A comparative study of multiple resting state functional magnetic resonance imaging indices to detect epileptic activity in childhood absence epilepsy
10.3760/cma.j.cn113694-20221204-00904
- VernacularTitle:多个静息态功能磁共振成像指标检测儿童失神性癫痫脑活动的比较研究
- Author:
Qianqian YU
1
;
Qiang XU
;
Gaoping LIU
;
Qirui ZHANG
;
Guangming LU
;
Zhiqiang ZHANG
Author Information
1. 南京大学医学院附属金陵医院放射诊断科，南京210002
- Keywords:
Epilepsy, absence;
Child;
Magnetic resonance imaging;
Electroencephalography;
Resting state functional magnetic resonance imaging
- From:
Chinese Journal of Neurology
2023;56(9):1009-1017
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To comprehensively evaluate the ability of common resting state functional magnetic resonance imaging (rs-fMRI) indices to detect abnormal brain activity in childhood absence epilepsy (CAE).Methods:Simultaneous electroencephalography-functional magnetic resonance imaging (fMRI) data of 20 patients with CAE who were treated in the Jinling Hospital, Nanjing University School of Medicine from February 2010 to September 2021 were retrospectively collected. After excluding 2 patients with CAE with greater head movement, 44 fMRI data containing discharges from 18 patients were obtained finally. The generalized spike and slow-wave discharges (GSWD) related fMRI activation mappings were obtained by using the generalized linear model. At the same time, 94 age- and sex-matched healthy controls underwent rs-fMRI scanning. Meanwhile, 12 indices of rs-fMRI were calculated respectively [amplitude of low frequency fluctuation (ALFF), fractional amplitude of low frequency fluctuation (fALFF), regional homogeneity (ReHo), functional connectivity density (FCD), long FCD, local FCD, granger causality density (GCD)-in, GCD-out, GCD-int, resting state functional magnetic resonance imaging lag analysis (RSLA), Hurst index and brain entropy]. Two-sample t-tests were employed to detect significant differences in 12 indices of rs-fMRI. The Dice coefficient was used to evaluate the overlap between different brain maps of 12 indices of rs-fMRI and the GSWD-related blood oxygenation level dependent (BOLD) activation. Results:Positive activation of GSWD-related BOLD in CAE was mainly in the bilateral thalamus, and negative activation was mainly in default mode network (DMN) related brain regions. There was a significant overlap between the abnormal brain regions detected by various resting-state indicators: compared with normal controls, ALFF, fALFF, ReHo, GCD-in, GCD-out and local FCD were elevated in the bilateral thalamus, while FCD, long FCD, GCD-int and RSLA were decreased in CAE; ALFF, fALFF, ReHo, local FCD, GCD-out, RSLA and brain entropy were decreased in the DMN, while FCD, long FCD, GCD-in and GCD-int were increased in CAE. The Dice coefficient of long FCD was the highest (0.365),FCD was 0.362, while the Hurst index showed the lowest (0.142).Conclusions:Rs-fMRI indices variously revealed abnormal brain activity in CAE, in which the FCD is better for detection of epileptic activity. Rs-fMRI could be helpful to understand the pathophysiological mechanism of CAE, and to find reliable imaging markers.
