Stroke detection and classification from computed tomography (CT) remains a critical and challenging task in medical imaging due to the complexity of lesion patterns, noise variations and unbalanced datasets. In this study, we propose a novel hybrid deep learning model, StrokeFuse-AttnNet, which integrates both global (ResNet50) and local (DenseNet121) convolutional feature extractors with a self-attention mechanism to improve spatial focus and semantic interpretability. A hierarchical feature fusion strategy concatenates multi-scale features, which are then processed by a self-attention module to highlight key stroke regions and reduce irrelevant activations. We use data augmentation and SMOTE on training samples to address imbalance and improve generalization. The proposed model was evaluated on both publicly and privately available brain CT datasets. StrokeFuse-AttnNet achieved an accuracy of 98.27

Accurate prediction of stroke risk at an early stage is essential for timely intervention and prevention, especially given the serious health consequences and economic burden that strokes can cause. In this study, we proposed a class-balanced and data-augmented (CBDA-ResNet50) deep learning model to improve the prediction accuracy of the well-known ResNet50 architecture for stroke risk. Our approach uses advanced techniques such as class balancing and data augmentation to address common challenges in medical imaging datasets, such as class imbalance and limited training examples. In most cases, these problems lead to biased or less reliable predictions. To address these issues, the proposed model assures that the predictions are still accurate even when some stroke risk factors are absent in the data. The performance of CBDA-ResNet50 improves by using the Adam optimizer and the ReduceLROnPlateau scheduler to adjust the learning rate. The application of weighted cross entropy removes the imbalance between classes and significantly improves the results. It achieves an accuracy of 97.87% and a balanced accuracy of 98.27%, better than many of the previous best models. This shows that we can make more reliable predictions by combining modern deep-learning models with advanced data-processing techniques. CBDA-ResNet50 has the potential to be a model for early stroke prevention, aiming to improve patient outcomes and reduce healthcare costs.

Rehabilitation after a stroke is vital for regaining functional abilities. However, a shortage of rehabilitation professionals leads to many patients with severe disabilities. Traditional rehabilitation methods can be time-consuming and hard to measure for progress. This study introduces an innovative machine learning (ML) approach for lower limb rehabilitation in stroke patients. The proposed methodology integrates two models: a fuzzy logic rule-based system and a K-Nearest Neighbor(K-NN) machine learning model. The rule-based model utilizes the Fugl-Meyer Assessment to evaluate lower limb angles during exercises using a camera without human intervention. The hybrid fuzzy logic-based ML model continuously tracks the desired angle, counts exercise repetitions, and provides real-time feedback on patient progress. Furthermore, it measures the Range of Motion (ROM) for each repetition, presenting a graphical visualization of ROMs for ten repetitions simultaneously. The model facilitates real-time evaluation of rehabilitation progress by clinicians, with the lowest observed error rate of $$0.34^\circ$$ of angle measurement. The K-NN model assesses rehabilitation exercise accuracy levels, presenting results graphically, with machine learning accuracy rates of $$97\%$$ , $$92\%$$ , and $$91\%$$ for hip flexion, hip external rotation, and knee extension rehabilitation exercises. Model training utilized data from 30 experienced physical therapists at King Chulalongkorn Memorial Hospital, Bangkok, Thailand, garnering positive evaluations from rehabilitation doctors. The proposed ML-based models offer real-time and prerecorded video capabilities, enabling telerehabilitation applications. This research highlights the potential of ML-based methodologies in stroke rehabilitation to enhance accuracy, efficiency, and patient outcomes.

