Automated identification of thrombectomy amenable vessel occlusion on computed tomography angiography using deep learning.

Introduction: We developed and externally validated a fully automated algorithm using deep learning to detect large vessel occlusion (LVO) in computed tomography angiography (CTA). Method: A total of 2,045 patients with acute ischemic stroke who underwent CTA were included in the development of our model. We validated the algorithm using two separate external datasets: one with 64 patients (external 1) and another with 313 patients (external 2), with ischemic stroke. In the context of current clinical practice, thrombectomy amenable vessel occlusion (TAVO) was defined as an occlusion in the intracranial internal carotid artery (ICA), or in the M1 or M2 segment of the middle cerebral artery (MCA). We employed the U-Net for vessel segmentation on the maximum intensity projection images, followed by the application of the EfficientNetV2 to predict TAVO. The algorithm's diagnostic performance was evaluated by calculating the area under the receiver operating characteristics curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results: The mean age in the training and validation dataset was 68.7 ± 12.6; 56.3% of participants were men, and 18.0% had TAVO. The algorithm achieved AUC of 0.950 (95% CI, 0.915-0.971) in the internal test. For the external datasets 1 and 2, the AUCs were 0.970 (0.897-0.997) and 0.971 (0.924-0.990), respectively. With a fixed sensitivity of 0.900, the specificities and PPVs for the internal test, external test 1, and external test 2 were 0.891, 0.796, and 0.930, and 0.665, 0.583, and 0.667, respectively. The algorithm demonstrated a sensitivity and specificity of approximately 0.95 in both internal and external datasets, specifically for cases involving intracranial ICA or M1-MCA occlusion. However, the diagnostic performance was somewhat reduced for isolated M2-MCA occlusion; the AUC for the internal and combined external datasets were 0.903 (0.812-0.944) and 0.916 (0.816-0.963), respectively. Conclusion: We developed and externally validated a fully automated algorithm that identifies TAVO. Further research is needed to evaluate its effectiveness in real-world clinical settings. This validated algorithm has the potential to assist early-career physicians, thereby streamlining the treatment process for patients who can benefit from endovascular treatment.

Automated segmentation of MRI white matter hyperintensities in 8,421 patients with acute ischemic stroke.

BACKGROUND AND PURPOSE: To date, only a few small studies have attempted deep learning-based automatic segmentation of white matter hyperintensity (WMH) lesions in patients with cerebral infarction, which is complicated because stroke-related lesions can obscure WMH borders. We developed and validated deep learning algorithms to segment WMH lesions accurately in patients with cerebral infarction, using multisite datasets involving 8,421 patients with acute ischemic stroke. MATERIALS AND METHODS: We included 8,421 stroke patients from 9 centers in Korea. 2D UNet and SE-Unet models were trained using 2,408 FLAIR MRI from 3 hospitals and validated using 6,013 FLAIR MRIs from 6 hospitals. WMH segmentation performance was assessed by calculating DSC, correlation coefficient, and concordance correlation coefficient compared to a human-segmented gold standard. In addition, we obtained an uncertainty index that represents overall ambiguity in the voxel classification for WMH segmentation in each patient based on the Kullback-Leibler divergence. RESULTS: In the training dataset, the mean age was 67.4±13.0 years and 60.4% were men. The mean (95% CI) DSCs for Unet in internal testing and external validation were respectively 0.659 (0.649-0.669) and 0.710 (0.707-0.714), which were slightly lower than the reliability between humans (DSC=0.744; 95% CI=0.738-0.751; P=.031). Compared with the Unet, the SE-Unet demonstrated better performance, achieving a mean DSC of 0.675 (0.666-0.685; P<.001) in the internal testing and 0.722 (0.719-0.726; P<.001) in the external validation; moreover, it achieved high DSC values (ranging from 0.672 to 0.744) across multiple validation datasets. We observed a significant correlation between WMH volumes that were segmented automatically and manually for the Unet (r=0.917, P<.0001) and even stronger for the SE-Unet (r=0.933, P<.0001). The SE-Unet also attained a high concordance correlation coefficient (ranging from 0.841 to 0.956) in external test datasets. In addition, the uncertainty indices in the majority of patients (86%) in the external datasets were below 0.35, with an average DSC of 0.744 in these patients. CONCLUSIONS: We developed and validated deep learning algorithms to segment WMH in patients with acute cerebral infarction using the largest-ever MRI datasets. In addition, we showed that the uncertainty index can be used to identify cases where automatic WMH segmentation is less accurate and requires human review. ABBREVIATIONS: WMH = white matter hyperintensity; CNN = convolutional neural networks; SE = squeeze-and-excitation; KL = Kullback-Leibler; ReLU = rectified linear unit; LKW = last known well; mRS = modified Rankin Scale; NIHSS = National Institute of Health Stroke Scale; LAA = large artery atherosclerosis; SVO = small vessel occlusion; CE = cardioembolism.

