Our study aims to develop an effective integrated machine learning (ML) scheme to predict vascular events and bleeding in patients with nonvalvular atrial fibrillation taking dabigatran and identify important risk factors. This study is a post-hoc analysis from the Randomized Evaluation of Long-Term Anticoagulant Therapy trial database. One traditional prediction method, logistic regression (LGR), and four ML techniques—naive Bayes, random forest (RF), classification and regression tree, and extreme gradient boosting (XGBoost)—were combined to construct our scheme. Area under the receiver operating characteristic curve (AUC) of RF (0.780) and XGBoost (0.717) was higher than that of LGR (0.674) in predicting vascular events. In predicting bleeding, AUC of RF (0.684) and XGBoost (0.618) showed higher values than those generated by LGR (0.605). Our integrated ML feature selection scheme based on the two convincing prediction techniques identified age, history of congestive heart failure and myocardial infarction, smoking, kidney function, and body mass index as major variables of vascular events; age, kidney function, smoking, bleeding history, concomitant use of specific drugs, and dabigatran dosage as major variables of bleeding. ML is an effective data analysis algorithm for solving complex medical data. Our results may provide preliminary direction for precision medicine.

Nonvalvular atrial fibrillation (NVAF) and carotid stenosis are important risk factors for stroke. Carotid angioplasty and stent placement (CAS) is recommended for patients with symptomatic high-grade carotid stenosis. The optimal medical management for patients with NVAF after CAS remains unclear. We aimed to clarify this issue using real-world data from the Taiwanese National Health Insurance Research Database (NHIRD). In total, 2116 consecutive NVAF patients who received CAS between January 1, 2010, and December 31, 2016, from NHIRD were divided into groups based on post-procedure medication as follows: only antiplatelet agent (OAP, n = 587); only anticoagulation agent (OAC, n = 477); dual antiplatelet agents (DAP, n = 49); and a combination of antiplatelet and anticoagulation agents (CAPAC, n = 304). Mortality, vascular events, and major bleeding episodes were compared after matching with the Charlson comorbidity index and CHA2DS2-VASc score. The CAPAC and the OAC groups had lower mortality rates than the OAP group (P = 0.0219), with no statistical differences in major bleeding, ischemic stroke, or vascular events. Conclusively, OAC therapy after CAS appears suitable for NVAF patients. CAPAC therapy might be considered as initial therapy or when there is concern about vascular events.

Radiation vasculopathy is one of the rare causes of ischemic stroke. Carotid stenosis with large volume infarction may occur years after radiation therapy for head or neck cancer. We report a case of a patient with bilateral internal carotid artery occlusion presenting with left middle cerebral artery infarct 10 years after receiving treatment for tongue cancer. A literature review and discussion of treatment for such patients are presented.

The new generation of nonvitamin K antagonists are broadly applied for stroke prevention due to their notable efficacy and safety. Our study aimed to develop a suggestive utilization of dabigatran through an integrated machine learning (ML) decision-tree model. Participants taking different doses of dabigatran in the Randomized Evaluation of Long-Term Anticoagulant Therapy trial were included in our analysis and defined as the 110 mg and 150 mg groups. The proposed scheme integrated ML methods, namely naive Bayes, random forest (RF), classification and regression tree (CART), and extreme gradient boosting (XGBoost), which were used to identify the essential variables for predicting vascular events in the 110 mg group and bleeding in the 150 mg group. RF (0.764 for 110 mg; 0.747 for 150 mg) and XGBoost (0.708 for 110 mg; 0.761 for 150 mg) had better area under the receiver operating characteristic curve (AUC) values than logistic regression (benchmark model; 0.683 for 110 mg; 0.739 for 150 mg). We then selected the top ten important variables as internal nodes of the CART decision tree. The two best CART models with ten important variables output tree-shaped rules for predicting vascular events in the 110 mg group and bleeding in the 150 mg group. Our model can be used to provide more visualized and interpretable suggestive rules to clinicians managing NVAF patients who are taking dabigatran.

Background and Purpose: Intracranial atherosclerosis is a significant contributor to ischemic stroke, often through plaques without luminal stenosis. Current imaging techniques struggle to detect vessel wall pathology effectively. This study evaluates short-TE susceptibility-weighted imaging magnitude images (STE-MI) for identifying cerebral microbleeds and basilar artery (BA) wall changes in acute isolated pontine infarction, with reduced artifacts compared to SWAN. Methods: In this post hoc analysis, 2,475 brain MR exams conducted between March 2022 and March 2025 using a 3.0 T scanner compared STE-MI (TE=7.2 ms) with SWAN for microbleed detection and BA artifact reduction. Among 131 patients with acute posterior circulation stroke (January 2023–April 2025), vessel wall pathology was assessed in 92 patients with vertebrobasilar stenosis (VBASO) and 39 with non-stenotic non-cardioembolic (NSNCE) stroke using STE-MI. Interobserver agreement and comparative statistics were analyzed. Results: STE-MI detected 86% of microbleeds identified by SWAN in 33 subjects (Cohen’s kappa=0.828 for interobserver agreement). Blooming artifacts around the BA were reduced in 96.9% of 2,475 cases with STE-MI compared to 5.5% with SWAN. In the NSNCE group, 10 of 16 patients with isolated pontine infarction exhibited extraluminal BA pathology on STE-MI (mixed hypo-/intermediate intensity in 6, hypointensity in 4), correlating with diffusion-weighted imaging (DWI) lesions. Conclusions: STE-MI effectively minimizes artifacts and detects BA wall pathology in acute isolated pontine infarction, particularly in NSNCE stroke. This approach may enhance the identification of culprit lesions, potentially improving risk stratification and guiding targeted therapies.