Introduction: Clopidogrel and aspirin were proved to have benefit in symptomatic intracranial stenosis. CYP2C19 polymorphism (CYP2C19*1, CYP2C19*2, CYP2C19*3, and CYP2C19*17 alleles) affects efficacy of clopidogrel. Epidemiologic study of CYP2C19 polymorphism has been conducted in Thai population. There was no data showed the frequency of allelic variants of CYP2C19 in Thai symptomatic intracranial stenosis patients. The aim of this study was to determine the prevalence of CYP2C19 polymorphism in symptomatic intracranial stenosis patients. Methods: The study group included 100 Thai symptomatic intracranial stenosis patients. Genotyping of CYP2C19 alleles (CYP2C19*1, CYP2C19*2, CYP2C19*3, and CYP2C19*17 alleles) was carried out by real-time polymerase chain reaction (rt-PCR) technique. Results: The allele frequency of CYP2C19*1, CYP2C19*2, CYP2C19*3, and CYP2C19*17 were 70.5%, 26%, 2.5%, and 1%, respectively. The result showed that 53% of symptomatic intracranial stenosis patients are normal metabolizers, while intermediate and poor metabolizer were 36 and 10 percent, respectively. Conclusion: Almost one-half of Thai symptomatic intracranial stenosis patients were intermediate or poor metabolizers. Usage of combination of aspirin and clopidogrel might not be effective in this group of patients.

Introduction: The stroke scale for the mid-level personnel (SML) was designed for emergency medical services personnel to predict acute ischemic stroke due to large vessel occlusion (LVO) in both prehospital and in-hospital settings. This study aimed to validate and determine the appropriate cut point of the SML score in this regard. Methods: This single-centered, prospective validation study to assess a novel LVO triage tool was performed in a tertiary care hospital in Bangkok. Patients presenting within 24 hours of onset of acute stroke were included in the study. The scale is designed for mid-level providers and emergency medical services (EMS) personnel including paramedics, emergency medical technicians (EMTs) and emergency department (ED) nurses. LVO was confirmed by brain and neck computed tomography angiography (CTA). Area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratios (LRs), and correctly classified instances (CCI) were calculated. Youden’s index was used to determine an appropriate cut point of the SML score for LVO prediction. Results: 200 cases with the median age of 64.0 (56.5-73.0) years were included (53.5% female). 83 (41.5%) cases were affiliated to the LVO and 117 (58.5%) to the non-LVO group. The median SML scores for non-LVO and LVO stroke patients were 3 (2 - 3) and 6 (5 - 7), respectively (p < 0.001). The most common presentations in both groups were facial palsy, arm weakness and speech impairment or dysarthria. There was significantly higher prevalence of neglect (8 (6.8%) vs. 5 (4.3%); p < 0.001) and eye deviation (39 (47%) vs. 29 (35%); p < 0.001) in the LVO stroke group than in the non-LVO group. LVO patients scored higher in all categories when compared to non-LVO cases. SML scores of 4 and 5 had the highest Youden's index of 0.82 and 0.67, respectively. SML score of 4 yielded the highest correctly classified instances (CCI) of 90% with sensitivity and specificity of 96.4% (95% confidence interval (CI): 89.9-99.3%) and 85.3% (95% CI: 77.6-91.2), respectively. SML score of 4 also achieved the lowest negative LR of 0.04 and an odds ratio of 157 (95% CI: 46.7-521). The AUC of SML in cutoff point of 4 was 0.901 (95%CI: 0.853 - 0.949). Conclusions: SML score may be helpful for mid-level medical providers and also EMS personnel in detecting LVOs since prehospital phase. According to the results, we recommend a cut point SML score ≥ 4 for enhanced sensitivity and NPV.

BACKGROUND:Stroke related to infections represents a less common but significant cause, particularly in low- and middle-income countries. This review examines the pathophysiology of stroke from infections, involving both direct and indirect mechanisms. SUMMARY:Bacterial infections such as tuberculous meningitis and infective endocarditis can directly cause strokes through local inflammation, arteritis, and septic embolism. Viral infections like Varicella zoster virus and HIV increase stroke risk through chronic immune activation, vasculopathy, and endothelial dysfunction. Parasitic infections, particularly malaria and neurocysticercosis, can cause strokes via vascular occlusion and inflammatory responses. Fungal infections like aspergillosis and mucormycosis can lead to strokes through vasculitis and direct invasion of the CNS. KEY MESSAGES:Understanding the mechanisms by which various infectious agents contribute to stroke pathogenesis is essential for developing targeted therapeutic strategies and improving patient outcomes. Further research is needed to establish effective prevention and treatment for infection-related strokes, especially in low- and middle-income countries.

