Pedersen Mangor, Curwood Evan K, Vaughan David N, Omidvarnia Amir H, Jackson Graeme D
1 The Florey Institute of Neuroscience and Mental Health , Austin Campus, Melbourne, Australia .
2 The University of Melbourne , Florey Department of Neuroscience and Mental Health, Melbourne, Australia .
Brain Connect. 2016 Apr;6(3):208-15. doi: 10.1089/brain.2015.0367. Epub 2016 Jan 7.
Individuals with focal epilepsy have heterogeneous sites of seizure origin. However, there may be brain regions that are common to most cases of intractable focal epilepsy. In this study, we aim to identify these using multivariate analysis of task-free functional MRI. Fourteen subjects with extratemporal focal epilepsy and 14 healthy controls were included in the study. Task-free functional MRI data were used to calculate voxel-wise regional connectivity with regional homogeneity (ReHo) and weighted degree centrality (DCw), in addition to regional activity using fraction of amplitude of low-frequency fluctuations (fALFF). Multivariate pattern analysis was applied to each of these metrics to discriminate brain areas that differed between focal epilepsy subjects and healthy controls. ReHo and DCw classified focal epilepsy subjects from healthy controls with high accuracy (89.3% and 75%, respectively). However, fALFF did not significantly classify patients from controls. Increased regional network activity in epilepsy subjects was seen in the ipsilateral piriform cortex, insula, and thalamus, in addition to the dorsal anterior cingulate cortex and lateral frontal cortices. Decreased regional connectivity was observed in the ventromedial prefrontal cortex, as well as lateral temporal cortices. Patients with extratemporal focal epilepsy have common areas of abnormality (ReHo and DCw measures), including the ipsilateral piriform cortex, temporal neocortex, and ventromedial prefrontal cortex. ReHo shows additional increase in the "salience network" that includes anterior insula and anterior cingulate cortex. DCw showed additional effects in the ipsilateral thalamus and striatum. These brain areas may represent key regional network properties underlying focal epilepsy.
局灶性癫痫患者的癫痫发作起源部位具有异质性。然而，大多数难治性局灶性癫痫病例可能存在一些共同的脑区。在本研究中，我们旨在通过对静息态功能磁共振成像进行多变量分析来识别这些脑区。该研究纳入了14名颞叶外局灶性癫痫患者和14名健康对照者。除了使用低频振幅分数（fALFF）来计算区域活动外，静息态功能磁共振成像数据还用于通过区域一致性（ReHo）和加权度中心性（DCw）来计算体素水平的区域连通性。对这些指标分别应用多变量模式分析，以区分局灶性癫痫患者和健康对照者之间存在差异的脑区。ReHo和DCw能够以较高的准确率将局灶性癫痫患者与健康对照者区分开来（分别为89.3%和75%）。然而，fALFF未能显著区分患者与对照者。癫痫患者的区域网络活动增加，表现于同侧梨状皮质、岛叶和丘脑，以及背侧前扣带回皮质和外侧额叶皮质。前额叶腹内侧皮质以及颞叶外侧皮质的区域连通性降低。颞叶外局灶性癫痫患者存在共同的异常区域（ReHo和DCw测量指标），包括同侧梨状皮质、颞叶新皮质和前额叶腹内侧皮质。ReHo显示在包括前岛叶和前扣带回皮质的“突显网络”中有额外增加。DCw在同侧丘脑和纹状体有额外影响。这些脑区可能代表了局灶性癫痫潜在的关键区域网络特性。
