Michels Lars, Muthuraman Muthuraman, Anwar Abdul R, Kollias Spyros, Leh Sandra E, Riese Florian, Unschuld Paul G, Siniatchkin Michael, Gietl Anton F, Hock Christoph
Clinic of Neuroradiology, University Hospital of ZurichZurich, Switzerland.
MR-Center, University Children's Hospital ZurichZurich, Switzerland.
Front Aging Neurosci. 2017 Sep 20;9:304. doi: 10.3389/fnagi.2017.00304. eCollection 2017.
The assessment of effects associated with cognitive impairment using electroencephalography (EEG) power mapping allows the visualization of frequency-band specific local changes in oscillatory activity. In contrast, measures of coherence and dynamic source synchronization allow for the study of functional and effective connectivity, respectively. Yet, these measures have rarely been assessed in parallel in the context of mild cognitive impairment (MCI) and furthermore it has not been examined if they are related to risk factors of Alzheimer's disease (AD) such as amyloid deposition and apolipoprotein ε4 (ApoE) allele occurrence. Here, we investigated functional and directed connectivities with Renormalized Partial Directed Coherence (RPDC) in 17 healthy controls (HC) and 17 participants with MCI. Participants underwent ApoE-genotyping and Pittsburgh compound B positron emission tomography (PiB-PET) to assess amyloid deposition. We observed lower spectral source power in MCI in the alpha and beta bands. Coherence was stronger in HC than MCI across different neuronal sources in the delta, theta, alpha, beta and gamma bands. The directed coherence analysis indicated lower information flow between fronto-temporal (including the hippocampus) sources and unidirectional connectivity in MCI. In MCI, alpha and beta RPDC showed an inverse correlation to age and gender; global amyloid deposition was inversely correlated to alpha coherence, RPDC and beta and gamma coherence. Furthermore, the ApoE status was negatively correlated to alpha coherence and RPDC, beta RPDC and gamma coherence. A classification analysis of cognitive state revealed the highest accuracy using EEG power, coherence and RPDC as input. For this small but statistically robust (Bayesian power analyses) sample, our results suggest that resting EEG related functional and directed connectivities are sensitive to the cognitive state and are linked to ApoE and amyloid burden.
使用脑电图（EEG）功率映射评估与认知障碍相关的效应，能够可视化振荡活动中特定频段的局部变化。相比之下，相干性测量和动态源同步测量分别用于研究功能连接和有效连接。然而，在轻度认知障碍（MCI）的背景下，这些测量很少同时进行评估，此外，尚未研究它们是否与阿尔茨海默病（AD）的风险因素如淀粉样蛋白沉积和载脂蛋白ε4（ApoE）等位基因出现有关。在此，我们使用重新归一化偏相干（RPDC）研究了17名健康对照（HC）和17名MCI参与者的功能连接和定向连接。参与者接受了ApoE基因分型和匹兹堡化合物B正电子发射断层扫描（PiB-PET）以评估淀粉样蛋白沉积。我们观察到MCI患者在α和β频段的频谱源功率较低。在δ、θ、α、β和γ频段，HC患者不同神经源之间的相干性比MCI患者更强。定向相干分析表明，MCI患者额颞叶（包括海马体）源之间的信息流较低且存在单向连接。在MCI患者中，α和β RPDC与年龄和性别呈负相关；全局淀粉样蛋白沉积与α相干性、RPDC以及β和γ相干性呈负相关。此外，ApoE状态与α相干性和RPDC、β RPDC以及γ相干性呈负相关。认知状态的分类分析显示，以EEG功率、相干性和RPDC作为输入时准确率最高。对于这个规模虽小但具有统计学稳健性（贝叶斯功效分析）的样本，我们的结果表明，静息EEG相关的功能连接和定向连接对认知状态敏感，并与ApoE和淀粉样蛋白负荷相关。
