table of Contents
This share is a brain-computer learner Rose finishing published in the public number: BCI community (Micro Signal: Brain_Computer) .QQ exchange group: 903 290 195
REVIEW
Brain computer interface (BCI) is a communication system, the brain signals by converting into machine instructions to help the user to interact with the external environment. EEG availability and reliability make BCI most commonly used method. Many EEG based brain-computer interface devices are using conventional wet or microelectromechanical system (MEMS) type EEG sensor development. However, these conventional sensors would cause an uncomfortable feeling when in contact with skin. Therefore, in a comfortable and convenient way to get EEG is an important part of a new BCI devices. In the present study, we have developed a device based on BCI wearable, portable and wireless EEG, the apparatus having a dry foam based EEG sensor, and a presentation by the game control applications. Dry non conductive plastic EEG sensor; however, they can provide good conductivity, the impedance can be effectively acquire electroencephalogram signals by adapting the forehead skin surface irregularities of the skin and maintain an appropriate sensor. The authors also demonstrated the use of portable devices proposed controls a game of real-time learning phase detection applications. The results show that use of such brain-computer interface device based portable EEG, can easily and effectively control the outside world, provides a way for the study of the rehabilitation project.
Research results
In this study, the researchers developed a wearable EEG-based BCI devices, and equipped with new sensors based on dry foam, and demonstrates the application of cognitive control of the game. The apparatus consists of a wireless EEG acquisition device and the computer. Wireless devices comprises a dry foam EEG acquisition sensors and wireless EEG acquisition module. Dry foam proposed sensor can work without the use of a conductive adhesive. These sensors can provide good electrical conductivity to efficiently obtain the EEG signals. Furthermore, the sensor may be suitably integrated into the wireless EEG acquisition device. BCI and other portable devices need to wet skin preparation process sensors compared, a user using the device in everyday life can be faster, more comfortable and more effective monitoring of the state of their EEG, and EEG signals can be transmitted directly to a personal computer signal processing. Further, to achieve a focus detection algorithm in real-time device, as an EEG-based game interface, in a comfortable way real-time detection of cognitive state of the user. Use the device complements other existing BCI methods used to study the behavioral responses of neurons activated state cognition and daily life.
Methods and Materials
A: Dry EEG sensors Design
EEG sensor based on the proposed specific design dry bulb using a conductive polymer material made of a polyurethane foam in contact with the skin of the forehead, the compression set of about 5 to 10%, as shown in Figure A and B. Covering the conductive foam material taffeta 0.2 mm thick, made of a conductive polymer fabric (conductivity of about 0.07 ohm / cm), and coated with a nickel / copper (Ni / Cu) on all of its surface to establish EEG electrical contact with a sensor similar to silver. Using 0.2 mm thick copper (Cu) as an adhesive layer, connected to the wireless EEG acquisition module. Dry bubble EEG proposed sensor is 20 × 20 × 9 mm3. Study design specification and an equivalent circuit of dry skin sensor interface EEG sensors follow.
