Value of Automatically Derived Full Thrombus Characteristics: An Explorative Study of Their Associations with Outcomes in Ischemic Stroke Patients

(1) Background: For acute ischemic strokes caused by large vessel occlusion, manually assessed thrombus volume and perviousness have been associated with treatment outcomes. However, the manual assessment of these characteristics is time-consuming and subject to inter-observer bias. Alternatively, a recently introduced fully automated deep learning-based algorithm can be used to consistently estimate full thrombus characteristics. Here, we exploratively assess the value of these novel biomarkers in terms of their association with stroke outcomes. (2) Methods: We studied two applications of automated full thrombus characterization as follows: one in a randomized trial, MR CLEAN-NO IV (n = 314), and another in a Dutch nationwide registry, MR CLEAN Registry (n = 1839). We used an automatic pipeline to determine the thrombus volume, perviousness, density, and heterogeneity. We assessed their relationship with the functional outcome defined as the modified Rankin Scale (mRS) at 90 days and two technical success measures as follows: successful final reperfusion, which is defined as an eTICI score of 2b-3, and successful first-pass reperfusion (FPS). (3) Results: Higher perviousness was significantly related to a better mRS in both MR CLEAN-NO IV and the MR CLEAN Registry. A lower thrombus volume and lower heterogeneity were only significantly related to better mRS scores in the MR CLEAN Registry. Only lower thrombus heterogeneity was significantly related to technical success; it was significantly related to a higher chance of FPS in the MR CLEAN-NO IV trial (OR = 0.55, 95% CI: 0.31-0.98) and successful reperfusion in the MR CLEAN Registry (OR = 0.88, 95% CI: 0.78-0.99). (4) Conclusions: Thrombus characteristics derived from automatic entire thrombus segmentations are significantly related to stroke outcomes.

Thrombus radiomics in patients with anterior circulation acute ischemic stroke undergoing endovascular treatment

BACKGROUND

Thrombus radiomics (TR) describe complex shape and textural thrombus imaging features. We aimed to study the relationship of TR extracted from non-contrast CT with procedural and functional outcome in endovascular-treated patients with acute ischemic stroke.

METHODS

Thrombi were segmented on thin-slice non-contrast CT (≤1 mm) from 699 patients included in the MR CLEAN Registry. In a pilot study, we selected 51 TR with consistent values across two raters' segmentations (ICC >0.75). Random forest models using TR in addition or as a substitute to baseline clinical variables (CV) and manual thrombus measurements (MTM) were trained with 499 patients and evaluated on 200 patients for predicting successful reperfusion (extended Thrombolysis in Cerebral Ischemia (eTICI) ≥2B), first attempt reperfusion, reperfusion within three attempts, and functional independence (modified Rankin Scale (mRS) ≤2). Three texture and shape features were selected based on feature importance and related to eTICI ≥2B, number of attempts to eTICI ≥2B, and 90-day mRS with ordinal logistic regression.

RESULTS

Random forest models using TR, CV or MTM had comparable predictive performance. Thrombus texture (inverse difference moment normalized) was independently associated with reperfusion (adjusted common OR (acOR) 0.85, 95% CI 0.72 to 0.99). Thrombus volume and texture were also independently associated with the number of attempts to successful reperfusion (acOR 1.36, 95% CI 1.03 to 1.88 and acOR 1.24, 95% CI 1.04 to 1.49).

CONCLUSIONS

TR describing thrombus volume and texture were associated with more attempts to successful reperfusion. Compared with models using CV and MTM, TR had no added value for predicting procedural and functional outcome.

Thrombus imaging characteristics within acute ischemic stroke: similarities and interdependence

BACKGROUND

The effects of thrombus imaging characteristics on procedural and clinical outcomes after ischemic stroke are increasingly being studied. These thrombus characteristics - for eg, size, location, and density - are commonly analyzed as separate entities. However, it is known that some of these thrombus characteristics are strongly related. Multicollinearity can lead to unreliable prediction models. We aimed to determine the distribution, correlation and clustering of thrombus imaging characteristics based on a large dataset of anterior-circulation acute ischemic stroke patients.

METHODS

We measured thrombus imaging characteristics in the MR CLEAN Registry dataset, which included occlusion location, distance from the intracranial carotid artery to the thrombus (DT), thrombus length, density, perviousness, and clot burden score (CBS). We assessed intercorrelations with Spearman's coefficient (ρ) and grouped thrombi based on 1) occlusion location and 2) thrombus length, density and perviousness using unsupervised clustering.

