arXiv 2403.173662024-03-26physics.med-ph

癫痫发作与心电图脑灌注指数的时空相关性

Spatio-Temporal Correlation of Epileptic Seizures with The Electrocardiography Brain Perfusion Index

Samuel J van Bohemen, Joe O Nardo, Jeffrey M Rogers, Eleanor Stephens, Chong H Wong, Andrew F Bleasel, Andre Z Kyme

AI总结：

本研究探讨心电图脑灌注指数(EBPi)与癫痫发作的时空相关性，发现EBPi对发作起止及位置具有时间敏感性和一定空间预测能力，可辅助增强发作期间的脑电图(EEG)数据。

摘要：

心电图脑灌注指数（EBPi）是一种基于心电图（ECG）的新型指标，可作为脑血流量（CBF）的替代指标。我们研究了EBPi与癫痫发作事件之间的时空相关性。EBPi是回顾性地从30名癫痫患者在癫痫发作期间捕获的临床脑电图（EEG）和ECG数据中计算得出的。将EBPi的显著变化在时间上与临床定义的真实癫痫发作起始和结束时间进行了比较。我们还评估了EBPi指标与临床定义的真实癫痫发作位置之间的空间相关性。在真实癫痫发作起始后10.5秒[-6, 53]（中位数[95%置信区间(CI)]）检测到EBPi显著增加，在真实癫痫发作结束后5秒[-42, 74]（中位数[95% CI]）检测到EBPi显著降低。在检测真实癫痫发作位置方面，EBPi表现出61.5% [33.3, 75]（中位数[95% CI]）的阳性预测值和57.1% [38.5, 66.7]（中位数[95% CI]）的敏感度。EBPi信号对癫痫发作事件表现出时间敏感性，在某些情况下与癫痫发作位置在空间上具有相关性。因此，与CBF相关联的EBPi似乎包含与癫痫发作活动相关的时空信息，可能在增强癫痫发作期间捕获的EEG数据方面具有有用的应用。

英文摘要：

The Electrocardiography Brain Perfusion index (EBPi) is a novel electrocardiography (ECG)-based metric that may function as a proxy for cerebral blood flow (CBF). We investigated the spatio-temporal correlation between EBPi and epileptic seizure events. EBPi was computed retrospectively from clinical EEG and ECG data captured previously from 30 epilepsy patients during seizures. Significant EBPi changes were compared temporally with clinically defined ground-truth seizure onset and offset times. We also assessed the spatial correlation between EBPi metrics and clinically defined ground-truth seizure locations. A significant increase in EBPi was detected 10.5 s [-6, 53] (median [95% confidence interval (CI)]) after ground-truth seizure onset, and a significant decrease in EBPi was detected 5 s [-42, 74] (median [95% CI]) after ground-truth seizure offset. EBPi demonstrated a positive predictive value of 61.5% [33.3, 75] (median [95% CI]) and a sensitivity of 57.1% [38.5, 66.7] (median [95% CI]) for the detection of ground truth seizure locations. EBPi signals exhibited a temporal sensitivity to seizure events and in some cases were correlated spatially to the seizure location. Therefore, EBPi, which has been linked to CBF, appears to contain spatio-temporal information related to seizure activity and might have useful application in augmenting EEG data captured during seizures.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。