Better performance of deep learning pulmonary nodule detection using chest radiography with pixel level labels in reference to computed tomography: data quality matters.

Labeling errors can significantly impact the performance of deep learning models used for screening chest radiographs. The deep learning model for detecting pulmonary nodules is particularly vulnerable to such errors, mainly because normal chest radiographs and those with nodules obscured by ribs appear similar. Thus, high-quality datasets referred to chest computed tomography (CT) are required to prevent the misclassification of nodular chest radiographs as normal. From this perspective, a deep learning strategy employing chest radiography data with pixel-level annotations referencing chest CT scans may improve nodule detection and localization compared to image-level labels. We trained models using a National Institute of Health chest radiograph-based labeling dataset and an AI-HUB CT-based labeling dataset, employing DenseNet architecture with squeeze-and-excitation blocks. We developed four models to assess whether CT versus chest radiography and pixel-level versus image-level labeling would improve the deep learning model's performance to detect nodules. The models' performance was evaluated using two external validation datasets. The AI-HUB dataset with image-level labeling outperformed the NIH dataset (AUC 0.88 vs 0.71 and 0.78 vs. 0.73 in two external datasets, respectively; both p < 0.001). However, the AI-HUB data annotated at the pixel level produced the best model (AUC 0.91 and 0.86 in external datasets), and in terms of nodule localization, it significantly outperformed models trained with image-level annotation data, with a Dice coefficient ranging from 0.36 to 0.58. Our findings underscore the importance of accurately labeled data in developing reliable deep learning algorithms for nodule detection in chest radiography.

Deep Learning-Based Automatic Classification of Ischemic Stroke Subtype Using Diffusion-Weighted Images.

BACKGROUND AND PURPOSE: Accurate classification of ischemic stroke subtype is important for effective secondary prevention of stroke. We used diffusion-weighted image (DWI) and atrial fibrillation (AF) data to train a deep learning algorithm to classify stroke subtype. METHODS: Model development was done in 2,988 patients with ischemic stroke from three centers by using U-net for infarct segmentation and EfficientNetV2 for subtype classification. Experienced neurologists (n=5) determined subtypes for external test datasets, while establishing a consensus for clinical trial datasets. Automatically segmented infarcts were fed into the model (DWI-only algorithm). Subsequently, another model was trained, with AF included as a categorical variable (DWI+AF algorithm). These models were tested: (1) internally against the opinion of the labeling experts, (2) against fresh external DWI data, and (3) against clinical trial dataset. RESULTS: In the training-and-validation datasets, the mean (±standard deviation) age was 68.0±12.5 (61.1% male). In internal testing, compared with the experts, the DWI-only and the DWI+AF algorithms respectively achieved moderate (65.3%) and near-strong (79.1%) agreement. In external testing, both algorithms again showed good agreements (59.3%-60.7% and 73.7%-74.0%, respectively). In the clinical trial dataset, compared with the expert consensus, percentage agreements and Cohen's kappa were respectively 58.1% and 0.34 for the DWI-only vs. 72.9% and 0.57 for the DWI+AF algorithms. The corresponding values between experts were comparable (76.0% and 0.61) to the DWI+AF algorithm. CONCLUSION: Our model trained on a large dataset of DWI (both with or without AF information) was able to classify ischemic stroke subtypes comparable to a consensus of stroke experts.

Comparison of two automated CT perfusion software packages in patients with ischemic stroke presenting within 24 h of onset.

Background: We compared the ischemic core and hypoperfused tissue volumes estimated by RAPID and JLK-CTP, a newly developed automated computed tomography perfusion (CTP) analysis package. We also assessed agreement between ischemic core volumes by two software packages against early follow-up infarct volumes on diffusion-weighted images (DWI). Methods: This retrospective study analyzed 327 patients admitted to a single stroke center in Korea from January 2021 to May 2023, who underwent CTP scans within 24 h of onset. The concordance correlation coefficient (ρ) and Bland-Altman plots were utilized to compare the volumes of ischemic core and hypoperfused tissue volumes between the software packages. Agreement with early (within 3 h from CTP) follow-up infarct volumes on diffusion-weighted imaging (n = 217) was also evaluated. Results: The mean age was 70.7 ± 13.0 and 137 (41.9%) were female. Ischemic core volumes by JLK-CTP and RAPID at the threshold of relative cerebral blood flow (rCBF) < 30% showed excellent agreement (ρ = 0.958 [95% CI, 0.949 to 0.966]). Excellent agreement was also observed for time to a maximum of the residue function (T max) > 6 s between JLK-CTP and RAPID (ρ = 0.835 [95% CI, 0.806 to 0.863]). Although early follow-up infarct volume showed substantial agreement in both packages (JLK-CTP, ρ = 0.751 and RAPID, ρ = 0.632), ischemic core volumes at the threshold of rCBF <30% tended to overestimate ischemic core volumes. Conclusion: JLK-CTP and RAPID demonstrated remarkable concordance in estimating the volumes of the ischemic core and hypoperfused area based on CTP within 24 h from onset.

