ischemic Core in CT Perfusion: Influence of Infarct Growth and Tissue-Specific Thresholds

BACKGROUND AND PURPOSE

CTP allows estimating ischemic core in patients with acute stroke. However, these estimations have limited accuracy compared with MR imaging. We studied the effect of applying WM- and GM-specific thresholds and analyzed the infarct growth from baseline imaging to reperfusion.

MATERIALS AND METHODS

This was a single-center cohort of consecutive patients (n = 113) with witnessed strokes due to proximal carotid territory occlusions with baseline CT perfusion, complete reperfusion, and follow-up DWI. We segmented GM and WM, coregistered CTP with DWI, and compared the accuracy of the different predictions for each voxel on DWI through receiver operating characteristic analysis. We assessed the yield of different relative CBF thresholds to predict the final infarct volume and an estimated infarct growth-corrected volume (subtracting the infarct growth from baseline imaging to complete reperfusion) for a single relative CBF threshold and GM- and WM-specific thresholds.

RESULTS

The fixed threshold underestimated lesions in GM and overestimated them in WM. Double GM- and WM-specific thresholds of relative CBF were superior to fixed thresholds in predicting infarcted voxels. The closest estimations of the infarct on DWI were based on a relative CBF of 25% for a single threshold, 35% for GM, and 20% for WM, and they decreased when correcting for infarct growth: 20% for a single threshold, 25% for GM, and 15% for WM. The combination of 25% for GM and 15% for WM yielded the best prediction.

CONCLUSIONS

GM- and WM-specific thresholds result in different estimations of ischemic core in CTP and increase the global accuracy. More restrictive thresholds better estimate the actual extent of the infarcted tissue.

FLAIR vascular hyperintensities predict functional outcome after endovascular thrombectomy in patients with large ischemic cores

OBJECTIVES

To establish whether collateral circulation was associated with functional outcome in stroke patients with large infarct size (Alberta Stroke Program Early CT Score [ASPECTS] ≤ 5) undergoing endovascular thrombectomy (EVT) METHODS: Consecutive patients with acute ischemic stroke due to large-vessel occlusion in the anterior circulation and an ASPECTS of ≤ 5 were analyzed. Quantification of collateral circulation was performed using a fluid-attenuated inversion recovery vascular hyperintensity (FVH)-ASPECTS rating system (score ranging from 0 [no FVH] to 7 [FVHs abutting all ASPECTS cortical areas]) by two independent neuroradiologists. Good functional outcome was defined by modified Rankin Scale (mRS) score of 0 to 3 at 3 months. We determined the association between FVH score and clinical outcome using multivariable regression analyses.

RESULTS

A total of 139 patients (age, 63.1 ± 20.8 years; men, 51.8%) admitted between March 2012 and December 2017 were included. Good functional outcome (mRS 0-3) was observed in 65 (46.8%) patients, functional independence (mRS 0-2) was achieved in 43 (30.9%) patients, and 33 (23.7%) patients died at 90 days. The median FVH score was 4 (IQR, 3-5). FVH score was independently correlated with good outcome (adjusted OR = 1.41 [95% CI, 1.03-1.92]; p = 0.03 per 1-point increase).

CONCLUSIONS

In stroke patients with large-volume infarcts, good collaterals as measured by the FVH-ASPECTS rating system are associated with improved outcomes and may help select patients for reperfusion therapy.

KEY POINTS

• Endovascular thrombectomy can allow almost 1 in 2 patients with large infarct cores to achieve good functional outcome (modified Rankin Scale [mRS] of 0-3) and 1 in 3 patients to regain functional independence (mRS 0-2) at 3 months. • The extent of FVH score (as reflected by FLAIR vascular hyperintensity [FVH]-Alberta Stroke Program Early CT Score [ASPECTS] values) is associated with functional outcome at 3 months in this patient group.

Computed tomography angiography-based deep learning method for treatment selection and infarct volume prediction in anterior cerebral circulation large vessel occlusion

Background

Computed tomography perfusion (CTP) is the mainstay to determine possible eligibility for endovascular thrombectomy (EVT), but there is still a need for alternative methods in patient triage.

Purpose

To study the ability of a computed tomography angiography (CTA)-based convolutional neural network (CNN) method in predicting final infarct volume in patients with large vessel occlusion successfully treated with endovascular therapy.

Materials and Methods

The accuracy of the CTA source image-based CNN in final infarct volume prediction was evaluated against follow-up CT or MR imaging in 89 patients with anterior circulation ischemic stroke successfully treated with EVT as defined by Thrombolysis in Cerebral Infarction category 2b or 3 using Pearson correlation coefficients and intraclass correlation coefficients. Convolutional neural network performance was also compared to a commercially available CTP-based software (RAPID, iSchemaView).

Results

A correlation with final infarct volumes was found for both CNN and CTP-RAPID in patients presenting 6–24 h from symptom onset or last known well, with r = 0.67 (p < 0.001) and r = 0.82 (p < 0.001), respectively. Correlations with final infarct volumes in the early time window (0–6 h) were r = 0.43 (p = 0.002) for the CNN and r = 0.58 (p < 0.001) for CTP-RAPID. Compared to CTP-RAPID predictions, CNN estimated eligibility for thrombectomy according to ischemic core size in the late time window with a sensitivity of 0.38 and specificity of 0.89.

