The aim of this study was to develop a convolutional neural network (CNN) that automatically detects and segments intra-arterial thrombi on baseline non-contrast computed tomography (NCCT) scans. We ...retrospectively collected computed tomography (CT)-scans of patients with an anterior circulation large vessel occlusion (LVO) from the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands trial, both for training (n = 86) and validation (n = 43). For testing we included patients with (n = 58) and without (n = 45) an LVO from our comprehensive stroke center. Ground truth was established by consensus between two experts using both CT angiography and NCCT. We evaluated the CNN for correct identification of a thrombus, its location and thrombus segmentation and compared these with the results of a neurologist in training and expert neuroradiologist. Sensitivity of the CNN thrombus detection was 0.86, vs. 0.95 and 0.79 for the neuroradiologists. Specificity was 0.65 for the network vs. 0.58 and 0.82 for the neuroradiologists. The CNN correctly identified the location of the thrombus in 79% of the cases, compared to 81% and 77% for the neuroradiologists. The sensitivity and specificity for thrombus identification and the rate for correct thrombus location assessment by the CNN were similar to those of expert neuroradiologists.
BACKGROUND AND PURPOSE—Intra-arterial treatment by means of retrievable stents has been proven safe and effective. In MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for ...Acute Ischemic Stroke in the Netherlands), the choice of the type of thrombectomy device was left to the discretion of the interventionist. The aim of this study was to explore the differences in functional outcome, neurological recovery, reperfusion, extent of infarction, and adverse events according to stent type and make.
METHODS—The primary outcome was functional outcome at 90 days, assessed with the modified Rankin Scale (mRS). Neuroimaging outcomes included occlusion on computed tomographic angiography at 24 hours, infarct volume at 5 to 7 days, and modified thrombolysis in cerebral infarction scores. Safety outcomes included death within 90 days and any symptomatic intracerebral hemorrhage. We analyzed possible interactions between stent type and treatment with multiple regression models. Treatment effects were adjusted for patient age, stroke severity, and collateral score.
RESULTS—Of the 500 patients included in the trial, 233 were allocated to intervention. Of these, 124 (53%) were first treated with Trevo (adjusted common odds ratio for shift on the mRS acOR, 1.98; 95% confidence interval, 1.30–2.92), 31 (13%) with Solitaire (acOR, 1.90; 95% confidence interval, 0.97–3.73), 40 (17%) with other retrievable stents or mechanical devices (acOR, 0.96; 95% confidence interval, 0.51–3.93, and 38 (16%) could not be treated. There was no interaction between device and treatment effect on functional outcome and all other secondary and safety outcomes.
CONCLUSIONS—We found no evidence for a differential effect of thrombectomy for acute ischemic stroke by type of stent.
CLINICAL TRIAL REGISTRATION—URLhttp://www.isrctn.com. Unique identifierISRCTN10888758.
Introduction
Endovascular treatment (EVT) has been proven to be both effective and cost-effective for patients with acute ischaemic stroke. We investigated the budget impact of large-scale ...implementation of EVT for acute ischaemic stroke patients in the Netherlands for 2015–2021.
Methods
An analysis was performed from a healthcare perspective as a preplanned substudy of the Multicenter Randomized Clinical trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN). Estimated yearly costs during follow-up after stroke for patients who had or had not been treated with EVT as add-on to usual care were linked to numbers of new patients retrieved from 2 Dutch registries of EVT that started after the last inclusion in MR CLEAN (2014). Aggregated costs and costs per care sector were calculated based on prevalence using a population dynamic tool.
Results
From 2015, the yearly number of new acute ischaemic stroke patients receiving EVT increased almost threefold, from 812 in 2015 to 2,370 in 2021. The introduction of EVT plus usual care resulted in estimated net annual savings that increased from € 2.9 million in 2015 to € 58 million in 2021.
Conclusion
Offering EVT as add-on to usual care for acute ischaemic stroke patients was increasingly cost saving from a national healthcare perspective but affected distinct healthcare sectors differently.
Infarct volume is a valuable outcome measure in treatment trials of acute ischemic stroke and is strongly associated with functional outcome. Its manual volumetric assessment is, however, too ...demanding to be implemented in clinical practice.
To assess the value of convolutional neural networks (CNNs) in the automatic segmentation of infarct volume in follow-up CT images in a large population of patients with acute ischemic stroke.
We included CT images of 1026 patients from a large pooling of patients with acute ischemic stroke. A reference standard for the infarct segmentation was generated by manual delineation. We introduce three CNN models for the segmentation of subtle, intermediate, and severe hypodense lesions. The fully automated infarct segmentation was defined as the combination of the results of these three CNNs. The results of the three-CNNs approach were compared with the results from a single CNN approach and with the reference standard segmentations.