B: wireless EEG acquisition module
下面两幅图为无线EEG采集模块及其在游戏控制中的应用。它用于从干式脑电图传感器获取脑电图信号，包括(INA2126, Texas Instruments)、一个采集组件(AD8609，模拟设备)、一个微处理器组件(MSP430,Texas Instruments)和一个无线传输组件(BM0403, Unigrand Ltd.)。为了放大和过滤脑电图信号，研究人员在电路板中嵌入了一个前置放大器、一个带通滤波器(0.5~ 50hz)和一个模数转换器(ADC)作为生物信号放大器和采集组件模块。放大器和采集组件的增益设置为大约5500。采用12位分辨率的ADC对脑电图信号进行数字化处理，对放大、滤波后的脑电图信号采样率为256hz。在微处理器组件中，使用ADC探测的脑电图信号被数字存储。在无线传输之前，使用频率为60hz的移动平均滤波器来排除电力线的干扰。蓝牙模块BM0403(Unigrand Ltd.)包含在电路的无线传输部分。需要注意的是，该模块完全符合蓝牙v2.0+ EDR和印刷电路板(PCB)天线的规范。总的来说，本论文提出的无线脑电图采集模块的尺寸大约为4.5×3×0.6 cm3，能够将该模块嵌入到基于可穿戴脑电图的BCI设备的机制中。该模块使用3.7 v直流电源，工作于31.58 mA。最重要的是，这个模块能够使用商用750毫安电池连续工作23小时。
C: 基于EEG的可穿戴BCI设备的机制
设计基于脑电图的BCI装置的快速放置机制，使干燥的脑电图传感器轻松快速地附着在用户的前额(F10)，如下图A所示。该装置由三个干泡传感器和一个无线脑电图采集模块组成，该模块包含一块电池。橡皮筋可以根据用户的头部大小进行调节，如下图A所示。利用干式脑电图传感器探测脑电图信号时，该机制也被用来最大化皮肤传感器的接触面积，以保持低阻抗。这一机制未对额头皮肤导致任何永久性或有害的影响。注意，多孔设备的所有通道都使用了基于干泡沫的电极。可穿戴式脑电图采集装置的应用，使用户更方便、舒适地监测脑电图信号。
D: EEG信号进行游戏控制
为了演示基于脑电图的干式传感器BCI装置在日常生活应用中的性能，研究人员提出了一个由用户通过脑电图信号的精神聚焦来控制的计算机游戏。游戏界面如图3所示。所有玩这个射箭游戏的用户都配备了这个基于脑电图的BCI设备。用户在游戏中进行射击，然后根据箭矢到靶心的距离来打分。屏幕右侧有一个栏，屏幕中心有一个目标，屏幕右上角有一个分数(下图A)。条形图显示了该用户在游戏期间的聚焦水平(FL)(下图B和下图C)。换句话说，FL值是游戏的主控制器。如果FL值较高，则说明射击距离目标中心较近，此时游戏得分较高。如果FL值较低，则射击距离目标中心较远，因此得分较低。用户的任务是使FL值尽可能高，通过射击接近目标的中心。用户有10秒的时间来完成一个镜头，10次之后计算总分。
评价指标
为了测量用户的FL值，提出了一种简单、实时、智能的游戏控制聚焦水平检测算法。该FL检测算法流程图如图4所示。FL检测算法包括三个主要步骤:
1)去除伪迹信号;
2)提取聚焦特征;
3)确定FL值。
首先对原始脑电图信号进行预处理，去除噪声信号。众所周知，精神集中状态与脑电图前额区域的alpha节律(8~ 12hz)高度相关，并且噪声伪影位于不同于alpha节律频率范围的频率区域。因此，为了剔除伪影，对信号进行快速傅里叶变换，得到信号的功率谱模式，保留在alpha波段内的信号。
其次，对alpha波段内的功率谱进行聚焦特征提取。之前的研究表明，随着用户的精神状态从专注状态转变为非专注状态，脑电图的alpha节律的力量也随之增强。因此，alpha波段在本研究中用来表示用户聚焦状态的主频带，FL检测算法选择原始脑电图信号的8~ 12hz频带。将Focus Feature (FF)定义为alpha节奏中平均功率的倒数，如下式所示:
实验结果
作者报告了干式传感器和电路的测试结果，以确保它们在日常生活中可用于测量EEG信号。基于可穿戴式EEG的BCI设备的主要组件包括干式EEG传感器及其相应的读出电路。关于信号质量和皮肤传感器界面之间的阻抗，对干式EEG传感器进行了实验表征。
下图中显示了用于验证信号质量的预测试实验。该预测实验的目的是识别在EEG测量期间由干式EEG传感器引起的任何失真。首先，使用带有导电胶的标准EEG传感器预先记录EEG数据，并将其存储在计算机中。接下来，将脑电图数据输入到可编程功能发生器中，并通过分压器生成模拟人脑电图信号。然后将模拟的EEG信号输入到干式EEG传感器中，并与记录的和预记录的EEG数据进行比较。
下图显示了预先记录的EEG信号和研究人员提出的干式EEG传感器记录的信号。预记录的EEG信号与使用干式EEG传感器获得的信号高度相关，达到97.68％的水平。预记录的EEG信号与使用干式EEG传感器获得的数据之间的高度相关性证实了使用基于干式泡沫的传感器记录的EEG信号的清晰度。
接下来，研究了传统湿式EEG传感器和干式EEG传感器之间的相关性。下图显示了在用户的额头上使用干/常规EEG传感器对之后传感器的位置和EEG测量的结果（F10）。对于前额，使用干式EEG传感器和常规湿式EEG传感器获得的信号之间的相关性通常超过95.56％。 因此，使用基于泡沫的干式EEG传感器进行的EEG信号测量的性能与常规的湿式EEG传感器相同。