Conclusion

A CTA-based CNN method had moderate correlation with final infarct volumes in the late time window in patients successfully treated with EVT.

[Progress in computer-assisted Alberta stroke program early computer tomography score of acute ischemic stroke based on different modal images]

Clinically, non-contrastive computed tomography (NCCT) is used to quickly diagnose the type and area of ​​stroke, and the Alberta stroke program early computer tomography score (ASPECTS) is used to guide the next treatment. However, in the early stage of acute ischemic stroke (AIS), it's difficult to distinguish the mild cerebral infarction on NCCT with the naked eye, and there is no obvious boundary between brain regions, which makes clinical ASPECTS difficult to conduct. The method based on machine learning and deep learning can help physicians quickly and accurately identify cerebral infarction areas, segment brain areas, and operate ASPECTS quantitative scoring, which is of great significance for improving the inconsistency in clinical ASPECTS. This article describes current challenges in the field of AIS ASPECTS, and then summarizes the application of computer-aided technology in ASPECTS from two aspects including machine learning and deep learning. Finally, this article summarizes and prospects the research direction of AIS-assisted assessment, and proposes that the computer-aided system based on multi-modal images is of great value to improve the comprehensiveness and accuracy of AIS assessment, which has the potential to open up a new research field for AIS-assisted assessment.

Predictive Value of CT Brain Perfusion Studies in Acute Ischemic Infarct Taking MRI Stroke Protocol As Gold Standard

Background Acute ischemic stroke is the leading cause of serious chronic disability worldwide. Imaging plays a key role in early diagnosis and intervention, thus reducing mortality and morbidity related to ischemic stroke. Computed tomography (CT) perfusion study is a valuable imaging tool for the assessment of acute infarction. The objective of this study was to determine the predictive value of CT perfusion in diagnosing acute ischemic infarction taking Magnetic Resonance Imaging (MRI) stroke protocol (including Diffusion Weighted Imaging (DWI)) as a gold standard. Methods The cross-sectional validation study was conducted at a teaching hospital in Islamabad from June 2019 to December 2019. The study comprised a total of 125 patients of either gender with suspected acute ischemic stroke. The patients were scanned for CT perfusion and MRI stroke protocol on the same day. Scans were reported separately for the detection of acute ischemic infarction by the same consultant radiologist. The predictive value of CT perfusion was calculated accordingly. Results Of the 125 patients, 58% were male and 42% were female. The age of selected patients ranged between 38 to 70 years with a mean age of 56.12 ± 9.69 years. Acute ischemic infarction was detected in 86 (69%) patients by CT perfusion study and in 120 (96%) patients by MRI stroke protocol. The positive predicted value of CT perfusion for the detection of acute infarction was calculated as 98.83 and the negative predicted value was 10.25. Conclusion CT perfusion study provides adequate sensitivity and specificity with good predictive value in the detection of acute ischemic infarct in stroke patients. This widely available and time-effective modality aids in the triage of patients for immediate endovascular intervention leading to maximal neurological benefit and improving outcomes.

Automated prediction of final infarct volume in patients with large-vessel occlusion acute ischemic stroke

OBJECTIVE

In patients with large-vessel occlusion (LVO) acute ischemic stroke (AIS), determinations of infarct size play a key role in the identification of candidates for endovascular stroke therapy (EVT). An accurate, automated method to quantify infarct at the time of presentation using widely available imaging modalities would improve screening for EVT. Here, the authors aimed to compare the performance of three measures of infarct core at presentation, including an automated method using machine learning.

METHODS

Patients with LVO AIS who underwent successful EVT at four comprehensive stroke centers were identified. Patients were included if they underwent concurrent noncontrast head CT (NCHCT), CT angiography (CTA), and CT perfusion (CTP) with Rapid imaging at the time of presentation, and MRI 24 to 48 hours after reperfusion. NCHCT scans were analyzed using the Alberta Stroke Program Early CT Score (ASPECTS) graded by neuroradiology or neurology expert readers. CTA source images were analyzed using a previously described machine learning model named DeepSymNet (DSN). Final infarct volume (FIV) was determined from diffusion-weighted MRI sequences using manual segmentation. The primary outcome was the performance of the three infarct core measurements (NCHCT-ASPECTS, CTA with DSN, and CTP-Rapid) to predict FIV, which was measured using area under the receiver operating characteristic (ROC) curve (AUC) analysis.

RESULTS

Among 76 patients with LVO AIS who underwent EVT and met inclusion criteria, the median age was 67 years (IQR 54-76 years), 45% were female, and 37% were White. The median National Institutes of Health Stroke Scale score was 16 (IQR 12-22), and the median NCHCT-ASPECTS on presentation was 8 (IQR 7-8). The median time between when the patient was last known to be well and arrival was 156 minutes (IQR 73-303 minutes), and between NCHCT/CTA/CTP to groin puncture was 73 minutes (IQR 54-81 minutes). The AUC was obtained at three different cutoff points: 10 ml, 30 ml, and 50 ml FIV. At the 50-ml FIV cutoff, the AUC of ASPECTS was 0.74; of CTP core volume, 0.72; and of DSN, 0.82. Differences in AUCs for the three predictors were not significant for the three FIV cutoffs.