The median infarct volume was 48 mL (IQR 15-125 mL). Comparison between the volumes of the three-CNNs approach and manually delineated infarct volumes showed excellent agreement, with an intraclass correlation coefficient (ICC) of 0.88. Even better agreement was found for severe and intermediate hypodense infarcts, with ICCs of 0.98 and 0.93, respectively. Although the number of patients used for training in the single CNN approach was much larger, the accuracy of the three-CNNs approach strongly outperformed the single CNN approach, which had an ICC of 0.34.
Convolutional neural networks are valuable and accurate in the quantitative assessment of infarct volumes, for both subtle and severe hypodense infarcts in follow-up CT images. Our proposed three-CNNs approach strongly outperforms a more straightforward single CNN approach.
Infarct volume (FIV) on follow-up diffusion-weighted imaging (FU-DWI) is only moderately associated with functional outcome in acute ischemic stroke patients. However, FU-DWI may contain other ...imaging biomarkers that could aid in improving outcome prediction models for acute ischemic stroke. We included FU-DWI data from the HERMES, ISLES, and MR CLEAN-NO IV databases. Lesions were segmented using a deep learning model trained on the HERMES and ISLES datasets. We assessed the performance of three classifiers in predicting functional independence for the MR CLEAN-NO IV trial cohort based on: (1) FIV alone, (2) the most important features obtained from a trained convolutional autoencoder (CAE), and (3) radiomics. Furthermore, we investigated feature importance in the radiomic-feature-based model. For outcome prediction, we included 206 patients: 144 scans were included in the training set, 21 in the validation set, and 41 in the test set. The classifiers that included the CAE and the radiomic features showed AUC values of 0.88 and 0.81, respectively, while the model based on FIV had an AUC of 0.79. This difference was not found to be statistically significant. Feature importance results showed that lesion intensity heterogeneity received more weight than lesion volume in outcome prediction. This study suggests that predictions of functional outcome should not be based on FIV alone and that FU-DWI images capture additional prognostic information.
The presence of extracranial carotid disease (ECD) is associated with less favorable clinical outcomes in patients with acute ischemic stroke caused by intracranial proximal occlusion. Acute ...intra-arterial treatment (IAT) in the setting of extracranial and intracranial lesions is considered challenging, and whether it yields improved outcomes remains uncertain.
To examine whether the presence of ECD modified the effect of IAT for intracranial proximal anterior circulation occlusion.
Prespecified subgroup analysis of a randomized clinical trial of endovascular treatment for acute ischemic stroke in the Netherlands. (Trial registrations: NTR1804 Netherlands Trial Register and ISRCTN10888758).
16 hospitals in the Netherlands.
Acute ischemic stroke caused by proximal intracranial arterial occlusion of the anterior circulation. Extracranial carotid disease was defined as cervical internal carotid artery stenosis (>50%) or occlusion.
IAT treatment versus no IAT.
The primary outcome was functional outcome, as measured by the modified Rankin Scale at 90 days and reported as adjusted common odds ratio (acOR) for a shift in direction of a better outcome. Multivariable ordinal logistic regression analysis with an interaction term was used to estimate treatment effect modification by ECD.
The overall acOR was 1.67 (95% CI, 1.21 to 2.30) in favor of the intervention. The acOR was 3.1 (CI, 1.7 to 5.8) in the prespecified subgroup of patients with ECD versus 1.3 (CI, 0.9 to 1.9) in patients presenting without ECD. Both acORs are in favor of the intervention (P for interaction = 0.07).
The study was not powered for subgroup analysis.
Intra-arterial treatment may be at least as effective in patients with ECD as in those without ECD, and it should not be withheld in these complex patients with acute ischemic stroke.
Dutch Heart Foundation, AngioCare BV, Medtronic/Covidien/EV3, MEDAC Gmbh/LAMEPRO, Penumbra, Stryker, and Top Medical/Concentric.