Trends in Dual Antiplatelet Therapy of Aspirin and Clopidogrel and Outcomes in Ischemic Stroke Patients Noneligible for POINT/CHANCE Trial Treatment.

BACKGROUND: Recent clinical trials established the benefit of dual antiplatelet therapy with aspirin and clopidogrel (DAPT-AC) in early-presenting patients with minor ischemic stroke. However, the impact of these trials over time on the use and outcomes of DAPT-AC among the patients with nonminor or late-presenting stroke who do not meet the eligibility criteria of these trials has not been delineated. METHODS AND RESULTS: In a multicenter stroke registry, this study examined yearly changes from April 2008 to August 2022 in DAPT-AC use for stroke patients ineligible for CHANCE/POINT (Clopidogrel in High-Risk Patients with Acute Nondisabling Cerebrovascular Events/Platelet-Oriented Inhibition in New TIA and Minor Ischemic Stroke) clinical trials due to National Institutes of Health Stroke Scale >4 or late arrival beyond 24 hours of onset. A total of 32 118 patients (age, 68.1±13.1 years; male, 58.5%) with National Institutes of Health Stroke Scale of 4 (interquartile range, 1-7) were analyzed. In 2008, DAPT-AC was used in 33.0%, other antiplatelets in 62.7%, and no antiplatelet in 4.3%. The frequency of DAPT-AC was relatively unchanged through 2013, when the CHANCE trial was published, and then increased steadily, reaching 78% in 2022, while other antiplatelets decreased to 17.8% in 2022 (Ptrend<0.001). From 2011 to 2022, clinical outcomes nonsignificantly improved, with an average relative risk reduction of 2%/y for the composite of stroke, myocardial infarction, and all-cause mortality, both among patients treated with DAPT-AC and patients treated with other antiplatelets. CONCLUSIONS: Use of DAPT-AC in stroke patients with stroke ineligible for recent DAPT clinical trials increased markedly and steadily after CHANCE publication in 2013, reaching deployment in nearly 4 of every 5 patients by 2022. The secondary prevention in patients with ischemic stroke seems to be gradually improving, possibly due to the enhancement of risk factor control.

Automated Prediction of Proximal Middle Cerebral Artery Occlusions in Noncontrast Brain Computed Tomography.

BACKGROUND: Early identification of large vessel occlusion (LVO) in patients with ischemic stroke is crucial for timely interventions. We propose a machine learning-based algorithm (JLK-CTL) that uses handcrafted features from noncontrast computed tomography to predict LVO. METHODS: We included patients with ischemic stroke who underwent concurrent noncontrast computed tomography and computed tomography angiography in seven hospitals. Patients from 5 of these hospitals, admitted between May 2011 and March 2015, were randomly divided into training and internal validation (9:1 ratio). Those from the remaining 2 hospitals, admitted between March 2021 and September 2021, were designated for external validation. From each noncontrast computed tomography scan, we extracted differences in volume, tissue density, and Hounsfield unit distribution between bihemispheric regions (striatocapsular, insula, M1-M3, and M4-M6, modified from the Alberta Stroke Program Early Computed Tomography Score). A deep learning algorithm was used to incorporate clot signs as an additional feature. Machine learning models, including ExtraTrees, random forest, extreme gradient boosting, support vector machine, and multilayer perceptron, as well as a deep learning model, were trained and evaluated. Additionally, we assessed the models' performance after incorporating the National Institutes of Health Stroke Scale scores as an additional feature. RESULTS: Among 2919 patients, 83 were excluded. Across the training (n=2463), internal validation (n=275), and external validation (n=95) datasets, the mean ages were 68.5±12.4, 67.6±13.8, and 67.9±13.6 years, respectively. The proportions of men were 57%, 53%, and 59%, with LVO prevalences of 17.0%, 16.4%, and 26.3%, respectively. In the external validation, the ExtraTrees model achieved a robust area under the curve of 0.888 (95% CI, 0.850-0.925), with a sensitivity of 80.1% (95% CI, 72.0-88.1) and a specificity of 88.6% (95% CI, 84.7-92.5). Adding the National Institutes of Health Stroke Scale score to the ExtraTrees model increased sensitivity (from 80.1% to 92.1%) while maintaining specificity. CONCLUSIONS: Our algorithm provides reliable predictions of LVO using noncontrast computed tomography. By enabling early LVO identification, our algorithm has the potential to expedite the stroke workflow.