CONCLUSIONS

In a cohort of patients with LVO AIS in whom reperfusion was achieved, determinations of infarct core at presentation by NCHCT-ASPECTS and a machine learning model analyzing CTA source images were equivalent to CTP in predicting FIV. These findings have suggested that the information to accurately predict infarct core in patients with LVO AIS was present in conventional imaging modalities (NCHCT and CTA) and accessible by machine learning methods.

On the Sensitivity Analysis of Porous Finite Element Models for Cerebral Perfusion Estimation

Computational physiological models are promising tools to enhance the design of clinical trials and to assist in decision making. Organ-scale haemodynamic models are gaining popularity to evaluate perfusion in a virtual environment both in healthy and diseased patients. Recently, the principles of verification, validation, and uncertainty quantification of such physiological models have been laid down to ensure safe applications of engineering software in the medical device industry. The present study sets out to establish guidelines for the usage of a three-dimensional steady state porous cerebral perfusion model of the human brain following principles detailed in the verification and validation (V&V 40) standard of the American Society of Mechanical Engineers. The model relies on the finite element method and has been developed specifically to estimate how brain perfusion is altered in ischaemic stroke patients before, during, and after treatments. Simulations are compared with exact analytical solutions and a thorough sensitivity analysis is presented covering every numerical and physiological model parameter. The results suggest that such porous models can approximate blood pressure and perfusion distributions reliably even on a coarse grid with first order elements. On the other hand, higher order elements are essential to mitigate errors in volumetric blood flow rate estimation through cortical surface regions. Matching the volumetric flow rate corresponding to major cerebral arteries is identified as a validation milestone. It is found that inlet velocity boundary conditions are hard to obtain and that constant pressure inlet boundary conditions are feasible alternatives. A one-dimensional model is presented which can serve as a computationally inexpensive replacement of the three-dimensional brain model to ease parameter optimisation, sensitivity analyses and uncertainty quantification. The findings of the present study can be generalised to organ-scale porous perfusion models. The results increase the applicability of computational tools regarding treatment development for stroke and other cerebrovascular conditions.

Ischemic Core Overestimation on Computed Tomography Perfusion

[Figure: see text].

A porous circulation model of the human brain for in silico clinical trials in ischaemic stroke

The advancement of ischaemic stroke treatment relies on resource-intensive experiments and clinical trials. In order to improve ischaemic stroke treatments, such as thrombolysis and thrombectomy, we target the development of computational tools for in silico trials which can partially replace these animal and human experiments with fast simulations. This study proposes a model that will serve as part of a predictive unit within an in silico clinical trial estimating patient outcome as a function of treatment. In particular, the present work aims at the development and evaluation of an organ-scale microcirculation model of the human brain for perfusion prediction. The model relies on a three-compartment porous continuum approach. Firstly, a fast and robust method is established to compute the anisotropic permeability tensors representing arterioles and venules. Secondly, vessel encoded arterial spin labelling magnetic resonance imaging and clustering are employed to create an anatomically accurate mapping between the microcirculation and large arteries by identifying superficial perfusion territories. Thirdly, the parameter space of the problem is reduced by analysing the governing equations and experimental data. Fourthly, a parameter optimization is conducted. Finally, simulations are performed with the tuned model to obtain perfusion maps corresponding to an open and an occluded (ischaemic stroke) scenario. The perfusion map in the occluded vessel scenario shows promising qualitative agreement with computed tomography images of a patient with ischaemic stroke caused by large vessel occlusion. The results highlight that in the case of vessel occlusion (i) identifying perfusion territories is essential to capture the location and extent of underperfused regions and (ii) anisotropic permeability tensors are required to give quantitatively realistic estimation of perfusion change. In the future, the model will be thoroughly validated against experiments.

Computed tomography-based triage of extensive baseline infarction: ASPECTS and collaterals versus perfusion imaging for outcome prediction

BACKGROUND

Patients presenting with large baseline infarctions are often excluded from mechanical thrombectomy (MT) due to uncertainty surrounding its effect on outcome. We hypothesized that computed tomography perfusion (CTP)-based selection may be predictive of functional outcome in low Alberta Stroke Program Early CT Score (ASPECTS) patients.

METHODS

This was a double-center, retrospective analysis of patients presenting with ASPECTS≤5 who received multimodal admission CT imaging between May 2015 and June 2020. The predicted ischemic core (pCore) was defined as a reduction in cerebral blood flow (rCBF), while mismatch volume was defined using time to maximum (Tmax). The pCore perfusion mismatch ratio (CPMR) was also calculated. These parameters (pCore, mismatch volume, and CPMR), as well as a combined radiological score consisting of ASPECTS and collateral status (ASCO score), were tested in logistic regression and receiver operating characteristic (ROC) analyses. The primary outcome was favorable modified Rankin Scale (mRS) at discharge (≤3).

RESULTS

A total of 113 patients met the inclusion criteria. The median ischemic core volume was 74.1 mL (IQR 43.8-121.8). The ASCO score was associated with favorable outcome at discharge (aOR 3.7, 95% CI 1.8 to 10.7, P=0.002), while no association was observed for the CTP parameters. A model including the ASCO score also had significantly higher area under the curve (AUC) values compared with the CTP-based model (0.88 vs 0.64, P=0.018).

CONCLUSIONS

The ASCO score was superior to the CTP-based model for the prediction of good functional outcome and could represent a quick, practical, and easily implemented method for the selection of low ASPECTS patients most likely benefit from MT.