RESULTS

We included 934 patients, of which 22% had an internal carotid artery (ICA) occlusion, 61% M1, 16% M2, and 1% another occlusion location. All thrombus characteristics were significantly correlated. Higher CBS was strongly correlated with longer DT (ρ=0.67, p<0.01), and moderately correlated with shorter thrombus length (ρ=-0.41, p<0.01). In more proximal occlusion locations, thrombi were significantly longer, denser, and less pervious. Unsupervised clustering analysis resulted in four thrombus groups; however, the cohesion within and distinction between the groups were weak.

CONCLUSIONS

Thrombus imaging characteristics are significantly intercorrelated - strong correlations should be considered in future predictive modeling studies. Clustering analysis showed there are no distinct thrombus archetypes - novel treatments should consider this thrombus variability.

Accuracy of Four Different CT Perfusion Thresholds for Ischemic Core Volume and Location Estimation Using IntelliSpace Portal

Computed tomography perfusion (CTP) is frequently used in the triage of ischemic stroke patients for endovascular thrombectomy (EVT). We aimed to quantify the volumetric and spatial agreement of the CTP ischemic core estimated with different thresholds and follow-up MRI infarct volume on diffusion-weighted imaging (DWI). Patients treated with EVT between November 2017 and September 2020 with available baseline CTP and follow-up DWI were included. Data were processed with Philips IntelliSpace Portal using four different thresholds. Follow-up infarct volume was segmented on DWI. In 55 patients, the median DWI volume was 10 mL, and median estimated CTP ischemic core volumes ranged from 10-42 mL. In patients with complete reperfusion, the intraclass correlation coefficient (ICC) showed moderate-good volumetric agreement (range 0.55-0.76). A poor agreement was found for all methods in patients with successful reperfusion (ICC range 0.36-0.45). Spatial agreement (median Dice) was low for all four methods (range 0.17-0.19). Severe core overestimation was most frequently (27%) seen in Method 3 and patients with carotid-T occlusion. Our study shows moderate-good volumetric agreement between ischemic core estimates for four different thresholds and subsequent infarct volume on DWI in EVT-treated patients with complete reperfusion. The spatial agreement was similar to other commercially available software packages.

Bifurcation occlusions and endovascular treatment outcome in acute ischemic stroke

BACKGROUND

A thrombus in the M1 segment of the middle cerebral artery (MCA) can occlude this main stem only or extend into the M1-M2 bifurcation. The occlusion pattern may affect endovascular treatment (EVT) success, as a bifurcated thrombus may be more prone to fragmentation during retrieval.

OBJECTIVE

To investigate whether bifurcated thrombus patterns are associated with EVT procedural and clinical outcomes.

METHODS

Occlusion patterns of MCA thrombi on CT angiography from MR CLEAN Registry patients were classified into three groups: main stem occlusion, bifurcation occlusion extending into one M2 branch, and bifurcation occlusion extending into both M2 branches. Procedural parameters, procedural outcomes (reperfusion grade and embolization to new territory), and clinical outcomes (24-48 hour National Institutes of Health Stroke Scale [NIHSS_FU] score, change in NIHSS scores between 24 and 48 hours and baseline ∆ [NIHSS], and 90-day modified Rankin Scale [mRS] scores) were compared between occlusion patterns.

RESULTS

We identified 1023 patients with an MCA occlusion of whom 370 (36%) had a main stem occlusion, 151 (15%) a single branch, and 502 (49%) a double branch bifurcation occlusion. There were no statistically significant differences in retrieval method, procedure time, number of retrieval attempts, reperfusion grade, and embolization to new territory between occlusion patterns. Patients with main stem occlusions had lower NIHSS_FU scores than patients with single (7 vs 11, p=0.01) or double branch occlusions (7 vs 9, p=0.04). However, there were no statistically significant differences in ∆ NIHSS or in 90-day mRS scores.

CONCLUSIONS

In our population, EVT procedural and long-term clinical outcomes were similar for MCA bifurcation occlusions and MCA main stem occlusions.

Association between thrombus composition and stroke etiology in the MR CLEAN Registry biobank

PURPOSE

The composition of thrombi retrieved during endovascular thrombectomy (EVT) in acute ischemic stroke (AIS) due to large vessel occlusion (LVO) may differ depending on their origin. In this study, we investigated the association between thrombus composition and stroke etiology in a large population of patients from the Dutch MR CLEAN Registry treated with EVT in daily clinical practice.

METHODS

The thrombi of 332 patients with AIS were histologically analyzed for red blood cells (RBC), fibrin/platelets (F/P), and white blood cells (leukocytes) using a machine learning algorithm. Stroke etiology was assessed using the Trial of Org 10,172 in acute stroke treatment (TOAST) classification.

