Predicting Epilepsy Surgery Outcomes from Presurgical Temporal Lobe Network Architecture
Neurology(2018)
摘要
Objective: We aimed to determine whether presurgical plastic reorganization of temporal lobe structures is related to the likelihood of seizure freedom after surgery. Background: Even though surgery for medial temporal lobe epilepsy (MTLE) is one of the most effective treatments in modern medicine, many patients remain with seizures after surgery, and the reasons for suboptimal outcomes are not well known. Design/Methods: We constructed the individual presurgical structural brain connectome of 50 patients with MTLE who underwent surgery for epilepsy and computed betweenness centrality (BC) (an integration measure of regional network hubness derived from graph theory), of all gray matter regions of interest (ROIs) throughout the brain. Only ipsilateral temporal ROIs were entered into a discriminant analysis function and then cross-validated classification outcomes by using a leave-one-out approach. We also derived discriminant functions from clinical variables alone and a combination of clinical and connectome measures. Results: Our results revealed that a discriminatory function constituted by the BC of six ipsilateral temporal ROIs had a classification accuracy of 90% for the original cases and 82% on cross-validation, and was especially reliable at predicting which patients would be seizure free after surgery (97.2% of the original group, 91.6% on cross-validation). In addition, the discriminatory function based on connectome measures alone was more accurate in classifying surgical outcomes than the functions based on clinical data alone (accuracy of 46%). Conclusions: Our results indicate that increased structural integration of gray matter regions in the temporal lobe ipsilateral to seizure onset is largely associated with lower probability of seizure freedom after surgery. Disclosure: Dr Gleichgerrcht has nothing to disclose. Dr. Bhatia has nothing to disclose. Dr. Edwards has nothing to disclose. Dr. Vandergrift has nothing to disclose. Dr. Kuzniecky has nothing to disclose. Dr. Bonilha has nothing to disclose.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络