Impact of Initial Imaging Protocol on Likelihood of Endovascular Stroke Therapy

Background and Purpose:

Noncontrast head CT and CT perfusion (CTP) are both used to screen for endovascular stroke therapy (EST), but the impact of imaging strategy on likelihood of EST is undetermined. Here, we examine the influence of CTP utilization on likelihood of EST in patients with large vessel occlusion (LVO).

Methods:

We identified patients with acute ischemic stroke at 4 comprehensive stroke centers. All 4 hospitals had 24/7 CTP and EST capability and were covered by a single physician group (Neurology, NeuroIntervention, NeuroICU). All centers performed noncontrast head CT and CT angiography in the initial evaluation. One center also performed CTP routinely with high CTP utilization (CTP-H), and the others performed CTP optionally with lower utilization (CTP-L). Primary outcome was likelihood of EST. Multivariable logistic regression was used to determine whether facility type (CTP-H versus CTP-L) was associated with EST adjusting for age, prestroke mRS, National Institutes of Health Stroke Scale, Alberta Stroke Program Early CT Score, LVO location, time window, and intravenous tPA (tissue-type plasminogen activator).

Results:

Among 3107 patients with acute ischemic stroke, 715 had LVO, of which 403 (56%) presented to CTP-H and 312 (44%) presented to CTP-L. CTP utilization among LVO patients was greater at CTP-H centers (72% versus 18%, CTP-H versus CTP-L, P <0.01). In univariable analysis, EST rates for patients with LVO were similar between CTP-H versus CTP-L (46% versus 49%). In multivariable analysis, patients with LVO were less likely to undergo EST at CTP-H (odds ratio, 0.59 [0.41–0.85]). This finding was maintained in multiple patient subsets including late time window, anterior circulation LVO, and direct presentation patients. Ninety-day functional independence (odds ratio, 1.04 [0.70–1.54]) was not different, nor were rates of post-EST PH-2 hemorrhage (1% versus 1%).

Conclusions:

We identified an increased likelihood for undergoing EST in centers with lower CTP utilization, which was not associated with worse clinical outcomes or increased hemorrhage. These findings suggest under-treatment bias with routine CTP.

Efficient Multimodal MRI Evaluation for Endovascular Thrombectomy of Anterior Circulation Large Vessel Occlusion

BACKGROUND

MRI and CT modalities are both current standard-of-care options for initial imaging in patients with acute ischemic stroke due to large vessel occlusion (AIS-LVO). MR provides greater lesion conspicuity and spatial resolution, but few series have demonstrated multimodal MR may be performed efficiently.

METHODS

In a prospective comprehensive stroke center registry, we analyzed all anterior circulation LVO thrombectomy patients between 2012-2017 who: (1) arrived directly by EMS from the field, and (2) had initial NIHSS ≥6. Center imaging policy was multimodal MRI (including DWI/GRE/MRA w/wo PWI) as the initial evaluation in all patients without contraindications, and multimodal CT (including CT with CTA, w/wo CTP) in the remainder.

RESULTS

Among 106 EMS-arriving endovascular thrombectomy patients, initial imaging was MRI 62.3%, CT in 37.7%. MRI and CT patients were similar in age (72.5 vs 71.3), severity (NIHSS 16.4 v 18.2), and medical history, though MRI patients had longer onset-to-door times. Overall, door-to-needle (DTN) and door-to-puncture (DTP) times did not differ among MR and CT patients, and were faster for both modalities in 2015-2017 versus 2012-2014. In the 2015-2017 period, for MR-imaged patients, the median DTN 42m (IQR 34-55) surpassed standard (60m) and advanced (45m) national targets and the median DTP 86m (IQR 71-106) surpassed the standard national target (90m).

CONCLUSIONS

AIS-LVO patients can be evaluated by multimodal MR imaging with care speeds faster than national recommendations for door-to-needle and door-to-puncture times. With its more sensitive lesion identification and spatial resolution, MRI remains a highly viable primary imaging strategy in acute ischemic stroke patients, though further workflow efficiency improvements are desirable.

Prognostic value of cerebral infarction coefficient in patients with massive cerebral infarction

OBJECTIVE

We proposed the concept of the cerebral infarction coefficient, which is cerebral infarction volume/brain volume. This study aimed to evaluate the prognostic value of the cerebral infarction coefficient in patients with massive cerebral infarction (MCI).

METHODS

According to the modified Rankin score, 71 patients with acute MCI were divided into good prognosis and poor prognosis groups. Clinical and imaging data of the two groups were collected and univariate analysis was carried out. If there were significant differences in the data between the two groups, binary logistic regression analysis was performed.

RESULTS

The poor prognosis group had a significantly higher cerebral infarction volume, cerebral infarction coefficient, and D-dimer levels, older age, the highest body temperature, a higher rate of a history of atrial fibrillation, and a lower rate of a history of hypertension compared with the good prognosis group (all P < 0.05). Binary logistic regression analysis showed that the cerebral infarction coefficient was an independent risk factor for a poor prognosis of patients with MCI (P < 0.05, 95 % confidence interval, 2.091, 42.562), and the odds ratio was 8.506. The area under the receiver operating characteristic curve for the cerebral infarction coefficient was 0.753. When the cut-off value was 7.8 %, the sensitivity of predicting a poor prognosis of patients with MCI was 92.5 %.

CONCLUSION

The cerebral infarction coefficient may have predictive value in determining the prognosis of patients with MCI.