Brain Frailty and Outcomes of Acute Minor Ischemic Stroke With Large-Vessel Occlusion.

BACKGROUND AND PURPOSE: The influence of imaging features of brain frailty on outcomes were investigated in acute ischemic stroke patients with minor symptoms and large-vessel occlusion (LVO). METHODS: This was a retrospective analysis of a prospective, multicenter, nationwide registry of consecutive patients with acute (within 24 h) minor (National Institutes of Health Stroke Scale score=0-5) ischemic stroke with anterior circulation LVO (acute minor LVO). Brain frailty was stratified according to the presence of an advanced white-matter hyperintensity (WMH) (Fazekas grade 2 or 3), silent/old brain infarct, or cerebral microbleeds. The primary outcome was a composite of stroke, myocardial infarction, and all-cause mortality within 1 year. RESULTS: In total, 1,067 patients (age=67.2±13.1 years [mean±SD], 61.3% males) were analyzed. The proportions of patients according to the numbers of brain frailty burdens were as follows: no burden in 49.2%, one burden in 30.0%, two burdens in 17.3%, and three burdens in 3.5%. In the Cox proportional-hazards analysis, the presence of more brain frailty burdens was associated with a higher risk of 1-year primary outcomes, but after adjusting for clinically relevant variables there were no significant associations between burdens of brain frailty and 1-year vascular outcomes. For individual components of brain frailty, an advanced WMH was independently associated with an increased risk of 1-year primary outcomes (adjusted hazard ratio [aHR]=1.33, 95% confidence interval [CI]=1.03-1.71) and stroke (aHR=1.32, 95% CI=1.00-1.75). CONCLUSIONS: The baseline imaging markers of brain frailty were common in acute minor ischemic stroke patients with LVO. An advanced WMH was the only frailty marker associated with an increased risk of vascular events. Further research is needed into the association between brain frailty and prognosis in patients with acute minor LVO.

Differential impacts of admission LDL-cholesterol on early vascular outcomes by ischemic stroke subtypes.

BACKGROUND: Because ischemic stroke is heterogeneous, the associations between low-density lipoprotein (LDL)-cholesterol levels and early vascular outcomes might be different according to the stroke subtype in acute ischemic stroke patients. METHODS: This study was an analysis of a prospective, multicenter, stroke registry. Acute ischemic stroke patients previously not treated with statins were included. Admission LDL-cholesterol levels were divided into 7 groups at 20 mg/dl intervals for comparison. The primary early vascular outcome was a composite of stroke, myocardial infarction (MI) and all-cause mortality within 3 months. RESULTS: A total of 38,531 patients (age, 68.5 ± 12.8 yrs; male, 59.6%) were analyzed for this study. The 3-month cumulative incidences of the composite of stroke, MI, and all-cause mortality significantly differed among the LDL-cholesterol level groups, with the highest event rate (11.11%) in the lowest LDL-cholesterol group (<70 mg/dl). After adjustment, the U-shaped associations of LDL-cholesterol levels with primary outcome and all-cause mortality were observed. For the stroke subtypes, there were substantial interactions between the LDL-cholesterol groups and stroke subtype and all-cause mortality (Pinteraction=0.07). Different patterns, with higher risks of all-cause mortality in the lower LDL-cholesterol in the large artery atherosclerosis subtype (adjusted hazard ratio [aHR] 1.29, 95% confidence interval [CI] 0.98-1.69), but in the higher LDL-cholesterol in the cardioembolism subtype (aHR 1.71 95% CI [1.28-2.29]), were observed among stroke subtypes. CONCLUSION: We found that there were differential associations of admission LDL-cholesterol levels with all-cause mortality within 3 months among stroke subtypes. These results suggest that admission LDL-cholesterol and early vascular outcomes had complex relationships in patients with ischemic stroke according to the stroke subtypes.

Improvement in Delivery of Ischemic Stroke Treatments but Stagnation of Clinical Outcomes in Young Adults in South Korea.