RESULTS

The thrombi of cardioembolic origin contained less RBC and more F/P than those of non-cardioembolic origin (25.8% vs 41.2% RBC [p = 0.003] and 67.1% vs 54.5% F/P [p = 0.004]). The likelihood of a non-cardioembolic source of stroke increased with increasing thrombus RBC content (OR 1.02; [95% CI 1.00-1.06] for each percent increase) and decreased with a higher F/P content (OR 1.02; [95% CI 1.00-1.06]). Thrombus composition in patients with a cardioembolic origin and undetermined origin was similar.

CONCLUSION

Thrombus composition is significantly associated with stroke etiology, with an increase in RBC and a decrease in F/P raising the odds for a non-cardioembolic cause. No difference between composition of cardioembolic thrombi and of undetermined origin was seen. This emphasizes the need for more extensive monitoring for arrhythmias and/or extended cardiac analysis in case of an undetermined origin.

Agreement between estimated computed tomography perfusion ischemic core and follow-up infarct on diffusion-weighted imaging

BACKGROUND

Computed tomography perfusion (CTP) is frequently performed during the diagnostic workup of acute ischemic stroke patients. Yet, ischemic core estimates vary widely between different commercially available software packages. We assessed the volumetric and spatial agreement of the ischemic core on CTP with the follow-up infarct on diffusion-weighted imaging (DWI) using an automated software.

METHODS

We included successfully reperfused patients who underwent endovascular treatment (EVT) with CTP and follow-up DWI between November 2017 and September 2020. CTP data were processed with a fully automated software using relative cerebral blood flow (rCBF) < 30% to estimate the ischemic core. The follow-up infarct was segmented on DWI imaging data, which were acquired at approximately 24 h. Ischemic core on CTP was compared with the follow-up infarct lesion on DWI using intraclass correlation coefficient (ICC) and Dice similarity coefficient (Dice).

RESULTS

In 59 patients, the median estimated core volume on CTP was 16 (IQR 8-47) mL. The follow-up infarct volume on DWI was 11 (IQR 6-42) mL. ICC was 0.60 (95% CI 0.33-0.76), indicating moderate volumetric agreement. Median Dice was 0.20 (IQR 0.01-0.35). The median positive predictive value was 0.24 (IQR 0.05-0.57), and the median sensitivity was 0.3 (IQR 0.13-0.47). Severe core overestimation on computed tomography perfusion > 50 mL occurred in 4/59 (7%) of the cases.

CONCLUSIONS

In patients with successful reperfusion after EVT, CTP-estimated ischemic core showed moderate volumetric and spatial agreement with the follow-up infarct lesion on DWI, similar to the most used commercially available CTP software packages. Severe ischemic core overestimation was relatively uncommon.

Accuracy of CT perfusion ischemic core volume and location estimation: A comparison between four ischemic core estimation approaches using syngo.via

BACKGROUND AND OBJECTIVE

Computed tomography perfusion (CTP) is widely used in the evaluation of acute ischemic stroke patients for endovascular thrombectomy (EVT). The stability of CTP core estimation is suboptimal and varies between software packages. We aimed to quantify the volumetric and spatial agreement between the CTP ischemic core and follow-up infarct for four ischemic core estimation approaches using syngo.via.

METHODS

We included successfully reperfused, EVT-treated patients with baseline CTP and 24h follow-up diffusion weighted magnetic resonance imaging (DWI) (November 2017-September 2020). Data were processed with syngo.via VB40 using four core estimation approaches based on: cerebral blood volume (CBV)<1.2mL/100mL with and without smoothing filter, relative cerebral blood flow (rCBF)<30%, and rCBF<20%. The follow-up infarct was segmented on DWI.

RESULTS

In 59 patients, median estimated CTP core volumes for four core estimation approaches ranged from 12-39 mL. Median 24h follow-up DWI infarct volume was 11 mL. The intraclass correlation coefficient (ICC) showed moderate-good volumetric agreement for all approaches (range 0.61-0.76). Median Dice was low for all approaches (range 0.16-0.21). CTP core overestimation >10mL occurred least frequent (14/59 [24%] patients) using the CBV-based core estimation approach with smoothing filter.

CONCLUSIONS

In successfully reperfused patients who underwent EVT, syngo.via CTP ischemic core estimation showed moderate volumetric and spatial agreement with the follow-up infarct on DWI. In patients with complete reperfusion after EVT, the volumetric agreement was excellent. A CTP core estimation approach based on CBV<1.2 mL/100mL with smoothing filter least often overestimated the follow-up infarct volume and is therefore preferred for clinical decision making using syngo.via.

Determinants of Symptomatic Intracranial Hemorrhage After Endovascular Stroke Treatment: A Retrospective Cohort Study

Symptomatic intracranial hemorrhage (sICH) is a serious complication after endovascular treatment for ischemic stroke. We aimed to identify determinants of its occurrence and location.