Comparison of Accuracy of Arrival-Time-Insensitive and Arrival-Time-Sensitive CTP Algorithms for Prediction of Infarct Tissue Volumes

The purpose of this study was to compare the performance of arrival-time-insensitive (ATI) and arrival-time-sensitive (ATS) computed tomography perfusion (CTP) algorithms in Philips IntelliSpace Portal (v9, ISP) and to investigate optimal thresholds for ATI regarding the prediction of final infarct volume (FIV). Retrospective, single-center study with 54 patients (mean 67.0 ± 13.1 years, 68.5% male) who received Stroke-CT/CTP-imaging between 2010 and 2018 with occlusion of the middle cerebral artery in the M1-/proximal M2-segment or terminal internal carotid artery. FIV was determined on short-term follow-up imaging in two patient groups: A) not attempted or failed mechanical thrombectomy (MT) and B) successful MT. ATS (default settings) and ATI (full-range of threshold settings regarding FIV prediction) maps were coregistered in 3D with FIV using voxel-wise overlap measurement. Based on an average imaging follow-up of 2.6 ± 2.1 days, the estimation regarding penumbra (group A, ATI: r = 0.63/0.69, ATS: r = 0.64) and infarct core (group B, ATI: r = 0.60/0.68, ATS: r = 0.63) was slightly higher in ATI but the effect was not significant (p > 0.05). Regarding ATI, Tmax (AUC 0.9) was the best estimator of the penumbra (group A), CBF relative to the contralateral hemisphere (AUC 0.80) showed the best estimation of the infarct core (group B). There was a broad range of thresholds of optimal ATI settings in both groups. Prediction of FIV with ATI was slightly better compared to ATS. However, this difference was not significant. Since ATI showed a broad range of optimal thresholds, exact thresholds regarding the ATI algorithm should be evaluated in further prospective, clinical studies.

Acute ischemic stroke: improving access to intravenous tissue plasminogen activator

INTRODUCTION

Since approval by the United States Food and Drug Administration in 1996, alteplase utilization rates for acute ischemic stroke have increased. Despite its efficacy for improving stroke outcomes, however, the majority of ischemic stroke patients still do not receive alteplase. To address this issue, different methods for improving access to alteplase have been tested with varying degrees of success.

AREAS COVERED

This article gives an overview of the recent approaches pursued to improve access to alteplase for acute ischemic stroke patients. Utilization of stroke systems of care, quality metrics, and quality-improvement initiatives to improve alteplase treatment rates are discussed. The implementation of Telestroke networks to improve access and timely evaluation by a stroke specialist are also reviewed. Lastly, this review discusses the use of neuroimaging techniques to identify alteplase candidates in stroke of unknown symptom onset or beyond the 4.5-h treatment window.

EXPERT COMMENTARY

Expanding access to alteplase therapy for acute ischemic stroke is a multi-faceted approach. Specific considerations based on region, population, and health-care resources should be considered for each strategy. Neuroimaging approaches to identify alteplase-eligible patients beyond the 4.5-h treatment window are a recent development in acute stroke care that holds promise for increasing alteplase treatment rates.

The accuracy of ischemic core perfusion thresholds varies according to time to recanalization in stroke patients treated with mechanical thrombectomy: A comprehensive whole-brain computed tomography perfusion study

Computed tomography perfusion (CTP) allows the estimation of pretreatment ischemic core after acute ischemic stroke. However, CTP-derived ischemic core may overestimate final infarct volume. We aimed to evaluate the accuracy of CTP-derived ischemic core for the prediction of final infarct volume according to time from stroke onset to recanalization in 104 patients achieving complete recanalization after mechanical thrombectomy who had a pretreatment CTP and a 24-h follow-up MRI-DWI. A range of CTP thresholds was explored in perfusion maps at constant increments for ischemic core calculation. Time to recanalization modified significantly the association between ischemic core and DWI lesion in a non-linear fashion (p-interaction = 0.018). Patients with recanalization before 4.5 h had significantly lower intraclass correlation coefficient (ICC) values between CTP-predicted ischemic core and DWI lesion (n = 54; best threshold relative cerebral blood flow (rCBF) < 25%, ICC = 0.673, 95% CI = 0.495-0.797) than those with later recanalization (n = 50; best threshold rCBF < 30%, ICC = 0.887, 95% CI = 0.811-0.935, p = 0.013), as well as poorer spatial lesion agreement. The significance of the associations between CTP-derived ischemic core and clinical outcome at 90 days was lost in patients recanalized before 4.5 h. CTP-derived ischemic core must be interpreted with caution given its dependency on time to recanalization, primarily in patients with higher chances of early recanalization.

