Stroke; a journal of cerebral circulation
-
Background and Purpose- Although several clinical studies suggested the beneficial effects of edaravone in acute ischemic stroke, most were performed under settings that differ from those in the current treatment strategy, which has dramatically changed with progress in reperfusion therapies. This study aimed to evaluate the efficacy of edaravone in patients with acute ischemic stroke treated by emergent endovascular reperfusion therapy. Methods- We conducted a retrospective observational study using a national administrative database. ⋯ Results of the propensity score matching analysis corroborated these results. Conclusions- This retrospective analysis of a Japanese nationwide administrative database suggested that combination therapy with edaravone and endovascular reperfusion therapy could be a promising therapeutic strategy in acute ischemic stroke. Further randomized control trials are warranted.
-
Background and Purpose- Efficacy of endovascular thrombectomy has been demonstrated up to 24 hours after stroke onset in patients selected with perfusion imaging. We hypothesized that a persistent favorable perfusion profile exists in some patients beyond 24 hours from the onset and can be predicted by a lower baseline hypoperfusion intensity ratio, which indicates favorable collaterals. Methods- We identified control arm patients from the DEFUSE 3 trial (The Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke) with a diffusion weighted imaging and perfusion magnetic resonance imaging performed 24 hours following randomization and compared imaging and clinical variables between patients with persistent mismatch versus patients who no longer had a mismatch 24 hours after randomization. ⋯ Clinical trials are needed to determine if patients with a favorable perfusion profile benefit from reperfusion beyond 24 hours. Clinical Trial Registration- URL: https://www.clinicaltrials.gov. Unique identifier: NCT02586415.
-
Background and Purpose- Triaging of referrals to transient ischemic attack (TIA) clinics is aided by risk stratification. Deep learning-based natural language processing, a type of machine learning, may be able to assist with the prediction of cerebrovascular cause of TIA-like presentations from free-text information. Methods- Consecutive TIA clinic notes were retrieved from existing databases. ⋯ The greatest AUC was achieved when the convolutional neural network was provided with the history of presenting complaint and magnetic resonance imaging report (88.3±3.6). Conclusions- Deep learning-based natural language processing, in particular convolutional neural networks, based on medical free-text, may prove effective in prediction of the cause of TIA-like presentations. Future research investigating the role of the application of deep learning-based natural language processing to the automated triaging of clinic referrals in TIA, and potentially other specialty areas, is indicated.