Relation of Pre-Stroke Aspirin Use With Cerebral Infarct Volume and Functional Outcomes

Deep learning-based automatic segmentation of cerebral infarcts on diffusion MRI

We explored effects of (1) training with various sample sizes of multi-site vs. single-site training data, (2) cross-site domain adaptation, and (3) data sources and features on the performance of algorithms segmenting cerebral infarcts on Magnetic Resonance Imaging (MRI). We used 10,820 annotated diffusion-weighted images (DWIs) from 10 university hospitals. Algorithms based on 3D U-net were trained using progressively larger subsamples (ranging from 217 to 8661), while internal testing employed a distinct set of 2159 DWIs. External validation was conducted using three unrelated datasets (n = 2777, 50, and 250). For domain adaptation, we utilized 50 to 1000 subsamples from the 2777-image external target dataset. As the size of the multi-site training data increased from 217 to 1732, the Dice similarity coefficient (DSC) and average Hausdorff distance (AHD) improved from 0.58 to 0.65 and from 16.1 to 3.75 mm, respectively. Further increases in sample size to 4330 and 8661 led to marginal gains in DSC (to 0.68 and 0.70, respectively) and in AHD (to 2.92 and 1.73). Similar outcomes were observed in external testing. Notably, performance was relatively poor for segmenting brainstem or hyperacute (< 3 h) infarcts. Domain adaptation, even with a small subsample (n = 50) of external data, conditioned the algorithm trained with 217 images to perform comparably to an algorithm trained with 8661 images. In conclusion, the use of multi-site data (approximately 2000 DWIs) and domain adaptation significantly enhances the performance and generalizability of deep learning algorithms for infarct segmentation.

Automated Segmentation of MRI White Matter Hyperintensities in 8421 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; this issue 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 data sets involving 8421 patients with acute ischemic stroke.

MATERIALS AND METHODS

We included 8421 patients with stroke from 9 centers in Korea. 2D UNet and squeeze-and-excitation (SE)-UNet models were trained using 2408 FLAIR MRIs from 3 hospitals and validated using 6013 FLAIR MRIs from 6 hospitals. WMH segmentation performance was assessed by calculating the Dice similarity coefficient (DSC), the correlation coefficient, and the concordance correlation coefficient compared with a human-segmented criterion 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 data set, the mean age was 67.4 (SD, 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 (95% CI, 0.666-0.685; P < .001) in the internal testing and 0.722 (95% CI, 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 data sets. We observed a significant correlation between WMH volumes that were segmented automatically and manually for the UNet (r = 0.917, P < .001), and it was even stronger for the SE-UNet (r = 0.933, P < .001). The SE-UNet also attained a high concordance correlation coefficient (ranging from 0.841 to 0.956) in the external test data sets. In addition, the uncertainty indices in most patients (86%) in the external data sets were <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 data sets. In addition, we showed that the uncertainty index can be used to identify cases in which automatic WMH segmentation is less accurate and requires human review.

Effect of infarct location and volume on cognitive dysfunction in elderly patients with acute insular cerebral infarction

BACKGROUND

The aging of the population has become increasingly obvious in recent years, and the incidence of cerebral infarction has shown an increasing trend annually, with high death and disability rates.

AIM

To analyze the effects of infarct location and volume on cognitive dysfunction in elderly patients with acute insular cerebral infarction.

METHODS

Between January 2020 and December 2023, we treated 98 cases of elderly acute insula, patients with cerebral infarction in the cerebral infarction acute phase (3-4 weeks) and for the course of 6 months in Montreal Cognitive Assessment Scale (MoCA) for screening of cognition. Notably, 58 and 40 patients were placed in the cognitive impairment group and without-cognitive impairment group, respectively. In patients with cerebral infarction, magnetic resonance imaging was used to screen and clearly analyze the MoCA scores of two groups of patients with different infarctions, the relationship between the parts of the infarction volume, and analysis of acute insula cognitive disorder in elderly patients with cerebral infarction and the relationship between the two.

RESULTS

The number of patients with cognitive impairment in the basal ganglia and thalamus was significantly higher than that without cognitive impairment (P < 0.05). The total infarct volume in the cognitive impairment group was higher than that in the non-cognitive impairment group, and the difference was statistically significant (P < 0.05). The infarct volumes at different sites in the cognitive impairment group was higher than in the non-cognitive impairment group (P < 0.05). In the cognitive impairment group, the infarct volumes in the basal ganglia, thalamus, and mixed lesions were negatively correlated with the total MoCA score, with correlation coefficients of -0.67, -0.73, and -0.77, respectively.

CONCLUSION

In elderly patients with acute insular infarction, infarction in the basal ganglia, thalamus, and mixed lesions were more likely to lead to cognitive dysfunction than in other areas, and patients with large infarct volumes were more likely to develop cognitive dysfunction. The infarct volume in the basal ganglia, thalamus, and mixed lesions was significantly negatively correlated with the MoCA score.

