Voxel-Based Lesion Symptom Mapping to Predict Poststroke Epilepsy After Mechanical Thrombectomy

基于体素的病灶症状映射预测机械取栓术后卒中后癫痫

阅读:1

Abstract

INTRODUCTION: Poststroke epilepsy (PSE) is an important long-term complication after stroke. Data regarding predictors of PSE in patients with large-vessel occlusion stroke receiving mechanical thrombectomy (MT) are scarce. Voxel-based lesion symptom mapping on brain MRI might be a valuable tool in the risk prediction of PSE. This study aims to assess PSE risk after acute stroke treated with MT via voxel- and volumetric-based analyses. METHODS: In this bi-center study from two tertiary-care stroke centers, we included consecutive acute ischemic stroke patients who had received MT between 2011 and 2017, and had postinterventional brain MRI as well as long-term follow-up data available. Infarct volume and location were assessed on MRI. Following semiautomated lesion outlining and generation of binarized lesion masks, lesion symptom mapping was applied to identify relevant topographical lesion patterns in PSE. RESULTS: Of 348 analyzed patients, 97 cases had to be excluded due to insufficient image quality and inaccurate registration results. Finally, lesion maps from 251 patients (median age: 66, 45.4% women) were considered for lesion symptom mapping, including maps from 26 patients with PSE (10.4%). Mean infarct volume was higher in PSE patients (119.2 cm(3) vs. 43.9 cm(3), p < 0.0001). Lesion symptom mapping identified the orbitofrontal region, the operculum, and the temporal pole as brain regions associated with PSE. CONCLUSION: Apart from infarct volume, lesion symptom mapping on postinterventional brain MRI identified specific brain regions associated with PSE after large vessel occlusion stroke. This information might be helpful for PSE risk stratification and follow-up care in this specific population.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。