Purpose: In acute middle cerebral artery (MCA) occlusion, collateral vessels provide retrograde supply to the occluded territory. We hypothesized that such collateral flow reduces perfusion of the ...non-occluded donor region (steal effect).Materials and Methods: Patients with an MCA occlusion with opacification of both ipsi- and contralateral anterior cerebral arteries (ACA) on angiography prior to endovascular treatment were selected. Arteriovenous transit time (AVTT) for both ACA territories was compared for different grades of collateral supply to the MCA territory. In addition, the influence of diabetes and hypertension was analyzed. After successful revascularization, AVTT was re-assessed to determine reversibility.Results: Forty-one patients were analyzed. An AVTT of 8.6 seconds (standard deviation SD 2.4 seconds) was seen in the ACA territory of the affected hemisphere in comparison to 6.6 seconds (SD 2.1 seconds) for the contralateral side (P<0.001). A more prolonged (but not significant) AVTT was seen in cases with a higher collateral grade. No difference in AVTT was seen in patients with diabetes or hypertension. After successful MCA revascularization, AVTT delay was 7.4 seconds (SD 2.1 seconds).Conclusion: A cerebral steal effect occurs in patients with an acute MCA occlusion, probably related to augmented flow to the penumbra area.
BACKGROUND AND PURPOSE—We assessed the effect of endovascular treatment in acute ischemic stroke patients with severe neurological deficit (National Institutes of Health Stroke Scale score, ≥20) ...after a prespecified analysis plan.
METHODS—The pooled analysis of the Interventional Management of Stroke III (IMS III) and Multicenter Randomized Clinical Trial of Endovascular Therapy for Acute Ischemic Stroke in the Netherlands (MR CLEAN) trials included participants with an National Institutes of Health Stroke Scale score of ≥20 before intravenous tissue-type plasminogen activator (tPA) treatment (IMS III) or randomization (MR CLEAN) who were treated with intravenous tPA ≤3 hours of stroke onset. Our hypothesis was that participants with severe stroke randomized to endovascular therapy after intravenous tPA would have improved 90-day outcome (distribution of modified Rankin Scale scores), when compared with those who received intravenous tPA alone.
RESULTS—Among 342 participants in the pooled analysis (194 from IMS III and 148 from MR CLEAN), an ordinal logistic regression model showed that the endovascular group had superior 90-day outcome compared with the intravenous tPA group (adjusted odds ratio, 1.78; 95% confidence interval, 1.20–2.66). In the logistic regression model of the dichotomous outcome (modified Rankin Scale score, 0–2, or functional independence), the endovascular group had superior outcomes (adjusted odds ratio, 1.97; 95% confidence interval, 1.09–3.56). Functional independence (modified Rankin Scale score, ≤2) at 90 days was 25% in the endovascular group when compared with 14% in the intravenous tPA group.
CONCLUSIONS—Endovascular therapy after intravenous tPA within 3 hours of symptom onset improves functional outcome at 90 days after severe ischemic stroke.
CLINICAL TRIAL REGISTRATION—URLhttp://www.clinicaltrials.gov. Unique identifierNCT00359424 (IMS III) and ISRCTN10888758 (MR CLEAN).
Background Current American guidelines are uncertain regarding endovascular treatment (EVT) for patients with acute ischemic stroke with an Alberta Stroke Program Early Computed Tomography Score ...(ASPECTS) <6. Dutch guidelines do not specify ASPECTS‐based exclusion criteria for EVT. In this retrospective observational cohort study, we investigated outcomes of EVT in patients with low ASPECTS in the MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in The Netherlands) registry. Methods ASPECTS was trichotomized into 0 to 2, 3 to 5, and 6 to 10, according to the grouping used in the ongoing trials. The effect of ASPECTS (granular and trichotomized) on 90‐day functional outcome (modified Rankin Scale score) and symptomatic intracranial hemorrhage was assessed with multivariable logistic regression. We included multiplicative interaction terms to evaluate treatment interaction between ASPECTS and reperfusion (extended thrombolysis in cerebral infarction score 2B‐3) as a proxy for EVT. Results Among 3075 included patients, higher ASPECTS was associated with improved functional outcome (granular: adjusted common odds ratio acOR, 1.06; 95% CI, 1.02–1.10; trichotomized: acOR, 1.52; 95% CI, 1.23–1.87). Interaction with reperfusion was not significant ( P =0.99 for granular, P =0.76 for trichotomized ASPECTS). All ASPECTS subgroups showed benefit of reperfusion (0–2 n = 39: acOR, 7.40; 95% CI, 1.41–18.68; 3–5 n = 214: acOR, 1.95; 95% CI, 1.13–3.34; 6–10 n = 2822: acOR, 2.41; 95% CI, 2.08–2.80). ASPECTS was not associated with symptomatic intracranial hemorrhage (granular: acOR, 1.00; 95% CI, 0.92–1.10, trichotomized: acOR, 0.92; 95% CI, 0.60–1.41). Conclusion Benefit of reperfusion was not modified by baseline ASPECTS. Patients in all ASPECTS subgroups showed benefit of reperfusion. These findings do not support withholding EVT on the basis of low ASPECTS only.
PDF
