4.0 Article

An individual data-driven virtual resection model based on epileptic network dynamics in children with intractable epilepsy: a magnetoencephalography interictal activity application

期刊

BRAIN COMMUNICATIONS
卷 5, 期 3, 页码 -

出版社

OXFORD UNIV PRESS
DOI: 10.1093/braincomms/fcad168

关键词

magnetoencephalography; interictal activity; computational neurosurgery; children epilepsy

向作者/读者索取更多资源

Cuesta et al. developed an in silico personalized protocol for epilepsy surgery planning using magnetoencephalography spike network analysis. The model showed high performance in a cohort of 24 paediatric patients with drug-resistant epilepsy. The study aims to improve surgical outcomes by testing surgical hypotheses in silico based on individual patient data.
Cuesta et al. present, in a retrospective study, an in silico personalized (no use of group data) protocol for epilepsy surgery planning using magnetoencephalography spike network analysis. The performance of the model in a cohort of 24 paediatric patients with focal drug-resistant epilepsy reached high accuracy, sensitivity and specificity scores. Epilepsy surgery continues to be a recommended treatment for intractable (medication-resistant) epilepsy; however, 30-70% of epilepsy surgery patients can continue to have seizures. Surgical failures are often associated with incomplete resection or inaccurate localization of the epileptogenic zone. This retrospective study aims to improve surgical outcome through in silico testing of surgical hypotheses through a personalized computational neurosurgery model created from individualized patient's magnetoencephalography recording and MRI. The framework assesses the extent of the epileptic network and evaluates underlying spike dynamics, resulting in identification of one single brain volume as a candidate for resection. Dynamic-locked networks were utilized for virtual cortical resection. This in silico protocol was tested in a cohort of 24 paediatric patients with focal drug-resistant epilepsy who underwent epilepsy surgery. Of 24 patients who were included in the analysis, 79% (19 of 24) of the models agreed with the patient's clinical surgery outcome and 21% (5 of 24) were considered as model failures (accuracy 0.79, sensitivity 0.77, specificity 0.82). Patients with unsuccessful surgery outcome typically showed a model cluster outside of the resected cavity, while those with successful surgery showed the cluster model within the cavity. Two of the model failures showed the cluster in the vicinity of the resected tissue and either a functional disconnection or lack of precision of the magnetoencephalography-MRI overlapping could explain the results. Two other cases were seizure free for 1 year but developed late recurrence. This is the first study that provides in silico personalized protocol for epilepsy surgery planning using magnetoencephalography spike network analysis. This model could provide complementary information to the traditional pre-surgical assessment methods and increase the proportion of patients achieving seizure-free outcome from surgery.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.0
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据