Annu Int Conf IEEE Eng Med Biol Soc. 2016 Aug;2016:1830-1833. doi: 10.1109/EMBC.2016.7591075.
Recent studies have found long-range temporal correlations (LRTC) of neuronal oscillation under different physiological and pathological state. In this paper we aimed to explore the temporal correlations of broadband neuronal oscillation in poststroke depression. Resting state EEGs of 16 electrodes in 16 poststroke depression (PSD), 18 poststroke non-depression (PSND) and 18 healthy control (HC) subjects were analyzed by means of detrended fluctuation analysis (DFA), a quantitative measurement of LRTC. The results indicated that LRTC of broadband oscillation in PSD subjects was still preserved and had similar spatial distribution to non-depression subjects, but attenuated especially in left hemisphere and frontal and posterior temporal regions of right hemisphere. Moreover, a linear correlation between the severity of psychiatric symptoms and attenuation of LRTC was observed in stroke subjects. ROC curves indicated a high discriminating ability of scaling exponents in prefrontal regions distinguishing PSD from PSND, suggesting the feasibility for LRTC as assessment indicators to PSD.
近期研究发现，在不同生理和病理状态下神经元振荡存在长程时间相关性（LRTC）。在本文中，我们旨在探究脑卒中后抑郁中宽带神经元振荡的时间相关性。采用去趋势波动分析（DFA，一种LRTC的定量测量方法）对16例脑卒中后抑郁（PSD）患者、18例脑卒中后非抑郁（PSND）患者和18例健康对照（HC）者的16个电极的静息态脑电图进行分析。结果表明，PSD患者宽带振荡的LRTC仍然存在，且与非抑郁患者具有相似的空间分布，但有所减弱，尤其是在左半球以及右半球的额颞叶和后颞叶区域。此外，在脑卒中患者中观察到精神症状严重程度与LRTC减弱之间存在线性相关性。ROC曲线表明，前额叶区域的标度指数对区分PSD和PSND具有较高的鉴别能力，提示LRTC作为PSD评估指标的可行性。
