One-Sensor P300 EEG and Blink EOG from OpenBCI Cyton Board

Two different systems have been developed in this study. An EEG control system based on P300 EP and an EOG system based on eye blinks. Both developed systems, the objects that the user can interact with are presented on a laptop screen using a GUI. This stimulation program was designed and developed in Python employing the PsychoPy package. The GUI, composed by a 3×3 matrix containing nine images describing the objects and control functions to be executed by users. The first two rows correspond to home devices, such as TVs, digital house locks or electric lights, which have a state that can be switched (e.g., on/off or up/down) by selecting it with the HMI. The last row has three possible phone calls: the emergency call (i.e., an SOS number) and two favourite contacts in their phone-book. Two stimulation paradigms were assessed for the experiments. For the first paradigm, each element of the matrix is intensified one by one. This intensification is randomly performed, but all the elements are intensified the same number of times. Therefore, they will have the same probability of being intensified, that is, 𝑝=1/9. Moreover, an element cannot be intensified two consecutive times..For the second paradigm, instead of element by element, each row and column of the matrix is intensified. This is also done randomly, and they are intensified the same number of times. Therefore, the probability of one element being intensified at any given time is 𝑝=2/6=1/3. EEG and EOG data have been recorded using the OpenBCI Cyton Board with a sampling frequency of 250Hz. For EEG recordings, since the objective is to detect the P300 potential, the electrode for the input channel was located in the Cz position. Conversely, for EOG recordings, the electrode was placed at Fp2 in order to detect activity related to eye blinks. Both recording modes shared reference and ground electrodes, placed at the right mastoid (RM) and the left earlobe (A1), respectively. The participant group in our experiments included a total of nine volunteers who agreed to collaborate in this research. The participants indicated that they did not have hearing or visual impairments. Informed consents were obtained from all the participants in order to employ their data in our study. The experiments were carried out in a sound-attenuated room where the participants were invited to sit in a comfortable chair while focusing their attention on a 15.6-inch laptop screen where stimuli were presented. Recording sessions for each participant were divided into four independent experiments, one per each stimulation paradigm and recording mode that is: for the 1-by-1 paradigm with EEG signals, for the row/column paradigm with EEG signals, for the 1-by-1 paradigm with EOG signals and for the row/column paradigm with EOG data. Eight runs were recorded for each experiment. Each subject recording is stored in one separate folder, named s[number_subject], e.g., for subject 1: folder s1. Inside each subject folder we will find 32 folders corresponding with each experiment. The structure of the name is the following: subject_controlSignal_paradigm_date, eg, s1P3001by1-12-11-2020_12-57-15. Inside each experiment folder we will find 3 files. ExInfo : Contains data related to the subject, channel, paradigm and also the target element of that experiment. Raw-STIMS: Contains the stimuli presentation for that experiment. It is important to note that each specific stimuli is represented by a code. In the 1-by-1 paradigm these codes range from 101 to 109 and represent each of the elements of the 3x3 matrix, from the upper left corner to the bottom right corner. In the row-column paradigm these codes range from 101 to 106. Codes 101-103 represents the rows of the matrix from top to bottom. Codes 104-106 represent the columns of the matrix from left to right. Raw-EEG: Contains the EEG data recorded from the Cyton board. For more details about the recordings and the data plase see the related paper of this study and also the associated code: Sensors | Free Full-Text | Proposals and Comparisons from One-Sensor EEG and EOG Human-Machine Interfaces (mdpi.com) franciscolaport/Proposals-and-Comparisons-from-One-Sensor-EEG-and-EOG-Human-Machine-Interfaces (github.com)本研究开发了两套不同的系统：一套基于P300事件相关电位（P300 EP）的脑电（EEG）控制系统，以及一套基于眨眼动作的眼电（EOG）控制系统。在两套已开发的系统中，用户可交互的操作对象均通过图形用户界面（GUI）展示于笔记本电脑屏幕。本次实验的刺激程序采用Python语言编写，并依托PsychoPy工具包完成开发。该图形用户界面由3×3矩阵构成，矩阵内包含九张图片，分别代表用户可执行的操作对象与控制功能。矩阵前两行对应智能家居设备，如电视机、智能门锁或电灯，这类设备可通过人机交互界面（Human Machine Interface, HMI）选择切换状态（例如开启/关闭、上调/下调）。矩阵末行设置了三类通话选项：紧急呼叫（即SOS求救号码）以及手机通讯录中的两位常用联系人。 本实验共评估了两种刺激范式。在第一种范式中，矩阵内的每个元素将依次单独高亮显示。该高亮过程随机执行，但所有元素的高亮总次数保持一致，因此每个元素被高亮的概率均为𝑝=1/9。此外，任意元素不得连续两次被高亮。在第二种范式中，并非单独高亮单个元素，而是依次高亮矩阵的每一行与每一列。该过程同样随机执行，且所有行与列的高亮总次数保持一致，因此任意时刻单个元素被高亮的概率为𝑝=2/6=1/3。 本次实验采用OpenBCI Cyton采集板记录脑电与眼电数据，采样频率为250Hz。针对脑电记录，由于实验目标为检测P300电位，输入通道的电极放置于Cz点位；反之，眼电记录的电极放置于Fp2点位，以检测与眨眼相关的脑电活动。两种记录模式共用参考电极与接地电极，分别置于右乳突（RM）与左耳廓（A1）。 本实验的受试群体共包含9名自愿参与本研究的志愿者，所有受试者均无听力与视觉障碍。本研究已获取所有受试者的知情同意，以使用其实验数据。实验在隔音室内开展，受试者被邀请坐在舒适的椅子上，将注意力集中于呈现刺激的15.6英寸笔记本电脑屏幕。每位受试者的记录会话分为四个独立实验，分别对应两种刺激范式与两种记录模式：逐元素范式+脑电信号、行列范式+脑电信号、逐元素范式+眼电信号，以及行列范式+眼电数据。每个实验共记录8个试次。 每位受试者的记录数据存储于一个单独的文件夹，命名格式为s[受试者编号]，例如受试者1的文件夹为s1。每个受试者文件夹内包含32个实验子文件夹，对应每一个独立实验。文件夹命名结构如下：subject_controlSignal_paradigm_date，例如s1P3001by1-12-11-2020_12-57-15。每个实验子文件夹内包含3个文件： 1. ExInfo：包含受试者信息、通道信息、实验范式以及本次实验的目标元素相关数据。 2. Raw-STIMS：包含本次实验的刺激呈现数据。需注意，每个特定刺激均由一组代码标识。在逐元素范式中，代码范围为101至109，分别对应3×3矩阵从左上角到右下角的9个元素；在行列范式中，代码范围为101至106，其中101至103依次对应矩阵从上到下的三行，104至106依次对应矩阵从左到右的三列。 3. Raw-EEG：包含通过Cyton采集板记录的脑电数据。 如需获取更多关于数据采集与数据集的详细信息，请参阅本研究的相关论文及配套开源代码： Sensors | Free Full-Text | Proposals and Comparisons from One-Sensor EEG and EOG Human-Machine Interfaces (mdpi.com) franciscolaport/Proposals-and-Comparisons-from-One-Sensor-EEG-EOG-Human-Machine-Interfaces (github.com)