Signal variance-based collateral index in DSC perfusion: A novel method to assess leptomeningeal collateralization in acute ischaemic stroke

As a determinant of the progression rate of the ischaemic process in acute large-vessel stroke, the degree of collateralization is a strong predictor of the clinical outcome after reperfusion therapy and may influence clinical decision-making. Therefore, the assessment of leptomeningeal collateralization is of major importance. The purpose of this study was to develop and evaluate a quantitative and observer-independent method for assessing leptomeningeal collateralization in acute large-vessel stroke based on signal variance characteristics in T2*-weighted dynamic susceptibility contrast (DSC) perfusion-weighted MR imaging (PWI). Voxels representing leptomeningeal collateral vessels were extracted according to the magnitude of signal variance in the PWI raw data time series in 55 patients with proximal large-artery occlusion and an intra-individual collateral vessel index (CVI_PWI) was calculated. CVI_PWI correlated significantly with the initial ischaemic core volume (rho = -0.459, p = 0.0001) and the PWI/DWI mismatch ratio (rho = 0.494, p = 0.0001) as an indicator of the amount of salvageable tissue. Furthermore, CVI_PWI was significantly negatively correlated with NIHSS and mRS at discharge (rho = -0.341, p = 0.015 and rho = -0.305, p = 0.023). In multivariate logistic regression, CVI_PWI was an independent predictor of favourable functional outcome (mRS 0-2) (OR = 16.39, 95% CI 1.42-188.7, p = 0.025). CVI_PWI provides useful rater-independent information on the leptomeningeal collateral supply in acute stroke.

Deep Learning Detection of Penumbral Tissue on Arterial Spin Labeling in Stroke

Background and Purpose- Selection of patients with acute ischemic stroke for endovascular treatment generally relies on dynamic susceptibility contrast magnetic resonance imaging or computed tomography perfusion. Dynamic susceptibility contrast magnetic resonance imaging requires injection of contrast, whereas computed tomography perfusion requires high doses of ionizing radiation. The purpose of this work was to develop and evaluate a deep learning (DL)-based algorithm for assisting the selection of suitable patients with acute ischemic stroke for endovascular treatment based on 3-dimensional pseudo-continuous arterial spin labeling (pCASL). Methods- A total of 167 image sets of 3-dimensional pCASL data from 137 patients with acute ischemic stroke scanned on 1.5T and 3.0T Siemens MR systems were included for neural network training. The concurrently acquired dynamic susceptibility contrast magnetic resonance imaging was used to produce labels of hypoperfused brain regions, analyzed using commercial software. The DL and 6 machine learning (ML) algorithms were trained with 10-fold cross-validation. The eligibility for endovascular treatment was determined retrospectively based on the criteria of perfusion/diffusion mismatch in the DEFUSE 3 trial (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke). The trained DL algorithm was further applied on twelve 3-dimensional pCASL data sets acquired on 1.5T and 3T General Electric MR systems, without fine-tuning of parameters. Results- The DL algorithm can predict the dynamic susceptibility contrast-defined hypoperfusion region in pCASL with a voxel-wise area under the curve of 0.958, while the 6 ML algorithms ranged from 0.897 to 0.933. For retrospective determination for subject-level endovascular treatment eligibility, the DL algorithm achieved an accuracy of 92%, with a sensitivity of 0.89 and specificity of 0.95. When applied to the GE pCASL data, the DL algorithm achieved a voxel-wise area under the curve of 0.94 and a subject-level accuracy of 92% for endovascular treatment eligibility. Conclusions- pCASL perfusion magnetic resonance imaging in conjunction with the DL algorithm provides a promising approach for assisting decision-making for endovascular treatment in patients with acute ischemic stroke.

e-ASPECTS derived acute ischemic volumes on non-contrast-enhanced computed tomography images

Background and purpose

Validation of automatically derived acute ischemic volumes (AAIV) from e-ASPECTS on non-contrast computed tomography (NCCT).

Materials and methods

Data from three studies were reanalyzed with e-ASPECTS Version 7. AAIV was calculated in milliliters (ml) in all scored ASPECTS regions of the hemisphere detected by e-ASPECTS. The National Institute of Health Stroke Scale (NIHSS) determined stroke severity at baseline and clinical outcome was measured with the modified Rankin Scale (mRS) between 45 and 120 days. Spearman ranked correlation coefficients (R) of AAIV and e-ASPECTS scores with NIHSS and mRS as well as Pearson correlation of AAIV with diffusion-weighted imaging and CT perfusion-estimated ischemic “core” volumes were calculated. Multivariate regression analysis (odds ratio, OR with 95% confidence intervals, CI) and Bland–Altman plots were performed.

Results

We included 388 patients. Mean AAIV was 11.6 ± 18.9 ml and e-ASPECTS was 9 (8–10: median and interquartile range). AAIV, respectively e-ASPECTS correlated with NIHSS at baseline (R = 0.35, p < 0.001; R = −0.36, p < 0.001) and follow-up mRS (R = 0.29, p < 0.001; R = −0.3, p < 0.001). In subsets of patients, AAIV correlated strongly with diffusion-weighted imaging (n = 37, R = 0.68, p < 0.001) and computed tomography perfusion-derived ischemic “core” (n = 41, R = 0.76, p < 0.001) lesion volume and Bland–Altman plots showed a bias close to zero (−2.65 ml for diffusion-weighted imaging and 0.45 ml forcomputed tomography perfusion “core”). Within the whole cohort, the AAIV (OR 0.98 per ml, 95% CI 0.96–0.99) and e-ASPECTS scores (OR 1.3, 95%CI 1.07–1.57) were independent predictors of good outcome

Conclusion

AAIV on NCCT correlated moderately with clinical severity but strongly with diffusion-weighted imaging lesion and computed tomography perfusion ischemic “core” volumes and predicted clinical outcome.