The impact of pre-stroke aspirin exposure on radiographic appearance and disability outcomes: A post-hoc analysis of the SPS3 trial: Aspirin Use and Small Subcortical Stroke

OBJECTIVES

The effect of pre-stroke use of aspirin on small subcortical infarct dimensions or outcomes is not well described. We aimed to bridge this knowledge gap amongst a well-described and heterogeneous patient population.

MATERIALS AND METHODS

We performed a post-hoc analysis of the Secondary Prevention of Small Subcortical Stroke (SPS3) trial. The primary exposure was aspirin use ≤7 days of index stroke. The primary outcomes were infarct dimensions. Functional outcomes by modified Rankin Scale (mRS) was a secondary outcome. Age restricted (≥55 years) subgroup analyses were performed as a sensitivity analysis. Descriptive statistical and regression modeling were performed for data analysis.

RESULTS

We included 1423 participants of which 453(31.8 %) used aspirin. Aspirin use was associated with more cardiovascular risk diagnoses. Maximal infarct diameter did not differ with pre-stroke aspirin use (11.3±4.2 mm versus 11.8±4.1 mm, p=0.057) however infarct area was smaller with exposure (126.4±90.0 mm2 versus 137.4±97.0 mm2, p=0.037) regardless of aspirin strength. Participants ≥55 years had smaller infarct diameters (11.1±4.2 mm versus 11.9±4.4 mm, p=0.019) and area (123.4±87.1 mm2 versus 130.6±93.2 mm2, p=0.037) with aspirin use. mRS did not significantly differ in our analyses.

CONCLUSIONS

In this post-hoc analysis of the SPS3 trial, pre-stroke aspirin use was associated with a smaller infarct area regardless of aspirin strength and without impact on functional outcomes. These findings were more pronounced in participants ≥55 years.

REGISTRATION

https://clinicaltrials.gov/study/NCT00059306?term= %22sps3 %22&rank=1.

A Knowledge Map of the Relationship between Diabetes and Stroke: A Bibliometric Analysis Study

INTRODUCTION

The correlation between diabetes and stroke has been studied extensively in epidemiological research. Here, we used bibliometric software to visualize and analyze the literature related to diabetic stroke to provide an overview of the current state of research, hotspots, and future trends in the field.

METHODS

Based on the Web of Science Core Collection (WoSCC) database, we collected studies related to diabetic stroke from 2007 to December 2022. We used CiteSpace (version 6.1.R5), VOSviewer, and Scimago Graphica to create knowledge maps and conduct visual analyses on authors, countries, institutions, cited references, and keywords, and Origin for statistical analysis.

RESULTS

We included a total of 5,171 papers on diabetic stroke from the WoSCC database. Overall, there was a steady increase in the number of publications, with a high number of emerging scientists. The USA was the most productive and influential country, which dominated national collaborations. The most common subject category was "neurology." In total, 12 major clusters were generated from the cited references. Keyword analysis showed that keywords related to poststroke injury and treatment are those with the highest burst intensity and latest burst time.

CONCLUSIONS

Individual disease treatment remains a hot topic, and how to balance acute stroke treatment and glycemic control is currently a difficult clinical problem. At the same time, the mechanism of their interaction and the prevention and treatment of related causative factors remain a hot topic of current and future research.

Ischaemic stroke despite antiplatelet therapy: Causes and outcomes

BACKGROUND

Ischaemic stroke may occur despite antiplatelet therapy (APT). We aimed to investigate frequency, potential causes and outcomes in patients with ischaemic stroke despite APT.

METHODS

In this cohort study, we enrolled patients with imaging-confirmed ischaemic stroke from the Swiss Stroke Registry (01/2014-07/2022). We determined the frequency of prior APT, assessed stroke aetiology (modified TOAST classification) and determined the association of prior APT with unfavourable functional outcome (modified Rankin Scale score 3-6) and recurrent ischaemic stroke at 3 months using regression models.

RESULTS

Among 53,352 patients, 27,484 (51.5%) had no prior antithrombotic treatment, 17,760 (33.3%) were on APT, 7039 (13.2%) on anticoagulation and 1069 (2.0%) were on APT + anticoagulation. In patients with a history of ischaemic stroke/TIA (n = 11,948; 22.4%), 2401 (20.1%) had no prior antithrombotic therapy, 6594 (55.2%) were on APT, 2489 (20.8%) on anticoagulation and 464 (3.9%) on APT + anticoagulation. Amongst patients with ischaemic stroke despite APT, aetiology was large artery atherosclerosis in 19.8% (n = 3416), cardiac embolism in 23.6% (n = 4059), small vessel disease in 11.7% (n = 2011), other causes in 7.4% (n = 1267), more than one cause in 6.3% (n = 1078) and unknown cause in 31.3% (n = 5388). Prior APT was not independently associated with unfavourable outcome (aOR = 1.06; 95% CI: 0.98-1.14; p = 0.135) or death (aOR = 1.10; 95% CI: 0.99-1.21; p = 0.059) at 3-months but with increased odds of recurrent stroke (6.0% vs 4.3%; aOR 1.26; 95% CI: 1.11-1.44; p < 0.001).

