Generate the scale-free brain music from BOLD signals.
Generate the scale-free brain music from BOLD signals.
复制标题
从 BOLD 信号生成无标度大脑音乐
DOI：
10.1097/md.0000000000009628
复制
发表时间：
2018-01
期刊：
影响因子：
1.6
通讯作者：
Yao D
中科院分区：
文献类型：
--
作者：
Lu J;Guo S;Chen M;Wang W;Yang H;Guo D;Yao D
摘要将人类的脑电图（EEG）转换成音乐的方法有很多种。除了EEG，功能性磁共振成像（fMRI）是另一种用于研究大脑的方法，可以反映生理过程。2012年，我们建立了一种方法，使用同时记录的fMRI和EEG信号来产生EEG-fMRI音乐，这代表了无标度大脑音乐的一步。在这项研究中，我们使用了一个神经质量模型，Jansen-Rit模型，来模拟几个大脑皮层区域的活动。不同脑区之间的相互作用由平均归一化扩散张量成像（DTI）结构连接与耦合系数调制的耦合强度。78个大脑区域采用自动解剖标记（AAL）模板。此外，我们使用的神经活动转换成血氧水平依赖（BOLD）的信号的Sparkon-Windkessel血流动力学模型。由于fMRI BOLD信号变化缓慢，我们使用250 Hz的采样率来产生音乐生成的时间序列。 然后，使用这些模拟的BOLD信号为每个区域生成BOLD音乐。因为BOLD信号是无标度的，所以这些音乐片段也是无标度的，这与经典音乐类似。在这里，为了模拟癫痫患者的情况，我们改变了神经质量模型中确定兴奋性突触后电位（EPSP）幅度的参数。最后，我们获得了健康和癫痫患者的BOLD音乐。如果能够通过更多的真实的数据来证实这两段音乐之间的唤醒水平差异，那么这两段音乐之间的唤醒水平差异可能为区分不同人群提供潜在的工具。
Abstract Many methods have been developed to translate a human electroencephalogram (EEG) into music. In addition to EEG, functional magnetic resonance imaging (fMRI) is another method used to study the brain and can reflect physiological processes. In 2012, we established a method to use simultaneously recorded fMRI and EEG signals to produce EEG-fMRI music, which represents a step toward scale-free brain music. In this study, we used a neural mass model, the Jansen–Rit model, to simulate activity in several cortical brain regions. The interactions between different brain regions were represented by the average normalized diffusion tensor imaging (DTI) structural connectivity with a coupling coefficient that modulated the coupling strength. Seventy-eight brain regions were adopted from the Automated Anatomical Labeling (AAL) template. Furthermore, we used the Balloon–Windkessel hemodynamic model to transform neural activity into a blood-oxygen-level dependent (BOLD) signal. Because the fMRI BOLD signal changes slowly, we used a sampling rate of 250 Hz to produce the temporal series for music generation. Then, the BOLD music was generated for each region using these simulated BOLD signals. Because the BOLD signal is scale free, these music pieces were also scale free, which is similar to classic music. Here, to simulate the case of an epileptic patient, we changed the parameter that determined the amplitude of the excitatory postsynaptic potential (EPSP) in the neural mass model. Finally, we obtained BOLD music for healthy and epileptic patients. The differences in levels of arousal between the 2 pieces of music may provide a potential tool for discriminating the different populations if the differences can be confirmed by more real data.