BACKGROUND: There is limited information on the delivery of acute stroke therapies and secondary preventive measures and clinical outcomes over time in young adults with acute ischemic stroke. This study investigated whether advances in these treatments improved outcomes in this population. METHODS: Using a prospective multicenter stroke registry in Korea, young adults (aged 18-50 years) with acute ischemic stroke hospitalized between 2008 and 2019 were identified. The observation period was divided into 4 epochs: 2008 to 2010, 2011 to 2013, 2014 to 2016, and 2017 to 2019. Secular trends for patient characteristics, treatments, and outcomes were analyzed. RESULTS: A total of 7050 eligible patients (mean age, 43.1; men, 71.9%) were registered. The mean age decreased from 43.6 to 42.9 years (Ptrend=0.01). Current smoking decreased, whereas obesity increased. Other risk factors remained unchanged. Intravenous thrombolysis and mechanical thrombectomy rates increased over time from 2008 to 2010 to 2017 to 2019 (9.5%-13.8% and 3.2%-9.2%, respectively; Ptrend<0.01). Door-to-needle time improved (Ptrend <.001), but onset-to-door and door-to-puncture times remained constant. Secondary prevention, including dual antiplatelets for noncardioembolic minor stroke (26.7%-47.0%), direct oral anticoagulants for atrial fibrillation (0.0%-56.2%), and statins for large artery atherosclerosis (76.1%-95.3%) increased (Ptrend<0.01). Outcome data were available from 2011. One-year mortality (2.5% in 2011-2013 and 2.3% in 2017-2019) and 3-month modified Rankin Scale scores 0 to 1 (68.3%-69.1%) and 0 to 2 (87.6%-86.2%) remained unchanged. The 1-year stroke recurrence rate increased (4.1%-5.5%; Ptrend=0.04), although the difference was not significant after adjusting for sex and age. CONCLUSIONS: Improvements in the delivery of acute stroke treatments did not necessarily lead to better outcomes in young adults with acute ischemic stroke over the past decade, indicating a need for further progress.

Nonhypertensive White Matter Hyperintensities in Stroke: Risk Factors, Neuroimaging Characteristics, and Prognosis.

BACKGROUND: This study explored the risk factors, neuroimaging features, and prognostic implications of nonhypertensive white matter hyperintensity (WMH) in patients with acute ischemic stroke and transient ischemic attack. METHODS AND RESULTS: We included 2283 patients with hypertension and 1003 without from a pool of 10 602. Associations of moderate-to-severe WMH with known risk factors, functional outcome, and a composite of recurrent stroke, myocardial infarction, and all-cause mortality were evaluated. A subset of 351 patients without hypertension and age- and sex-matched pairs with hypertension and moderate-to-severe WMH was created for a detailed topographic examination of WMH, lacunes, and microbleeds. Approximately 35% of patients without hypertension and 65% of patients with hypertensive stroke exhibited moderate-to-severe WMH. WMH was associated with age, female sex, and previous stroke, irrespective of hypertension. In patients without hypertension, WMH was associated with initial systolic blood pressure and was more common in the anterior temporal region. In patients with hypertension, WMH was associated with small vessel occlusion as a stroke mechanism and was more frequent in the periventricular region near the posterior horn of the lateral ventricle. The higher prevalence of occipital microbleeds in patients without hypertension and deep subcortical lacunes in patients with hypertension were also observed. Associations of moderate-to-severe WMH with 3-month functional outcome and 1-year cumulative incidence of the composite outcome were significant (both P<0.01), although the latter lost significance after adjustments. The associations between WMH and outcomes were consistent across hypertensive status. CONCLUSIONS: One-third of patients without hypertension with stroke have moderate-to-severe WMH. The pathogenesis of WMH may differ between patients without and with hypertension, but its impact on outcome appears similar.

Statin Treatment in Patients With Stroke With Low-Density Lipoprotein Cholesterol Levels Below 70 mg/dL.

Background It is unclear whether statin treatment could reduce the risk of early vascular events when baseline low-density lipoprotein cholesterol (LDL-C) levels are already low, at <70 mg/dL, at the time of the index stroke. Methods and Results This study was an analysis of a prospective, multicenter, nationwide registry of consecutive patients with first-ever acute ischemic stroke with baseline low-density lipoprotein cholesterol levels <70 mg/dL and without statin pretreatment. An inverse probabilities of treatment weights method was applied to control for imbalances in baseline characteristics. The primary outcome was a composite of stroke (either hemorrhagic or ischemic), myocardial infarction, and all-cause death within 3 months. A total of 2850 patients (age, 69.5±13.4 years; men, 63.5%) were analyzed for this study. In-hospital statin treatment was used for 74.2% of patients. The primary composite outcome within 3 months occurred in 21.5% of patients in the nonstatin group and 6.7% of patients in the statin group (P<0.001), but the rates of stroke (2.65% versus 2.33%), hemorrhagic stroke (0.16% versus 0.10%), and myocardial infarction (0.73% versus 0.19%) were not significantly different between the 2 groups. After inverse probability of treatment weighting analysis, the primary composite outcome was significantly reduced in patients with statin therapy (weighted hazard ratio [HR], 0.54 [95% CI, 0.42-0.69]). However, statin treatment did not increase the risk of hemorrhagic stroke (weighted HR, 1.11 [95% CI, 0.10-12.28]). Conclusions Approximately three-quarters of the patients with first-ever ischemic stroke with baseline low-density lipoprotein cholesterol levels <70 mg/dL received in-hospital statin treatment. Statin treatment, compared with no statin treatment, was significantly associated with a reduced risk of the 3-month primary composite outcomes and all-cause death but did not alter the rate of stroke recurrence.