Estimation of treatment effects in observational stroke care data: comparison of statistical approaches

Introduction

Various statistical approaches can be used to deal with unmeasured confounding when estimating treatment effects in observational studies, each with its own pros and cons. This study aimed to compare treatment effects as estimated by different statistical approaches for two interventions in observational stroke care data.

Patients and methods

We used prospectively collected data from the MR CLEAN registry including all patients (n = 3279) with ischemic stroke who underwent endovascular treatment (EVT) from 2014 to 2017 in 17 Dutch hospitals. Treatment effects of two interventions – i.e., receiving an intravenous thrombolytic (IVT) and undergoing general anesthesia (GA) before EVT – on good functional outcome (modified Rankin Scale ≤2) were estimated. We used three statistical regression-based approaches that vary in assumptions regarding the source of unmeasured confounding: individual-level (two subtypes), ecological, and instrumental variable analyses. In the latter, the preference for using the interventions in each hospital was used as an instrument.

Results

Use of IVT (range 66–87%) and GA (range 0–93%) varied substantially between hospitals. For IVT, the individual-level (OR ~ 1.33) resulted in significant positive effect estimates whereas in instrumental variable analysis no significant treatment effect was found (OR 1.11; 95% CI 0.58–1.56). The ecological analysis indicated no statistically significant different likelihood (β = − 0.002%; P = 0.99) of good functional outcome at hospitals using IVT 1% more frequently. For GA, we found non-significant opposite directions of points estimates the treatment effect in the individual-level (ORs ~ 0.60) versus the instrumental variable approach (OR = 1.04). The ecological analysis also resulted in a non-significant negative association (0.03% lower probability).

Discussion and conclusion

Both magnitude and direction of the estimated treatment effects for both interventions depend strongly on the statistical approach and thus on the source of (unmeasured) confounding. These issues should be understood concerning the specific characteristics of data, before applying an approach and interpreting the results. Instrumental variable analysis might be considered when unobserved confounding and practice variation is expected in observational multicenter studies.

Occult blood flow patterns distal to an occluded artery in acute ischemic stroke

Residual blood flow distal to an arterial occlusion in patients with acute ischemic stroke (AIS) is associated with favorable patient outcome. Both collateral flow and thrombus permeability may contribute to such residual flow. We propose a method for discriminating between these two mechanisms, based on determining the direction of flow in multiple branches distal to the occluding thrombus using dynamic Computed Tomography Angiography (dynamic CTA). We analyzed dynamic CTA data of 30 AIS patients and present patient-specific cases that identify typical blood flow patterns and velocities. We distinguished patterns with anterograde (N = 10), retrograde (N = 9), and both flow directions (N = 11), with a large variability in velocities for each flow pattern. The observed flow patterns reflect the interplay between permeability and collaterals. The presented method characterizes distal flow and provides a tool to study patient-specific distal tissue perfusion.

Association of thrombus density and endovascular treatment outcomes in patients with acute ischemic stroke due to M1 occlusions

PURPOSE

We aimed to study the association of non-contrast CT (NCCT) thrombus density with procedural and clinical outcomes in patients with acute ischemic stroke who underwent endovascular treatment (EVT). Since thrombus density is associated with thrombus location, we focused on M1 occlusions only.

METHODS

Patients with available thin-slice (< 2.5 mm) NCCT were included from a nationwide registry. Regression models were used to assess the relation between thrombus density (per Hounsfield unit [HU]) and the following outcomes. For reperfusion grade, adjusted common odds ratios (acOR) indicated a 1-step shift towards improved outcome per HU increase in thrombus density. For the binary outcomes of first-pass reperfusion (first-pass extended thrombolysis in cerebral infarction [eTICI] 2C-3, FPR), functional independence [90-day modified Rankin Scale (mRS) score of 0-2] and mortality), aORs were reported. Adjusted β coefficients (aβ) were reported for 24-h NIHSS and procedure duration in minutes. Outcome differences between first-line treatment devices (stent retriever versus aspiration) were assessed with interaction terms.

RESULTS

In 566 patients with M1 occlusions, thrombus density was not associated with reperfusion (acOR 1.01, 95% CI 0.99-1.02), FPR (aOR 1.01, 95% CI 0.99-1.03), mortality (aOR 0.98, 95% CI 0.95-1.00), 24-h NIHSS (aβ - 0.7%, 95% CI - 1.4-0.2), or procedure duration (aβ 0.27, 95% CI - 0.05-0.58). In multivariable analysis, thrombus density was associated with functional independence (aOR 1.02, 95% CI 1.00-1.05). No interaction was found between thrombus density and first-line treatment device for any outcome.