Machine Learning-Enabled Automated Determination of Acute Ischemic Core From Computed Tomography Angiography

Background and Purpose- The availability of and expertise to interpret advanced neuroimaging recommended in the guideline-based endovascular stroke therapy (EST) evaluation are limited. Here, we develop and validate an automated machine learning-based method that evaluates for large vessel occlusion (LVO) and ischemic core volume in patients using a widely available modality, computed tomography angiogram (CTA). Methods- From our prospectively maintained stroke registry and electronic medical record, we identified patients with acute ischemic stroke and stroke mimics with contemporaneous CTA and computed tomography perfusion (CTP) with RAPID (IschemaView) post-processing as a part of the emergent stroke workup. A novel convolutional neural network named DeepSymNet was created and trained to identify LVO as well as infarct core from CTA source images, against CTP-RAPID definitions. Model performance was measured using 10-fold cross validation and receiver-operative curve area under the curve (AUC) statistics. Results- Among the 297 included patients, 224 (75%) had acute ischemic stroke of which 179 (60%) had LVO. Mean CTP-RAPID ischemic core volume was 23±42 mL. LVO locations included internal carotid artery (13%), M1 (44%), and M2 (21%). The DeepSymNet algorithm autonomously learned to identify the intracerebral vasculature on CTA and detected LVO with AUC 0.88. The method was also able to determine infarct core as defined by CTP-RAPID from the CTA source images with AUC 0.88 and 0.90 (ischemic core ≤30 mL and ≤50 mL). These findings were maintained in patients presenting in early (0-6 hours) and late (6-24 hours) time windows (AUCs 0.90 and 0.91, ischemic core ≤50 mL). DeepSymNet probabilities from CTA images corresponded with CTP-RAPID ischemic core volumes as a continuous variable with r=0.7 (Pearson correlation, P<0.001). Conclusions- These results demonstrate that the information needed to perform the neuroimaging evaluation for endovascular therapy with comparable accuracy to advanced imaging modalities may be present in CTA, and the ability of machine learning to automate the analysis.

Noncontrast Computed Tomography Alberta Stroke Program Early CT Score May Modify Intra-Arterial Treatment Effect in DAWN

Background and Purpose—

It is unknown whether noncontrast computed tomography (NCCT) can identify patients who will benefit from intra-arterial treatment (IAT) in the extended time window. We sought to characterize baseline Alberta Stroke Program Early CT Score (ASPECTS) in DAWN (DWI or CTP Assessment With Clinical Mismatch in the Triage of Wake-Up and Late Presenting Strokes Undergoing Neurointervention With Trevo) and to assess whether ASPECTS modified IAT effect.

Methods—

Core lab adjudicated ASPECTS scores were analyzed. The trial cohort was divided into 2 groups by qualifying imaging (computed tomography versus magnetic resonance imaging). ASPECTS-by-treatment interaction was tested for the trial coprimary end points (90-day utility-weighted modified Rankin Scale (mRS) score and mRS, 0–2), mRS 0 to 3, and ordinal mRS. ASPECTS was evaluated separately as an ordinal and a dichotomized (0–6 versus 7–10) variable.

Results—

Of 205 DAWN subjects, 123 (60%) had NCCT ASPECTS, and 82 (40%) had diffusion weighted imaging ASPECTS. There was a significant ordinal NCCT ASPECTS-by-treatment interaction for 90-day utility-weighted mRS (interaction P =0.04) and mRS 0 to 2 (interaction P =0.02). For both end points, IAT effect was more pronounced at higher NCCT ASPECTS. The dichotomized NCCT ASPECTS-by-treatment interaction was significant only for mRS 0 to 2 (interaction P =0.04), where greater treatment benefit was seen in the ASPECTS 7 to 10 group (odds ratio, 7.50 [2.71–20.77] versus odds ratio, 0.48 [0.04–5.40]). A bidirectional treatment effect was observed in the NCCT ASPECTS 0 to 6 group, with treatment associated with not only more mRS 0 to 3 outcomes (50% versus 25%) but also more mRS 5 to 6 outcomes (40% versus 25%). There was no significant modification of IAT effect by diffusion weighted imaging ASPECTS.

Conclusions—

Baseline NCCT ASPECTS appears to modify IAT effect in DAWN. Higher NCCT ASPECTS was associated with greater benefit from IAT. No treatment interaction was observed for diffusion weighted imaging ASPECTS.

Computer-aided imaging analysis in acute ischemic stroke - background and clinical applications

Tools for medical image analysis have been developed to reduce the time needed to detect abnormalities and to provide more accurate results. Particularly, tools based on artificial intelligence and machine learning techniques have led to significant improvements in medical imaging interpretation in the last decade. Automatic evaluation of acute ischemic stroke in medical imaging is one of the fields that witnessed a major development. Commercially available products so far aim to identify (and quantify) the ischemic core, the ischemic penumbra, the site of arterial occlusion and the collateral flow but they are not (yet) intended as standalone diagnostic tools. Their use can be complementary; they are intended to support physicians' interpretation of medical images and hence standardise selection of patients for acute treatment. This review provides an introduction into the field of computer-aided diagnosis and focuses on the automatic analysis of non-contrast-enhanced computed tomography, computed tomography angiography and perfusion imaging. Future studies are necessary that allow the evaluation and comparison of different imaging strategies and post-processing algorithms during the diagnosis process in patients with suspected acute ischemic stroke; which may further facilitate the standardisation of treatment and stroke management.