Association between blood pressure and endovascular treatment outcomes differs by baseline perfusion and reperfusion status.

We hypothesized that the association between BP and endovascular treatment (EVT) outcomes would differ by baseline perfusion and recanalization status. We identified 388 ICA or M1 occlusion patients who underwent EVT ≤ 24 h from onset with successful recanalization (TICI ≥ 2b). BP was measured at 5-min intervals from arrival and during the procedure. Systolic BPs (SBP) were summarized as dropmax (the maximal decrease over two consecutive measurements), incmax (the maximal increase), mean, coefficient of variation (cv), and standard deviation. Adequate baseline perfusion was defined as hypoperfusion intensity ratio (HIR) ≤ 0.5; infarct proportion as the volume ratio of final infarcts within the Tmax > 6 s region. In the adequate perfusion group, infarct proportion was closely associated with SBPdropmax (ß ± SE (P-value); 1.22 ± 0.48, (< 0.01)), SBPincmax (1.12 ± 0.33, (< 0.01)), SBPcv (0.61 ± 0.15 (< 0.01)), SBPsd (0.66 ± 0.08 (< 0.01)), and SBPmean (0.71 ± 0.37 (0.053) before recanalization. The associations remained significant only in SBPdropmax, SBPincmax, and SBPmean after recanalization. SBPincmax, SBPcv and SBPsd showed significant associations with modified Rankin Scale score at 3 months in the pre-recanalization period. In the poor perfusion group, none of the SBP indices was associated with any stroke outcomes regardless of recanalization status. BP may show differential associations with stroke outcomes by the recanalization and baseline perfusion status.

Admission LDL-cholesterol, statin pretreatment and early outcomes in acute ischemic stroke.

BACKGROUND: Lipid paradox of low LDL-C may cause physicians to be reluctant to use statins in acute ischemic stroke (AIS) patients with low LDL-C levels at admission. OBJECTIVE: This study investigated the association between LDL-C levels and early vascular outcomes and assessed the potential interaction effect between LDL-C and statin pretreatment on early outcomes. PATIENTS AND METHODS: This was a study of a prospective, multicenter, registry of AIS patients with admission LDL-C. The subjects were divided into 3 groups according to LDL-C levels: low LDL-C (≤100 mg/dL); intermediate LDL-C (>100, <130 mg/dL); and high LDL-C (≥130 mg/dL). The primary early vascular outcome was a composite of stroke (ischemic or hemorrhagic), myocardial infarction and all-cause mortality within 3 months. The associations of LDL-C levels as a continuous variable and the risks of primary outcome using Cox proportional hazards models with restricted cubic splines were explored. RESULTS: A total of 32,505 patients (age, 69 ± 12; male, 58.6%) were analyzed. The 3 groups showed significant differences in the 3-month primary outcome, with highest events in the low LDL-C group; after adjustment, no significant associations with the 3-month primary outcome remained. U-shaped nonlinear relationships of LDL-C levels with the 3-month primary outcome were observed (Pnon-linearity<0.001), with substantial relationships in the no pretreatment subgroup. CONCLUSIONS: The relationships between admission LDL-C levels and early outcomes are complex but appear to be paradoxical in patients with low LDL-C and no statin pretreatment. The results suggest that statin pretreatment might offset the paradoxical response of low LDL-C on early vascular outcomes. Further study would be warranted.

Comparison of Hospital Performance in Acute Ischemic Stroke Based on Mortality and Functional Outcome in South Korea.