CONCLUSION

In patients with M1 occlusions, thrombus density was not clearly associated with procedural and clinical outcomes after EVT.

A Randomized Trial of Intravenous Alteplase before Endovascular Treatment for Stroke

BACKGROUND

The value of administering intravenous alteplase before endovascular treatment (EVT) for acute ischemic stroke has not been studied extensively, particularly in non-Asian populations.

METHODS

We performed an open-label, multicenter, randomized trial in Europe involving patients with stroke who presented directly to a hospital that was capable of providing EVT and who were eligible for intravenous alteplase and EVT. Patients were randomly assigned in a 1:1 ratio to receive EVT alone or intravenous alteplase followed by EVT (the standard of care). The primary end point was functional outcome on the modified Rankin scale (range, 0 [no disability] to 6 [death]) at 90 days. We assessed the superiority of EVT alone over alteplase plus EVT, as well as noninferiority by a margin of 0.8 for the lower boundary of the 95% confidence interval for the odds ratio of the two trial groups. Death from any cause and symptomatic intracerebral hemorrhage were the main safety end points.

RESULTS

The analysis included 539 patients. The median score on the modified Rankin scale at 90 days was 3 (interquartile range, 2 to 5) with EVT alone and 2 (interquartile range, 2 to 5) with alteplase plus EVT. The adjusted common odds ratio was 0.84 (95% confidence interval [CI], 0.62 to 1.15; P = 0.28), which showed neither superiority nor noninferiority of EVT alone. Mortality was 20.5% with EVT alone and 15.8% with alteplase plus EVT (adjusted odds ratio, 1.39; 95% CI, 0.84 to 2.30). Symptomatic intracerebral hemorrhage occurred in 5.9% and 5.3% of the patients in the respective groups (adjusted odds ratio, 1.30; 95% CI, 0.60 to 2.81).

CONCLUSIONS

In a randomized trial involving European patients, EVT alone was neither superior nor noninferior to intravenous alteplase followed by EVT with regard to disability outcome at 90 days after stroke. The incidence of symptomatic intracerebral hemorrhage was similar in the two groups. (Funded by the Collaboration for New Treatments of Acute Stroke consortium and others; MR CLEAN-NO IV ISRCTN number, ISRCTN80619088.).

Value of infarct location in the prediction of functional outcome in patients with an anterior large vessel occlusion: results from the HERMES study

Purpose

Follow-up infarct volume (FIV) is moderately associated with functional outcome. We hypothesized that accounting for infarct location would strengthen the association of FIV with functional outcome.

Methods

We included 252 patients from the HERMES collaboration with follow-up diffusion weighted imaging. Patients received endovascular treatment combined with best medical management (n = 52%) versus best medical management alone (n = 48%). FIV was quantified in low, moderate and high modified Rankin Scale (mRS)-relevant regions. We used binary logistic regression to study the relation between the total, high, moderate or low mRS-relevant FIVs and favorable outcome (mRS < 2) after 90 days. The strength of association was evaluated using the c-statistic.

Results

Small lesions only occupied high mRS-relevant brain regions. Lesions additionally occupied lower mRS-relevant brain regions if FIV expanded. Higher FIV was associated with a higher risk of unfavorable outcome, as were volumes of tissue with low, moderate and high mRS relevance. In multivariable modeling, only the volume of high mRS-relevant infarct was significantly associated with favorable outcome. The c-statistic was highest (0.76) for the models that included high mRS-relevant FIV or the combination of high, moderate and low mRS-relevant FIV but was not significantly different from the model that included only total FIV (0.75).

Conclusion

This study confirms the association of FIV and unfavorable functional outcome but showed no strengthened association if lesion location was taken into account.

Endovascular Treatment Effect Diminishes With Increasing Thrombus Perviousness: Pooled Data From 7 Trials on Acute Ischemic Stroke

Supplemental Digital Content is available in the text.

Background and Purpose:

Thrombus perviousness estimates residual flow along a thrombus in acute ischemic stroke, based on radiological images, and may influence the benefit of endovascular treatment for acute ischemic stroke. We aimed to investigate potential endovascular treatment (EVT) effect modification by thrombus perviousness.