此外，还使用阻抗谱法(LCR4235,Wayne Kerr Electronics Ltd.)测量传感器-皮肤接触界面的阻抗。传统的脑电图传感器利用自粘特性附着在使用者前额左侧的皮肤上。干燥的脑电图传感器用一条3米长的一次性皮带固定，每次测量之间小心地更换，以避免皮肤表面的任何变化。使用者的皮肤被用带有2-丙醇的棉垫轻轻擦拭干净，在使用传感器之前，使2-丙醇蒸发。为了保证结果的可靠性和可重复性，将阻抗谱测试信号设置为1v，频率范围设置为0.5 ~1000Hz。对五名不同的参与者进行了十项测试，分别测试两种不同的脑电图传感器(湿式和干式)。下图显示了不同条件下的阻抗测量结果。在下图中，黑线表示未使用皮肤制剂或导电凝胶的干性EEG传感器对的阻抗。蓝色和红色的线分别表示未使用皮肤制剂和使用皮肤制剂时常规脑电图传感器的阻抗。常规脑电图传感器均采用导电凝胶法。结果表明，皮肤与无皮肤准备或导电凝胶的干性脑电图传感器之间的阻抗水平接近有皮肤准备和前额部位导电凝胶的湿性脑电图传感器(F10)。因此，干式脑电传感器在传导性能方面与传统的脑电传感器有一定的差距。
下图显示了不同传感器(干湿传感器)在长期EEG测量中的阻抗变化。在长期的脑电图测量中，导电凝胶和皮肤制备的常规脑电图传感器的阻抗变化高于干式脑电图传感器。干式脑电图传感器的阻抗变化范围为4~12 kOhm，在正常脑电图测量的可接受范围内。此外，在长期的EEG测量(2小时)中，与传统的EEG传感器相比，干式传感器在皮肤电极阻抗方面提供了可靠的信号质量(图10)。这一结果可以解释为，干燥的脑电图传感器不需要导电凝胶，在测量过程中容易干燥，因此降低了相对于湿传感器的稳定性。
结论
Experimental results show that, based on a dry foam EEG sensors successful, stable wireless EEG electroencephalogram measured by the respective collecting means; these results with conventional EEG sensor using a conductive gel substantially the same. Thus, compared with the conventional wet EEG-based BCI sensor device, based on a dry foam EEG sensors have the potential to allow regular and repeated measurement. And successfully portable, wireless, low-power EEG acquisition module for long-term monitoring of EEG. The dry EEG sensors and wireless EEG acquisition module embedded wearable electroencephalograph collecting apparatus. EEG-based BCI wearable devices, do not use conductive gel, allowing users more comfortable state EEG monitoring in everyday life.
In this study, the researchers also use this portable device to demonstrate the application based on the EEG cognitive control of the game. A personal computer as a platform to run real-time focus feature detection algorithms and EEG monitoring program to monitor a user's cognitive state. Experimental data show that BCI EEG-based wearable device and a corresponding control algorithm can be reliably used for a general user or application outside world researchers. The apparatus BCI complements other existing methods for the study of human neuron firing states and cognitive behavioral response in everyday life.
Reference
using EEG sensors game testing the effectiveness of a dry
This article was shared by brain-computer learner Rose notes, QQ exchange group: 903 290 195
more to share, please pay attention to the public No.