BackgroundThe Royal College of Physicians of Thailand (RCPT) published a Clinical Practice Guideline on Pharmacologic Therapy of Dyslipidemia for Atherosclerotic Cardiovascular Disease (ASCVD) Prevention in 2016. The availability of newer classes of medications for dyslipidemia, supported by extensive clinical research findings, indicates a significant need for the updating of the existing clinical practice guideline.ObjectivesTo serve as guidelines on the management of dyslipidemia for Thai adults.MethodsThe RCPT Dyslipidemia Guidelines Committee was established with representatives from selected professional societies to revise the 2016 Guideline by critically reviewing the latest evidence. Meetings were conducted from August to December 2023, culminating in a public hearing that engaged various stakeholders in January 2024. The final Thai version received approval in April 2024, while the English translation was completed in October 2024.ResultsLifestyle modifications and statins remain the cornerstone of therapy for dyslipidemia in adults across various clinical settings. Emerging evidence regarding newer classes of lipid-lowering medications indicates that these treatments are effective in lowering LDL-cholesterol levels and reducing atherosclerotic cardiovascular events. This suggests that they may serve as an add-on therapy for individuals who cannot achieve target levels or who are at high risk for future cardiovascular events. The Thai CV Risk Score is recommended due to its specificity for the Thai population.ConclusionsThe 2024 updated clinical practice guidelines establish a framework, provide recommendations, and serve as a comprehensive resource for the contemporary management of dyslipidemia in adults, with the goal of preventing ASCVD in Thailand.

Cerebrovascular diseases such as stroke are among the most common causes of death and disability worldwide and are preventable and treatable. Early detection of strokes and their rapid intervention play an important role in reducing the burden of disease and improving clinical outcomes. In recent years, machine learning methods have attracted a lot of attention as they can be used to detect strokes. The aim of this study is to identify reliable methods, algorithms, and features that help medical professionals make informed decisions about stroke treatment and prevention. To achieve this goal, we have developed an early stroke detection system based on CT images of the brain coupled with a genetic algorithm and a bidirectional long short-term Memory (BiLSTM) to detect strokes at a very early stage. For image classification, a genetic approach based on neural networks is used to select the most relevant features for classification. The BiLSTM model is then fed with these features. Cross-validation was used to evaluate the accuracy of the diagnostic system, precision, recall, F1 score, ROC (Receiver Operating Characteristic Curve), and AUC (Area Under The Curve). All of these metrics were used to determine the system’s overall effectiveness. The proposed diagnostic system achieved an accuracy of 96.5%. We also compared the performance of the proposed model with Logistic Regression, Decision Trees, Random Forests, Naive Bayes, and Support Vector Machines. With the proposed diagnosis system, physicians can make an informed decision about stroke.

INTRODUCTION:Moyamoya disease (MMD) and non-MMD intracranial cerebral artery stenosis (ICAS) have been linked to the RNF213 rs112735431 gene in Korean and Japanese populations. This cross-sectional study investigates the prevalence of the RNF213 rs112735431 gene in non-cardioembolic ischemic stroke (NCIS) among Thai patients. METHODS:A cross-sectional investigation was conducted on patients aged 18 years or older admitted to King Chulalongkorn Memorial Hospital between June 2015 and March 2016 with acute NCIS. ICAS and extracranial carotid artery stenosis (ECAS) were assessed through computer tomography angiography or magnetic resonance angiography. Blood samples were collected, and Sanger sequencing was performed. RESULTS:Among 234 acute NCIS cases, 113 exhibited ICAS, 12 had ECAS, 20 had both, and 89 had neither. The RNF213 rs112735431 gene variant was detected in 2 patients, both heterozygous A/G. The frequency of the RNF213 rs112735431 variant was 0.9% (2/234; 95% CI: 0-2.1%) in acute NCIS patients and 1.8% (2/113; 95% CI: 0-4.2%) in ICAS. All individuals with the RNF213 variant were males with hypertension, diabetes mellitus, dyslipidemia, and ICAS, without a family history of ischemic stroke. CONCLUSION:This study reveals that the RNF213 rs112735431 gene variant is uncommon among Thai NCIS patients, suggesting a discrepancy in the prevalence of this genetic variation between Thai and other Eastern Asian populations.