Perfusion Computed Tomography in Acute Ischemic Stroke

Occlusion of a cervical or cerebral artery may cause acute ischemic stroke (AIS). Recent advances in AIS treatment by endovascular thrombectomy have led to more widespread use of advanced computed tomography (CT) imaging, including perfusion CT (PCT). This article reviews PCT for the evaluation of AIS patients.

Systematic review with network meta-analysis: Diagnostic values of ultrasonography, computed tomography, and magnetic resonance imaging in patients with ischemic stroke

BACKGROUND AND OBJECTIVE

Ischemic stroke is a foremost cause for disability and death worldwide. This study is conducted in order to compare the diagnostic values between transcranial Doppler ultrasound (ultrasonography), computed tomography (CT), and magnetic resonance imaging (MRI) in patients suffering from ischemic stroke by performing a network meta-analysis.

METHODS

We made use of Cochrane Library, PubMed, and Embase in order to obtain literature and papers. The combination analysis of both direct and indirect evidence in terms of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy was conducted so as to assess the odds ratios (ORs) and surface under the cumulative ranking curve (SUCRA) values of the seven different imaging methods. These imaging techniques include ultrasonography, computed tomography (traditional CT, computed tomography angiography [CTA], computed tomography perfusion [CTP]), and MRI (traditional MRI, diffusion-weighted imaging [DWI], magnetic resonance angiography), in order to properly diagnose ischemic stroke patients.

RESULTS

Thirteen eligible diagnostic trials were enrolled into this network meta-analysis. The results of the traditional meta-analysis showed that among CT methods, CTP showed higher sensitivity, NPV, and accuracy; among MRI methods, DWI had relatively higher sensitivity, NPV, and accuracy. The results of network meta-analysis showed that DWI had relatively higher sensitivity, NPV, and accuracy when compared with traditional CT, CTA, magnetic resonance angiography and traditional MRI. CTP showed higher SUCRA among CT methods while DWI showed higher SUCRA among MRI methods. A cluster analysis revealed that DWI had the highest diagnostic value in terms of sensitivity, PPV, NPV, and accuracy amongst the aforementioned seven imaging techniques.

CONCLUSION

This network meta-analysis provides supporting evidence to the idea that DWI has a higher diagnostic value regarding ischemic stroke among MRI methods, and CTP has a poor diagnostic value among CT methods, which provide therapeutic considerations for Ischemic stroke intervention.

Magnetic Resonance Imaging Selection for Endovascular Stroke Therapy

Background and Purpose—

The GOLIATH trial (General or Local Anesthesia in Intra-Arterial Therapy) compared infarct growth and outcome in patients undergoing endovascular therapy under either general anesthesia or conscious sedation. Magnetic resonance imaging was performed before and after the procedure to study infarct growth. In this post hoc analysis of GOLIATH, we aimed to characterize the workflow of patients undergoing magnetic resonance imaging selection before endovascular therapy.

Methods—

We randomized 128 patients with anterior circulation large vessel occlusion stroke within 6 hours of onset to either general anesthesia or conscious sedation (1:1 allocation). We studied workflow time intervals to examine whether magnetic resonance imaging conferred a time delay in treatment when compared with computed tomography-based studies that emphasized rapid workflow.

Results—

Of 128 patients enrolled between March 2015 and February 2017, 65 were randomized to general anesthesia. Baseline demographic and clinical variables were balanced between the treatment arms. The median interval from scan to groin puncture was 56.5 minutes (interquartile range, 44.5–73.5) for all patients. The median interval from admission to groin puncture was 68 minutes (interquartile range, 54.5–87 minutes). Comparable intervals in recent randomized data were 51 minutes (interquartile range, 39–68) for scan to groin puncture in the ESCAPE trial (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion With Emphasis on Minimizing CT to Recanalization Times) and 90 minutes (interquartile range, 69–120 minutes) for door to groin puncture in the SWIFT-PRIME study (Solitaire With the Intention for Thrombectomy as Primary Endovascular Treatment).

Conclusions—

Workflow in GOLIATH demonstrates that magnetic resonance imaging selection for endovascular therapy can be accomplished rapidly and within a similar time frame as computed tomography-based selection.

Clinical Trial Registration—

URL: https://www.clinicaltrials.gov . Unique identifier: NCT02317237.

Neuroimaging Paradigms to Identify Patients for Reperfusion Therapy in Stroke of Unknown Onset

Despite the proven efficacy of intravenous alteplase or endovascular thrombectomy for the treatment of patients with acute ischemic stroke, only a minority receive these treatments. This low treatment rate is due in large part to delay in hospital arrival or uncertainty as to the exact time of onset of ischemic stroke, which renders patients outside the current guideline-recommended window of eligibility for receiving these therapeutics. However, recent pivotal clinical trials of late-window thrombectomy now force us to rethink the value of a simplistic chronological formulation that “time is brain.” We must recognize a more nuanced concept that the rate of tissue death as a function of time is not invariant, that still salvageable tissue at risk of infarction may be present up to 24 h after last-known well, and that those patients may strongly benefit from reperfusion. Multiple studies have sought to address this clinical dilemma using neuroimaging methods to identify a radiographic time-stamp of stroke onset or evidence of salvageable ischemic tissue and thereby increase the number of patients eligible for reperfusion therapies. In this review, we provide a critical analysis of the current state of neuroimaging techniques to select patients with unwitnessed stroke for revascularization therapies and speculate on the future direction of this clinically relevant area of stroke research.