基于OpenGL的脑电图机显控软件设计

发布时间：2018-04-25 15:48

本文选题：虚拟导联 + P300电位　；参考：《哈尔滨工程大学》2016年硕士论文

【摘要】：脑电信号(EEG)是大脑皮层中神经元细胞间交换物质、传递各感官反应和传送大脑指令时,电位变化产生的电压值。当人体受到外界刺激或主观思考时,脑电信号会发生变化,传统的脑电设备,显示功能较为单一,只能通过波形的变化来表示这些差异。本文设计的脑电图机显控软件,将头部模型引入显示端,并在模型中添加虚拟导联来模拟脑电测试时的电极片,在仿真中,通过对脑电信号的读取,将脑电信号的特征变化通过虚拟导联的不同状态来表征。此外,还对大脑模型进行功能区划分,每个功能区都与其包含的虚拟导联建立了特定的关系,当功能区内的虚拟导联发生状态变化时,该区域就会被标记为当前脑活动区,通过颜色变化在显示端直观表示。首先,本文通过3DMAX对头部模型建模,将头部模型依据脑功能区的划分标准,分割为5个区域,并根据10-20导联法为头部模型添加电极片模型。利用OpenGL图形接口完成对模型的读取,并设计模型的旋转、平移、拾取和颜色渲染的功能。在脑电的信号研究中科学家归纳一些脑电信号变化的规律,其中P300电位变化就是大脑应对刺激的一种规律反应。本文使用盲源分离算法对串口中传输的脑电信号做P300电位分析,通过对P300电位特点的先验理论,对含有P300电位的多导联混合脑电信号进行时频域滤波,去除多余数据,保留P300特征信号;然后使用FastICA算法对混合信号盲源分离,得到若干独立分量;对独立分量中时域峰值和能量熵分析,得到最接近P300电位的若干独立分量,最终选取这些独立分量中投影权值较大的若干导联作为当前活动电极。然后是对脑电图机软件人机交互和功能模块的实现,包括主界面设计模块、导联设置模块、串口配置模块以及波形绘制模块的人机交互实现和电极片状态显示、活动电极标识、功能区标识以及电极片拾取的功能实现。最后,通过4种特定刺激来验证算法获取活动电极的正确性以及对人机交互和功能模块进行测试。本文设计打破了传统脑电图机单一的波形输出功能,将脑电信号的变化在头部模型中直观显示,实现了信号量的可视化。本文所设计的脑电显控软件能够直观显示脑电信号变化动态,实现对电极片的模拟控制,可实现电极片不同工作状态下的区分显示,避免了在脑电信号测量时电极片未接入,或者漏测数据的问题,同时还实现了对虚拟导联的拾取功能,可对每一个电极片的工作状态单独查看,对异常工作的导联进行重新连接,返回到正常工作状态,还可以对无用电极进行断开连接,减少数据测试,节省资源。
[Abstract]:The EEG signal (EGG) is the voltage produced by the change of the electric potential when the neurons in the cerebral cortex are exchanged between the neurons to transmit the sensory responses and to transmit the instructions of the brain. When the human body is stimulated by the outside world or subjective thinking, the EEG signal will change. The traditional EEG equipment, the display function is relatively single, can only express these differences by the change of the waveform. In this paper, the EEG display control software is designed, the head model is introduced into the display terminal, and the virtual lead is added to the model to simulate the electroencephalogram (EEG) test. In the simulation, the EEG signal is read. The characteristic changes of EEG signals are represented by the different states of virtual leads. In addition, the functional regions of the brain model are divided, and each functional area has a specific relationship with the virtual lead it contains, and when the virtual lead in the functional area changes, the region is marked as the current brain active area. Visualize the display by changing the color. Firstly, the head model is modeled by 3DMAX, and the head model is divided into five regions according to the criteria of brain functional area, and the electrode sheet model is added to the head model according to the 10-20 lead method. The OpenGL graphic interface is used to read the model, and the functions of rotation, translation, pickup and color rendering of the model are designed. In the study of EEG signals, scientists conclude some laws of EEG changes, among which the change of P300 potential is a regular response of brain to stimulation. In this paper, the blind source separation algorithm is used to analyze the P300 potential of EEG transmitted in serial port. Through the priori theory of P300 potential characteristics, the multilead mixed EEG signal with P300 potential is filtered in time-frequency domain to remove the redundant data. The P300 characteristic signal is preserved, then the blind source of mixed signal is separated by FastICA algorithm, and some independent components are obtained by analyzing the time domain peak and energy entropy of the independent component, which is the closest to the P300 potential. Finally, some of these independent components with large projection weights are selected as the current active electrodes. Then the realization of man-machine interaction and function module of EEG software, including main interface design module, lead setting module, serial port configuration module and waveform drawing module, human-computer interaction realization and electrode sheet status display, active electrode identification, etc. The function of function area identification and electrode chip pickup is realized. Finally, four specific stimuli were used to verify the correctness of the algorithm to obtain the active electrode and to test the man-machine interaction and functional modules. In this paper, the function of single waveform output of traditional EEG machine is broken, the change of EEG signal is displayed intuitively in the head model, and the visualization of signal quantity is realized. The EEG display and control software designed in this paper can directly display the change of EEG signal, realize the simulation control of the electrode sheet, and realize the differential display of the electrode sheet in different working states, thus avoiding the non-connection of the electrode sheet when the EEG signal is measured. Or the problem of missing the test data, at the same time, the function of picking up the virtual lead is realized. The working state of each electrode sheet can be viewed separately, and the lead of abnormal work can be reconnected to return to the normal working state. It can also disconnect useless electrodes, reduce data testing and save resources.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.52;R318

【参考文献】