Methods:

We included 443 patients with thin-slice imaging available, out of 1766 patients from the pooled HERMES (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke trials) data set of 7 randomized trials on EVT in the early window (most within 8 hours). Control arm patients (n=233) received intravenous alteplase if eligible (212/233; 91%). Intervention arm patients (n=210) received additional EVT (prior alteplase in 178/210; 85%). Perviousness was quantified by thrombus attenuation increase on admission computed tomography angiography compared with noncontrast computed tomography. Multivariable regression analyses were performed including multiplicative interaction terms between thrombus attenuation increase and treatment allocation. In case of significant interaction, subgroup analyses by treatment arm were performed. Our primary outcome was 90-day functional outcome (modified Rankin Scale score), resulting in an adjusted common odds ratio for a one-step shift towards improved outcome. Secondary outcomes were mortality, successful reperfusion (extended Thrombolysis in Cerebral Infarction score, 2B–3), and follow-up infarct volume (in mL).

Results:

Increased perviousness was associated with improved functional outcome. After adding a multiplicative term of thrombus attenuation increase and treatment allocation, model fit improved significantly (P=0.03), indicating interaction between perviousness and EVT benefit. Control arm patients showed significantly better outcomes with increased perviousness (adjusted common odds ratio, 1.2 [95% CI, 1.1–1.3]). In the EVT arm, no significant association was found (adjusted common odds ratio, 1.0 [95% CI, 0.9–1.1]), and perviousness was not significantly associated with successful reperfusion. Follow-up infarct volume (12% [95% CI, 7.0–17] per 5 Hounsfield units) and chance of mortality (adjusted odds ratio, 0.83 [95% CI, 0.70–0.97]) decreased with higher thrombus attenuation increase in the overall population, without significant treatment interaction.

Conclusions:

Our study suggests that the benefit of best medical care including alteplase, compared with additional EVT, increases in patients with more pervious thrombi.

Influence of Onset to Imaging Time on Radiological Thrombus Characteristics in Acute Ischemic Stroke

Introduction: Radiological thrombus characteristics are associated with patient outcomes and treatment success after acute ischemic stroke. These characteristics could be expected to undergo time-dependent changes due to factors influencing thrombus architecture like blood stasis, clot contraction, and natural thrombolysis. We investigated whether stroke onset-to-imaging time was associated with thrombus length, perviousness, and density in the MR CLEAN Registry population.

Methods: We included 245 patients with M1-segment occlusions and thin-slice baseline CT imaging from the MR CLEAN Registry, a nation-wide multicenter registry of patients who underwent endovascular treatment for acute ischemic stroke within 6.5 h of onset in the Netherlands. We used multivariable linear regression to investigate the effect of stroke onset-to-imaging time (per 5 min) on thrombus length (in mm), perviousness and density (both in Hounsfield Units). In the first model, we adjusted for age, sex, intravenous thrombolysis, antiplatelet use, and history of atrial fibrillation. In a second model, we additionally adjusted for observed vs. non-observed stroke onset, CT-angiography collateral score, direct presentation at a thrombectomy-capable center vs. transfer, and stroke etiology. We performed exploratory subgroup analyses for intravenous thrombolysis administration, observed vs. non-observed stroke onset, direct presentation vs. transfer, and stroke etiology.

Results: Median stroke onset-to-imaging time was 83 (interquartile range 53–141) min. Onset to imaging time was not associated with thrombus length nor perviousness (β 0.002; 95% CI −0.004 to 0.007 and β −0.002; 95% CI −0.015 to 0.011 per 5 min, respectively) and was weakly associated with thrombus density in the fully adjusted model (adjusted β 0.100; 95% CI 0.005–0.196 HU per 5 min). The subgroup analyses showed no heterogeneity of these findings in any of the subgroups, except for a significantly positive relation between onset-to-imaging time and thrombus density in patients transferred from a primary stroke center (adjusted β 0.18; 95% CI 0.022–0.35).

Conclusion: In our population of acute ischemic stroke patients, we found no clear association between onset-to-imaging time and radiological thrombus characteristics. This suggests that elapsed time from stroke onset plays a limited role in the interpretation of radiological thrombus characteristics and their effect on treatment results, at least in the early time window.

Endovascular treatment for calcified cerebral emboli in patients with acute ischemic stroke

OBJECTIVE

Calcified cerebral emboli (CCE) are a rare cause of acute ischemic stroke. The authors aimed to assess the association of CCE with functional outcome, successful reperfusion, and mortality. Furthermore, they aimed to assess the effectiveness of intravenous alteplase treatment and endovascular treatment (EVT), as well as the best first-line EVT approach in patients with CCE.

METHODS

The Multicenter Randomized Controlled Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN) Registry is a prospective, observational multicenter registry of patients treated with EVT for acute ischemic stroke in 16 intervention hospitals in the Netherlands. The association of CCE with functional outcome, reperfusion, and mortality was evaluated using logistic regression models. Univariable comparisons were made to determine the effectiveness of intravenous alteplase treatment and the best first-line EVT approach in CCE patients.