CONCLUSIONS

One-third of ischaemic strokes occurred despite APT and 20% of patients with a history of ischaemic stroke had no antithrombotic therapy when having stroke recurrence. Aetiology of breakthrough strokes despite APT is heterogeneous and these patients are at increased risk of recurrent stroke.

Pre-aspirin use has no benefit on the neurological disability and mortality after cardiovascular events: A nation-wide population-based cohort study

To evaluate the effects of aspirin in the primary prevention, we evaluated disability grades and mortality after ischemic/hemorrhagic stroke and myocardial infarction (MI). A retrospective nation-wide propensity score-matched cohort study was performed using the Korean National Health Information Database. From 3,060,639 subjects who were older than 55 and performed national health examinations in 2004 and 2005, we selected the aspirin group (N = 8770) was composed of patients who had received aspirin prior to cardiovascular events. Cox proportional hazards model was used to compare the acquisition times for neurologic disability grades and survival times between the aspirin and control groups. Only in hemorrhagic stroke, the severe neurologic disability risk was higher in the aspirin group (hazard ratio [HR], 1.21; 95% confidence interval [CI], 1.02-1.42). The aspirin group was associated with higher 90-day (HR, 1.33; 95% CI, 1.23-1.44) and long-term mortality risk (HR, 1.06; 95% CI, 1.03-1.10) after pooling 3 events. The old age was a strong risk factor for 90-day mortality in hemorrhagic stroke (50s: reference; 60s: HR 2.21, 95% CI 1.50-3.25; 70s: HR 3.63, 95% CI 2.48-5.30; 80s: HR 6.69, 95% CI 4.54-9.65; >90s: HR 11.28, 95% CI 6.46-19.70). Pre-aspirin use in cardiovascular events has detrimental effects on severe neurological disability in hemorrhagic stroke and short-/long-term mortality in 3 cardiovascular events. The use of aspirin for the primary prevention especially in the elderly should be very cautious because the old age is a strong risk factor for 90-day mortality after hemorrhagic stroke.

David G. Sherman Lecture Award: 15-Year Experience of the Nationwide Multicenter Stroke Registry in Korea

The expected growth of stroke burden in Korea in early 2000s led to the initiation of a government-funded clinical research project with the goal of development and implementation of national stroke guidelines. The CRCS-K (Clinical Research Collaboration for Stroke in Korea) began as a part of this project. For stroke epidemiology and quality of care research, the CRCS-K developed a multicenter, prospective, stroke registry and began collection of data in 2008. Now, about 100 000 cases have been registered at 17 university hospitals or regional stroke centers and about 200 articles have been published based on the registry experience. The analysis of the 10-year secular trends showed overall improvement of stroke care and outcomes and areas for improvement. This large-scale, high-quality dataset provides opportunities to explore and compare treatment disparities using the comparative effectiveness research methods, design and conduct a registry-based randomized clinical trial, connect the registry data with other data sources including the national claims data and neuroimaging or genetic data, and collaborate with other international researchers. An international stroke registry consortium may be a viable future direction.

Cerebral Complications of Snakebite Envenoming: Case Studies

There are an estimated 5.4 million snakebite cases every year. People with snakebite envenoming suffer from severe complications, or even death. Although some review articles cover several topics of snakebite envenoming, a review of the cases regarding cerebral complications, especially rare syndromes, is lacking. Here, we overview 35 cases of snakebite by front-fanged snakes, including Bothrops, Daboia, Cerastes, Deinagkistrodon, Trimeresurus, and Crotalus in the Viperidae family; Bungarus and Naja in the Elapidae family, and Homoroselaps (rare cases) in the Lamprophiidae family. We also review three rare cases of snakebite by rear-fanged snakes, including Oxybelis and Leptodeira in the Colubridae family. In the cases of viper bites, most patients (17/24) were diagnosed with ischemic stroke and intracranial hemorrhage, leading to six deaths. We then discuss the potential underlying molecular mechanisms that cause these complications. In cases of elapid bites, neural, cardiac, and ophthalmic disorders are the main complications. Due to the small amount of venom injection and the inability to deep bite, all the rear-fanged snakebites did not develop any severe complications. To date, antivenom (AV) is the most effective therapy for snakebite envenoming. In the six cases of viper and elapid bites that did not receive AV, three cases (two by viper and one by elapid) resulted in death. This indicates that AV treatment is the key to survival after a venomous snakebite. Lastly, we also discuss several studies of therapeutic agents against snakebite-envenoming-induced complications, which could be potential adjuvants along with AV treatment. This article organizes the diagnosis of hemotoxic and neurotoxic envenoming, which may help ER doctors determine the treatment for unidentified snakebite.