BACKGROUND: Recent evidence suggests a correlation between modified Rankin Scale-based measures, an outcome measure commonly used in acute stroke trials, and mortality-based measures used by health agencies in the evaluation of hospital performance. We aimed to examine whether the 2 types of measures are interchangeable in relation to evaluation of hospital performance in acute ischemic stroke. METHODS: Five outcome measures, unfavorable functional outcome (3-month modified Rankin Scale score ≥2), death or dependency (3-month modified Rankin Scale score ≥3), 1-month mortality, 3-month mortality, and 1-year mortality, were collected for 8292 individuals who were hospitalized for acute ischemic stroke between January 2014 and May 2015 in 14 hospitals participating in the Clinical Research Collaboration for Stroke in Korea - National Institute of Health registry. Hierarchical regression models were used to calculate per-hospital risk-adjusted outcome rates for each measure. Hospitals were ranked and grouped based on the risk-adjusted outcome rates, and the correlations between the modified Rankin Scale-based and mortality-based ranking and their intermeasure reliability in categorizing hospital performance were analyzed. RESULTS: The comparison between the ranking based on the unfavorable functional outcome and that based on 1-year mortality resulted in a Spearman correlation coefficient of -0.29 and Kendall rank coefficient of -0.23, and the comparison of grouping based on these 2 types of ranks resulted in a weighted kappa of 0.123 for the grouping in the top 33%/middle 33%/bottom 33% and 0.25 for the grouping in the top 20%/middle 60%/bottom 20%, respectively. No significant correlation or similarity in grouping capacities were found between the rankings based on the functional outcome measures and those based on the mortality measures. CONCLUSIONS: This study shows that regardless of clinical correlation at an individual patient level, functional outcome-based measures and mortality-based measures are not interchangeable in the evaluation of hospital performance in acute ischemic stroke.

Body mass index and waist circumference as predictors of recurrent vascular events after a recent ischemic stroke.

OBJECTIVES: Although elevated body mass index (BMI) is a risk factor for stroke, it appears to protect against recurrent vascular events. We tried to evaluate BMI and waist circumference (WC) as predictors of recurrent stroke and vascular events in a cohort of stroke survivors who were followed for 12 months. MATERIALS AND METHODS: We analyzed the stroke registry database of 6 hospitals and recruited patients with a first-ever stroke who were admitted from January 2011 to November 2019 and had their BMI and WC measured. Cox proportional hazards models were used to compare risks of recurrent stroke and major vascular events (a composite of stroke, myocardial infarction, or vascular death) between different BMI and WC quintiles. Reference categories were patients in the lowest quintiles. RESULTS: A total of 14 781 patients were analyzed. Patients in the second quintile of BMI had the lowest risk of recurrent stroke (adjusted hazard ratio (HR) 0.72; 95% confidence interval (CI) 0.58-0.91); patients in the highest quintile had the lowest risk or a major vascular event (adjusted HR 0.71; 95% CI 0.58-0.86). Patients in the fourth quintile of WC had the lowest risk of recurrent stroke (adjusted HR 0.73; 95% CI 0.59-0.91) and a major vascular event (adjusted HR 0.72; 95 % CI 0.60-0.86). CONCLUSIONS: Our results show favorable effects of excess body weight and intra-abdominal fat on avoidance of vascular events after stroke and a favorable effect of intra-abdominal fat on avoidance of recurrent stroke.

Densely Convolutional Spatial Attention Network for nuclei segmentation of histological images for computational pathology.

Introduction: Automatic nuclear segmentation in digital microscopic tissue images can aid pathologists to extract high-quality features for nuclear morphometrics and other analyses. However, image segmentation is a challenging task in medical image processing and analysis. This study aimed to develop a deep learning-based method for nuclei segmentation of histological images for computational pathology. Methods: The original U-Net model sometime has a caveat in exploring significant features. Herein, we present the Densely Convolutional Spatial Attention Network (DCSA-Net) model based on U-Net to perform the segmentation task. Furthermore, the developed model was tested on external multi-tissue dataset - MoNuSeg. To develop deep learning algorithms for well-segmenting nuclei, a large quantity of data are mandatory, which is expensive and less feasible. We collected hematoxylin and eosin-stained image data sets from two hospitals to train the model with a variety of nuclear appearances. Because of the limited number of annotated pathology images, we introduced a small publicly accessible data set of prostate cancer (PCa) with more than 16,000 labeled nuclei. Nevertheless, to construct our proposed model, we developed the DCSA module, an attention mechanism for capturing useful information from raw images. We also used several other artificial intelligence-based segmentation methods and tools to compare their results to our proposed technique. Results: To prioritize the performance of nuclei segmentation, we evaluated the model's outputs based on the Accuracy, Dice coefficient (DC), and Jaccard coefficient (JC) scores. The proposed technique outperformed the other methods and achieved superior nuclei segmentation with accuracy, DC, and JC of 96.4% (95% confidence interval [CI]: 96.2 - 96.6), 81.8 (95% CI: 80.8 - 83.0), and 69.3 (95% CI: 68.2 - 70.0), respectively, on the internal test data set. Conclusion: Our proposed method demonstrates superior performance in segmenting cell nuclei of histological images from internal and external datasets, and outperforms many standard segmentation algorithms used for comparative analysis.