RESULTS

The study included 3077 patients from the MR CLEAN Registry. Fifty-five patients (1.8%) had CCE. CCE were not significantly associated with worse functional outcome (adjusted common OR 0.71, 95% CI 0.44-1.15), and 29% of CCE patients achieved functional independence. An extended Thrombolysis in Cerebral Infarction score ≥ 2B was significantly less often achieved in CCE patients compared to non-CCE patients (adjusted OR [aOR] 0.52, 95% CI 0.28-0.97). Symptomatic intracranial hemorrhage occurred in 8 CCE patients (15%) vs 171 of 3022 non-CCE patients (6%; p = 0.01). The median improvement on the National Institutes of Health Stroke Scale (NIHSS) was 2 in CCE patients versus 4 in non-CCE patients (p = 0.008). CCE were not significantly associated with mortality (aOR 1.16, 95% CI 0.64-2.12). Intravenous alteplase use in CCE patients was not associated with functional outcome or reperfusion. In CCE patients with successful reperfusion, stent retrievers were more often used as the primary treatment device (p = 0.04).

CONCLUSIONS

While patients with CCE had significantly lower reperfusion rates and less improvement on the NIHSS after EVT, CCE were not significantly associated with worse functional outcome or higher mortality rates. Therefore, EVT should still be considered in this specific group of patients.

Automatic segmentation of cerebral infarcts in follow-up computed tomography images with convolutional neural networks

Background and purpose

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.

Objective

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.

Materials and methods

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.

Results

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.

Conclusion

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.

Collateral Circulation and Outcome in Atherosclerotic Versus Cardioembolic Cerebral Large Vessel Occlusion

Supplemental Digital Content is available in the text.

Due to chronic hypoperfusion, cervical atherosclerosis may promote cerebral collateral circulation. We hypothesized that patients with ischemic stroke due to cervical carotid atherosclerosis have a more extensive collateral circulation and better outcomes than patients with cardioembolism. We tested this hypothesis in a population of patients who underwent endovascular treatment for large vessel occlusion.

Data-efficient deep learning of radiological image data for outcome prediction after endovascular treatment of patients with acute ischemic stroke

Treatment selection is becoming increasingly more important in acute ischemic stroke patient care. Clinical variables and radiological image biomarkers (old age, pre-stroke mRS, NIHSS, occlusion location, ASPECTS, among others) have an important role in treatment selection and prognosis. Radiological biomarkers require expert annotation and are subject to inter-observer variability. Recently, Deep Learning has been introduced to reproduce these radiological image biomarkers. Instead of reproducing these biomarkers, in this work, we investigated Deep Learning techniques for building models to directly predict good reperfusion after endovascular treatment (EVT) and good functional outcome using CT angiography images. These models do not require image annotation and are fast to compute. We compare the Deep Learning models to Machine Learning models using traditional radiological image biomarkers. We explored Residual Neural Network (ResNet) architectures, adapted them with Structured Receptive Fields (RFNN) and auto-encoders (AE) for network weight initialization. We further included model visualization techniques to provide insight into the network's decision-making process. We applied the methods on the MR CLEAN Registry dataset with 1301 patients. The Deep Learning models outperformed the models using traditional radiological image biomarkers in three out of four cross-validation folds for functional outcome (average AUC of 0.71) and for all folds for reperfusion (average AUC of 0.65). Model visualization showed that the arteries were relevant features for functional outcome prediction. The best results were obtained for the ResNet models with RFNN. Auto-encoder initialization often improved the results. We concluded that, in our dataset, automated image analysis with Deep Learning methods outperforms radiological image biomarkers for stroke outcome prediction and has the potential to improve treatment selection.

Thrombus Imaging Characteristics and Outcomes in Acute Ischemic Stroke Patients Undergoing Endovascular Treatment

Background and Purpose—

Thrombus imaging characteristics have been reported to be useful to predict functional outcome and reperfusion in acute ischemic stroke. However, conflicting data about this subject exist in patients undergoing endovascular treatment. Therefore, we aimed to evaluate whether thrombus imaging characteristics assessed on computed tomography are associated with outcomes in patients with acute ischemic stroke treated by endovascular treatment.

Methods—

The MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands) Registry is an ongoing, prospective, and observational study in all centers performing endovascular treatment in the Netherlands. We evaluated associations of thrombus imaging characteristics with the functional outcome (modified Rankin Scale at 90 days), mortality, reperfusion, duration of endovascular treatment, and symptomatic intracranial hemorrhage using univariable and multivariable regression models. Thrombus characteristics included location, clot burden score (CBS), length, relative and absolute attenuation, perviousness, and distance from the internal carotid artery terminus to the thrombus. All characteristics were assessed on thin-slice (≤2.5 mm) noncontrast computed tomography and computed tomography angiography, acquired within 30 minutes from each other.

