Collection of electroencephalogram signal?

According to the way of EEG production, there are two kinds: spontaneous EEG (EEG) and evoked EEG (EP).(按照脑电产生的方式可以分为自发脑电（EEG）和诱发脑电（EP）两种。)

The BCI system based on EEG realizes control by real-time or short-time extraction and analysis of signals reflecting different states of brain in EEG. It is worth mentioning that it is not realistic to extract the various thinking activities that people are doing through electroencephalogram reading. BCI is to generate easy-to-explain electroencephalograms, and then identify this electroencephalogram, make different choices or issue different instructions. The following six research methods were adopted according to different electroencephalogram signal and methods used in BCI: (基于脑电的 BCI 系统, 通过实时或短时提取分析出脑电中的反映大脑不同状态的信号来实现控制。值得提出的是,目前通过脑电读取出人正在进行的各种思维活动还不现实 ,BCI 是使人产生容易被解释的脑电,然后识别出这种脑电 ,作出不同的选择或发出不同指令。根据 BCI 利用的脑电信号和方式不同 , 采用以下6 种研究方法)

P300 event-related potential method: P300 is an event-related potential (ERP), its peak appears about 300 ms after the event, the smaller the probability of the event. The more significant the P300 is caused. The advantage of P300-based BCI is that users can generate P300 without training. (P300事件相关电位法：P300 是一种事件相关电位(ERP), 其峰值大约出现在事件发生后 300 ms, 相关事件发生的概率越小,所引起的P300 越显著。基于 P300 的 BCI 的优点是, 使用者无须通过训练就可产生P300);
Steady-state visual evoked potentials method: The first method is to generate SSVEP with 13.25 Hz sine-modulated white fluorescence stimulation. Subjects learn to control the amplitude of SSVEP through training. The second type does not need training. It identifies SSVEPs of different frequencies to realize control. There are two buttons on the screen that flash at different frequencies. The subject gazes at the button to be selected, and the amplitude of the corresponding frequency component in SSVEP increases. (稳态视觉诱发电位法：第一种是用 13 .25 Hz 正弦调制的白色荧光刺激产生 SSVEP , 受试者通过训练学习控制SSVEP 的幅度。第二种不需要训练, 是识别不同频率的 SSVEP 来实现控制 ,屏幕上有 2 个按钮, 以不同频率闪烁 ,受试者注视要选的按钮, SSVEP 中相应的频率成分的幅度增加);
Event-related synchronization or desynchronization: During unilateral limb movement or imagination movement, event-related desynchronization potential (ERD) is generated on the opposite side of the brain, and event-related synchronization potential (ERS) is generated on the same side of the brain. (事件相关同步或去同步法：单边的肢体运动或想象运动, 大脑对侧产生事件相关去同步电位(ERD), 大脑同侧产生事件相关同步电位(ERS));
Cortical slow potential method: The cortical slow potential (SCP) is the change of cortical potential, which lasts from hundreds of milliseconds to several seconds and can reflect the excitability of cortical layers I and II. Healthy people and paralyzed patients learn through feedback training. Make the SCP amplitude generate a positive or negative shift.(皮层慢电位法：皮层慢电位(SCP)是皮层电位的变化 , 持续时间为几百毫秒到几秒, 能反映皮质 I 和 II 层的兴奋性,健康人和瘫痪病人通过反馈训练学习 , 使 SCP 幅度产生正向或负向偏移);
Spontaneous EEG method: alpha rhythm and mu rhythm, frequency range 8-13 Hz, related to the relaxation state of the subject. The alpha rhythm detected in the occipital visual cortex reflected visual relaxation, and the mu rhythm detected in the sensorimotor cortex reflected motor relaxation. Subjects' ability to generate strong spontaneous electroencephalogram (EEG) can be enhanced by biofeedback or manipulation training. By adjusting the EEG according to the feedback, subjects can learn to control the equipment through training. (自发脑电信号法：α节律和 mu 节律，频率范围为8~13Hz，与受试者的松弛状态有关。在枕部视觉皮层区检测到的 α节律可反应出受试者处于视觉松弛状态，在感觉运动皮层区检测的mu节律反应运动松弛状态。受试对象产生强的自发脑电（EEG）的能力可通过生物反馈或操作训练得到加强。根据反馈调节EEG，受试对象可通过训练来学会控制设备);
Implanted electrode method: The microelectrode implanted in the intracranial has high spatial and frequency resolution, can provide the electrical activity information of a few neurons near the electrode, has good positioning and high signal-to-noise ratio. The implanted electrode is not affected by muscle movement, can be fixed on the head for a long time, has good position stability, and is suitable for specific patients or specific occasions. (植入电极法：植入颅内的微电极具有较高的空间和频率分辨率,能提供电极附近少数神经元的电活动信息 ,定位性好,信噪比高。植入电极不受肌肉运动的影响 ,可以在头部固定较长时间, 具有较好的位置稳定性, 适合特定的病人或特定的场合); The above six research methods have their own characteristics and limitations. Both P300 and SSVEP belong to evoked potentials and do not need training. Because P300 appears at a specific time and SSVEP concentrates on a specific frequency, the signal detection and processing method is simple and accurate, and the disadvantage is that additional stimulation devices are required to provide stimulation. and depends on some kind of human perception (e.g. vision). Several other methods have the advantage of not relying on external stimuli to generate EEG signals for control, but require a lot of training before using the device. BCIs based on slow cortical potentials, alpha rhythms, and mu rhythms require long-term training to produce controlled and stable EEG signals.(上述 6 种研究方法, 具有各自的特点和局限。P300 和 SSVEP 都属于诱发电位, 不需要训练 , 由于P300 出现在特定时间, SSVEP 集中于特定频率 , 其信号检测和处理方法较简单且正确率较高 ,不足之处是,需要额外的刺激装置提供刺激, 并且依赖于人的某种知觉(如视觉)。其它几种方法的优点是不依赖外部刺激产生用于控制的脑电信号 ,但使用设备前需要大量的训练。基于皮层慢电位、α节律、mu 节律的BCI ,使用者需要进行长时间的训练, 以产生可以控制的稳定的脑电信号。)