Abstract Background Ischemic stroke (IS) is one of the leading causes of death among non-communicable diseases in Thailand. Patients who have survived an IS are at an increased risk of developing recurrent IS, which can result in worse outcomes and post-stroke complications. Objectives The study aimed to investigate the incidence of recurrent IS among patients with first-ever IS during a one-year follow-up period and to determine its associated risk factors. Methods Adult patients (aged ≥ 18 years) who were hospitalized at the Stroke Center, King Chulalongkorn Memorial Hospital (KCMH) in Bangkok, Thailand, due to first-ever IS between January and December 2019 and had at least one follow-up visit during the one-year follow-up period were included in this retrospective cohort study. IS diagnosis was confirmed by neurologists and imaging. The log-rank test was used to determine the event-free survival probabilities of recurrent IS in each risk factor. Results Of 418 patients hospitalized due to first-ever IS in 2019, 366 (87.6%) were included in the analysis. During a total of 327.2 person-years of follow-up, 25 (6.8%) patients developed recurrent IS, accounting for an incidence rate of 7.7 per 100 person-year (95% confidence interval [CI] 5.2–11.3). The median (interquartile range) time of recurrence was 35 (16–73) days. None of the 47 patients with atrial fibrillation developed recurrent IS. The highest incidence rate of recurrent IS occurred within 1 month after the first episode (34 per 100 person-years) compared to other follow-up periods. Patients with small vessel occlusion and large-artery atherosclerosis (LAA) constituted the majority of patients in the recurrent IS episode (48% and 40%, respectively), with LAA exhibiting a higher recurrence rate (13.5%). Additionally, smoking status was found to be associated with an increased risk of recurrence. Conclusion The incidence rate of the recurrence was moderate in our tertiary care setting, with a decreasing trend over time after the first episode. The various subtypes of IS and smoking status can lead to differences in event-free survival probabilities.

Stroke has a negative impact on people’s lives and is one of the leading causes of death and disability worldwide. Early detection of symptoms can significantly help predict stroke and promote a healthy lifestyle. Researchers have developed several methods to predict strokes using machine learning (ML) techniques. However, the proposed systems have suffered from the following two main problems. The first problem is that the machine learning models are biased due to the uneven distribution of classes in the dataset. Recent research has not adequately addressed this problem, and no preventive measures have been taken. Synthetic Minority Oversampling (SMOTE) has been used to remove bias and balance the training of the proposed ML model. The second problem is to solve the problem of lower classification accuracy of machine learning models. We proposed a learning system that combines an autoencoder with a linear discriminant analysis (LDA) model to increase the accuracy of the proposed ML model for stroke prediction. Relevant features are extracted from the feature space using the autoencoder, and the extracted subset is then fed into the LDA model for stroke classification. The hyperparameters of the LDA model are found using a grid search strategy. However, the conventional accuracy metric does not truly reflect the performance of ML models. Therefore, we employed several evaluation metrics to validate the efficiency of the proposed model. Consequently, we evaluated the proposed model’s accuracy, sensitivity, specificity, area under the curve (AUC), and receiver operator characteristic (ROC). The experimental results show that the proposed model achieves a sensitivity and specificity of 98.51% and 97.56%, respectively, with an accuracy of 99.24% and a balanced accuracy of 98.00%.