Results—

In total, 408 patients were analyzed. Thrombus with distal location, higher CBS, and shorter length were associated with better functional outcome (adjusted common odds ratio, 3.3; 95% CI, 2.0–5.3 for distal M1 occlusion compared with internal carotid artery occlusion; adjusted common odds ratio, 1.15; 95% CI, 1.07–1.24 per CBS point; and adjusted common odds ratio, 0.96; 95% CI, 0.94–0.99 per mm, respectively) and reduced duration of endovascular procedure (adjusted coefficient B, −14.7; 95% CI, −24.2 to −5.1 for distal M1 occlusion compared with internal carotid artery occlusion; adjusted coefficient B, −8.5; 95% CI, −14.5 to −2.4 per CBS point; and adjusted coefficient B, 7.3; 95% CI, 2.9–11.8 per mm, respectively). Thrombus perviousness was associated with better functional outcome (adjusted common odds ratio, 1.01; 95% CI, 1.00–1.02 per Hounsfield units increase). Distal thrombi were associated with successful reperfusion (adjusted odds ratio, 2.6; 95% CI, 1.4–4.9 for proximal M1 occlusion compared with internal carotid artery occlusion).

Conclusions—

Distal location, higher CBS, and shorter length are associated with better functional outcome and faster endovascular procedure. Distal thrombus is strongly associated with successful reperfusion, and a pervious thrombus is associated with better functional outcome.

Abstract WP70: Artificial Intelligence Based Detection of Large Vessel Occlusion on Non-Contrast Computed Tomography in Stroke

Introduction: Rapid detection of a large vessel occlusion (LVO) in acute ischemic stroke is beneficial as it may accelerate therapeutic management. The hyperdense artery sign (HAS) on non-contrast CT (NCCT) is result of an acute thrombolytic occlusion and is known as the earliest radiological marker to identify LVO. However, HAS is not apparent for all LVOs and detection by a less-trained eye can be sub-optimal. Artificial intelligence (AI) may support the non-trained radiologist in swift and accurate LVO detection. We evaluated the LVO detection accuracy of AI, and compared this to human expert performance.

Materials and Methods: We used a convolutional neural network (CNN) developed by Nico.lab (www.nico-lab.com) that automatically detects and segments thrombi on NCCT using a patch-based approach. The CNN was trained using thrombus and non-thrombus patches, combined with the contralateral side. In a retrospective analysis of thin-slice NCCTs, obtained from Amsterdam UMC patients, with (ICA-T, M1, or M2) and without occlusions (stroke mimics), the CNN and two observers (>15y and >4y of experience) assessed LVO-presence and location. Ground truth was established by consensus between two different experts with >5y of experience, using both CTA and NCCT. Thrombus segmentations were considered accurate if segmentations overlapped with the ground truth. Sensitivity and specificity of the LVO detection were assessed.

Results: We included 107 patients, of which 59 had proven LVOs. Nico.lab CNN showed a thrombus segmentation accuracy of 81%, vs. 81% and 77% of human experts. Sensitivity of LVO detection by CNN was 0.86, vs. 0.95 and 0.79 for human observers. Specificity was 0.65 for the CNN vs. 0.58 and 0.82 for human experts respectively.

Conclusion: AI-based LVO detection on NCCT showed comparable results to expert observers. This supportive tool could lead to earlier detection of LVO in acute stroke situations. Fig 1A. HAS; B. CNN output (red); C. clot on CTA (blue arrow)

Cranial shape comparison for automated objective 3D craniosynostosis surgery planning

Virtual planning of open cranial vault reconstruction is used to simulate and define an pre-operative plan for craniosynostosis surgery. However, virtual planning techniques are subjective and dependent on the experience and preferences of the surgical team. To develop an objective automated 3D pre-operative planning technique for open cranial vault reconstructions, we used curvature maps for the shape comparison of the patient's skull with an age-specific reference skull. We created an average skull for the age-group of 11-14 months. Also, we created an artificial test object and selected a cranial CT-scan of an 11 months old trigonocephaly patient as test case. Mesh data of skulls were created using marching cubes and raycasting. Curvature maps were computed using quadric surface fitting. The shape comparison was tested for the test object and the average skull. Finally, shape comparison was performed for the trigonocephalic skull with the average skull. Similar shapes and the area on the patient's skull that maximally corresponded in shape with the reference shape were correctly identified. This study showed that curvature maps allow the comparison of craniosynostosis skulls with age-appropriate average skulls and a first step towards an objective user-independent pre-operative planning technique for open cranial vault reconstructions is made.