The disclosed brain wave dataset is obtained in the following manner:(公开的脑电波数据集获取方式如下：)

PhysioNet is an open medical database that includes multiple brain wave datasets. It is available free of charge on the PhysioNet website (https://physionet.org/). (PhysioNet是一个公开的医学数据库，其中包括多个脑电波数据集。可以通过PhysioNet网站（https://physionet.org/）免费获取。);
BCI Competition is a brain-computer interface competition that includes multiple brain wave datasets. It can be obtained free of charge through the BCI Competition website at (http://www.bbci.de/competition/). (BCI Competition是一个脑机接口竞赛，其中包括多个脑电波数据集。可以通过BCI Competition网站（http://www.bbci.de/competition/）免费获取。);
OpenBCI is an open source brain-computer interface hardware and software platform that includes multiple brain wave datasets. It is available for free on the OpenBCI website (https://openbci.com/pages/datasets). (OpenBCI是一个开源的脑机接口硬件和软件平台，其中包括多个脑电波数据集。可以通过OpenBCI网站（https://openbci.com/pages/datasets）免费获取。);
Kaggle is a data science competition platform that includes multiple brainwave datasets. It is available for free on the Kaggle website at (https://www.kaggle.com/datasets?search=EEG). (Kaggle是一个数据科学竞赛平台，其中包括多个脑电波数据集。可以通过Kaggle网站（https://www.kaggle.com/datasets?search=EEG）免费获取。);
UCI Machine Learning Repository is an open database of machine learning datasets that include multiple brain wave datasets. It is available for free on the UCI Machine Learning Repository website (https://archive.ics.uci.edu/ml/datasets/EEG+Database). (UCI Machine Learning Repository是一个公开的机器学习数据集库，其中包括多个脑电波数据集。可以通过UCI Machine Learning Repository网站（https://archive.ics.uci.edu/ml/datasets/EEG+Database）免费获取。);
The MetaBCI platform is compiled based on the international open source language Python, standardizes the brain-computer interface data structure and preprocessing process, develops a general decoding algorithm framework, and uses dual processes and dual threads to improve the real-time efficiency of the online system. It can realize the whole process of inducing, obtaining, analyzing, and converting the user's brain intention. The platform contains 376 classes and functions, is compatible with 14 BCI public datasets, 16 data analysis methods, and 53 brain-computer decoding models. All its code has been publicly shared on GitHub, the world's largest open source programming and code hosting website, and is accompanied by an instruction manual. Provide platform-level technical support for global brain-computer interface developers and enthusiasts. The basic architecture of the MetaBCI platform consists of three modules. The Brainda platform, which is oriented to offline analysis requirements, unifies the existing open dataset interface and integrates multiple main BCI data analysis methods and decoding algorithms to improve the algorithm development efficiency of researchers. The Brainstim platform, which is oriented to stimulation presentation requirements, provides simple and efficient paradigm design modules to quickly create brain-computer interface paradigm stimulation interfaces. The Brainflow platform, which is oriented to online development requirements, realizes real-time and high-speed data reading, data processing, and result feedback, which significantly lowers the technical threshold of the brain-computer interface online system.(MetaBCI平台基于国际通行开源语言Python编写，规范了脑机接口数据结构与预处理流程，开发了通用的解码算法框架，利用双进程和双线程提高在线系统的实时效率，能够实现对用户大脑意图的诱发、获取、分析和转换等全流程处理。平台共包含 376 个类和函数，兼容 14 种 BCI 公开数据集，涵盖 16 种数据分析方法和 53 种脑机解码模型，其全部代码已在全球最大的开源编程及代码托管网站GitHub 公开共享，并配套使用说明手册，为全球脑机接口开发者、爱好者提供平台级的技术支持。MetaBCI 平台基本架构包含三大模块。其中，面向离线分析需求的Brainda平台，统一了现有公开数据集接口，集成多种主要BCI数据分析方法及解码算法，以此提高研究者的算法开发效率；面向刺激呈现需求的Brainstim平台，通过提供简洁高效的范式设计模块，可快速创建脑机接口范式刺激界面；面向在线开发需求的Brainflow 平台，实现了实时高速的数据读取、数据处理、结果反馈等功能，显著降低了脑机接口在线系统的技术门槛。)

Studies have shown that rehabilitative training activates a larger cortical area in the affected side of the brain. The activity of sensorimotor areas in the damaged brain increased with time, but the activity of undamaged brain areas did not change. For spectral characteristics, the higher frequency band 25-38 Hz is active at the beginning, but the lower band gradually becomes active during rehabilitation, replacing the higher band. The dynamic change of frequency band indicates that rhythm changes in time sequence with rehabilitation training.(研究表明,康复训练会练激活了患侧脑区更大的皮层区域。受损脑区感觉运动区的活跃程度随着时间不断增强，而未受损脑区活跃程度没有变化。对于频谱特性，在开始阶段，频率较高的波段25-38Hz 比较活跃，然而在康复过程中较低的波段逐渐活跃，代替了较高的波段。频率波段动态变化的现象表明节律随着康复训练是时序变化的。)

cmy2022 / neural-interface

Collection of electroencephalogram signal? #7