Biological Mechanism of Sex Difference in Stroke Manifestation and Outcomes.

BACKGROUND AND OBJECTIVES: Female patients tend to have greater disability and worse long-term outcomes after stroke than male patients. To date, the biological basis of sex difference in ischemic stroke remains unclear. We aimed to (1) assess sex differences in clinical manifestation and outcomes of acute ischemic stroke and (2) investigate whether the sex disparity is due to different infarct locations or different impacts of infarct in the same location. METHODS: This MRI-based multicenter study included 6,464 consecutive patients with acute ischemic stroke (<7 days) from 11 centers in South Korea (May 2011-January 2013). Multivariable statistical and brain mapping methods were used to analyze clinical and imaging data collected prospectively: admission NIH Stroke Scale (NIHSS) score, early neurologic deterioration (END) within 3 weeks, modified Rankin Scale (mRS) score at 3 months, and culprit cerebrovascular lesion (symptomatic large artery steno-occlusion and cerebral infarction) locations. RESULTS: The mean (SD) age was 67.5 (12.6) years, and 2,641 (40.9%) were female patients. Percentage infarct volumes on diffusion-weighted MRI did not differ between female patients and male patients (median 0.14% vs 0.14%, p = 0.35). However, female patients showed higher stroke severity (NIHSS score, median 4 vs 3, p < 0.001) and had more frequent END (adjusted difference 3.5%; p = 0.002) than male patients. Female patients had more frequent striatocapsular lesions (43.6% vs 39.8%, p = 0.001) and less frequent cerebrocortical (48.2% vs. 50.7% in patients older than 52 years, p = 0.06) and cerebellar (9.1% vs. 11.1%, p = 0.009) lesions than male patients, which aligned with angiographic findings: female patients had more prevalent symptomatic steno-occlusion of the middle cerebral artery (MCA) (31.1% vs 25.3%; p < 0.001) compared with male patients, who had more frequent symptomatic steno-occlusion of the extracranial internal carotid artery (14.2% vs 9.3%; p < 0.001) and vertebral artery (6.5% vs 4.7%; p = 0.001). Cortical infarcts in female patients, specifically left-sided parieto-occipital regions, were associated with higher NIHSS scores than expected for similar infarct volumes in male patients. Consequently, female patients had a higher likelihood of unfavorable functional outcome (mRS score >2) than male patients (adjusted absolute difference 4.5%; 95% CI 2.0-7.0; p < 0.001). DISCUSSION: Female patients have more frequent MCA disease and striatocapsular motor pathway involvement with acute ischemic stroke, along with left parieto-occipital cortical infarcts showing greater severity for equivalent infarct volumes than in male patients. This leads to more severe initial neurologic symptoms, higher susceptibility to neurologic worsening, and less 3-month functional independence, when compared with male patients.

Admission hyperglycemia, stroke subtypes, outcomes in acute ischemic stroke.

AIMS: Whether admission hyperglycemia is differentially associated with early vascular outcomes in acute ischemic stroke (AIS) depending on stroke subtype has been incompletely delineated. METHODS: In a multicenter, prospective stroke registry, patients with AIS were categorized based on admission glucose levels into normoglycemia, moderate hyperglycemia, and severe hyperglycemia (<140mg/dl, 140-179mg/dl, and ≥180mg/dl, respectively) groups. Multivariate analysis assessed the interaction between the hyperglycemia and ischemic stroke subtypes of large artery atherothrombosis (LAA), cardioembolism (CE), and small vessel occlusion (SVO) and early vascular outcomes (3-month stroke, all-cause mortality, and composite of stroke, MI, and all-cause mortality). RESULTS: Among the 32,772 patients (age:69.0±12.6yrs, male:58.4%) meeting eligibility criteria, 61.9% were in the normoglycemia group, 19.5% were in the moderate hyperglycemia group, and 18.7% were in the severe hyperglycemia group. Substantial interactions between hyperglycemia groups and stroke subtypes were observed for 3-month stroke (Pinteraction = 0.003) and composite of stroke, MI, and all-cause mortality (Pinteraction = 0.001), with differential recurrence strongest among CE, intermediate among LAA, and least among SVO. CONCLUSIONS: Hyperglycemia was differently associated with the risk of 3-month stroke by ischemic stroke subtype. The associations of hyperglycemia with 3-month stroke were greatest in CE subtype but not in SVO subtype. These results suggest that the effect of glucose-lowering treatment after AIS may differ according to stroke subtype.