China
Africa
Explore chapters and articles related to this topic
Functional Neuroimaging of the Central Auditory System
Published in Stavros Hatzopoulos, Andrea Ciorba, Mark Krumm, Advances in Audiology and Hearing Science, 2020
David L. McPherson, Richard Harris, David Sorensen
The above discussion serves as an example of the power and usefulness of this technique. The reader can obtain public domain software and tutorials for analysis from Brainstorm (http://neuroimage.usc.edu/brainstorm/Introduction) a collaborative public domain project that provides open-source applications and tutorials. Likewise, the Schwartz Center for Computational Neuroscience, University of California, San Diego (https://sccn.ucsd.edu/wiki/EEGLAB) has open-source software and tutorials. Since the above software requires MATLAB (MathWorks), it is necessary to purchase the base software either for personal use (https://www.mathworks.com/pricing-licensing.html?prodcode=ML&intendeduse=home) or, if a student, the student license (https://www.mathworks.com/pricing-licensing.html?prodcode=ML&intendeduse=student).
A Step-by-Step Tutorial for a Motor Imagery–Based BCI
Published in Chang S. Nam, Anton Nijholt, Fabien Lotte, Brain–Computer Interfaces Handbook, 2018
Hohyun Cho, Minkyu Ahn, Moonyoung Kwon, Sung Chan Jun
After recording MI data, we had five files (*.dat) from the BCI2000 system because we recorded five runs, as shown in Table 23.1. BCI2000 provides a MATLAB® code, “BCI2000import.m,” to convert each *.dat file to the EEGlab format of MATLAB. EEGlab is a well-known EEG signal processing toolbox (Delorme and Makeig 2004). In the EEGlab data structure, an “event” variable contains trigger information. The data length of each “event” variable is the same with respect to the number of stimuli. The “event” variable includes three variables: latency, position, and type. The “latency” variable contains the time point value of a triggered stimulus within the run data. The “position” variable indicates the position of the stimulus array of the configuration in BCI2000. This value can be considered a class label, because the first and second values in our MI experiment are left- and right-hand MI instructions, as shown in Figure 23.2. Last, the “type” variable indicates the type of stimulus code. In an online experiment, there are many types of stimulus codes in BCI2000, for example, the stimulus, target, and feedback codes, and so on; then, we have raw EEG and trigger information, including latency and stimulus labels. We can extract each trial data from five runs of data. Depending on the latency of each stimulus (onset), the data frame extracted was between −2000 and 5000 ms. We carried out the same procedure for non–task-related data as well.
Event-Related Potentials to Speech Relate to Speech Sound Production and Language in Young Children
Published in Developmental Neuropsychology, 2022
Vanessa Harwood, Jonathan Preston, Alisa Baron, Daniel Kleinman, Nicole Landi
EEG data were analyzed using the EEGLAB v2019.1 toolbox (Delorme & Makeig, 2004) except as noted below. As data from some participants were contaminated with line noise, notch filters were applied at line frequencies (60, 120, 180, and 240 Hz; order = 180). Next, the PREP pipeline (Bigdely-Shamlo et al., 2015) was used to identify electrodes that were bad throughout data collection (all active electrodes were considered candidates for replacement except for channels 125–128, which are located below and next to the eyes, and were never replaced at this stage); these electrodes were replaced using spherical spline interpolation (Perrin, Pernier, Bertrand, & Echallier, 1989); the mean number of electrodes replaced was 6.00 SD = 3.67, range = [0–14]). Data were band-pass filtered from 0.3 to 30 Hz (Butterworth filter, 12 db./oct roll-off); re-referenced to the average reference (the vertex reference, Cz, was used during recording); segmented into 700 ms epochs including a 100 ms pre-stimulus baseline and a 600 ms post-stimulus interval; and baseline-corrected using the mean of the pre-stimulus window (Junghofer, Elbert, Tucker, & Braun, 1999). The horizontal electrooculogram (HEOG) was measured as the difference between channels 125 and 128, and the vertical electrooculogram (VEOG) was represented by four pairwise differences between channels 126 and 8, 126 and 14, 127 and 21, and 127 and 25.
Longitudinal changes in blood-based biomarkers in chronic moderate to severe traumatic brain injury: preliminary findings
Published in Brain Injury, 2021
Caroline Schnakers, James Divine, Micah A. Johnson, Evan Lutkenhoff, Martin M. Monti, Katrina M. Keil, John Guthrie, Nader Pouratian, David Patterson, Gary Jensen, Vanessa C. Morales, Kathleen F. Weaver, Emily R. Rosario
EEG data were analyzed using the GUI (graphic user interface) of EEGLAB (version 14.0.0b). Preprocessing included channel location, band-pass filtered between 1 and 40 Hz (FIR filter), Independent Component Analyses (ICA; with jader decomposition algorithm) to identify and exclude ocular and motor artifacts related components (maximum of 3), and a final visual inspection of the EEG data in order to manually remove any residual artifacts. For the spectral analysis, the continuous signal of each recording was segmented into 1s epochs. Each epoch was fast Fourier transformed (FFT) with a Hanning-tapered window. Subsequently, all epochs were averaged, and the mean of the power spectral density in different frequency bands were exported for statistical analysis. Frequency bands were chosen at target electrode sites based on previous literature pinpointing sites of maximal amplitude for each band: alpha (8–13 Hz) at the occipital electrodes (O1, Oz, O2), theta (4–7 Hz) at the frontal electrodes (F7, F3, Fz, F4, F8), and delta (1–3 Hz) at the central electrodes (C3, Cz, C4). The averaged power spectral density extracted for each site (F, C, O) was expressed in absolute power (μV2) (22).
An Electrophysiological Study of Aging and Perceptual Letter-Matching
Published in Experimental Aging Research, 2021
Peter R. Mallik, Philip A. Allen, Mei-Ching Lien, Elliott Jardin, Michelle L. Houston, James R. Houston, Brianna K. Jurosic
After the hand removal of obvious artifacts, the data were analyzed using an independent component analysis (ICA) which uses specific algorithms to remove artifacts from actual EEG signal. After ICA identified components for removal, the data were epoched. This process involved breaking down data into specific “bins” for the averaging process, which is used to take EEG data and average it into ERP data. A baseline of −200 to 0 ms prior to the probe stimulus was established making everything prior to the probe have a microvolt value of 0. The epoch ranged in total from −200 ms prior to the probe stimulus presentation to 1500 ms post-probe stimulus presentation. Data went through artifact rejection using the epoched data. Data were tested for abnormal values, trends, spectra, and abnormal distribution using the EEGLAB artifact rejection tools. Once cleaned, P300 data were collected from the averaged EEG data from PZ.