ERPLAB Studio: Basic Processing Steps - ucdavis/erplab GitHub Wiki

This page provides an outline of the typical sequence of operations for analyzing an ERP experiment using ERPLAB Studio. You can find a more detailed discussion of the sequence of processing steps in (Appendix 3 of Applied ERP Data Analysis.

Make sure you have read our document on Important Background Concepts and Data Structures before reading this page.

1. Import an EEG file by clicking the Import button in the EEGsets panel of the EEG tab. This creates a dataset (also known as an EEGset). The original EEG file can come from one of the many different commercial EEG recording systems supported by EEGLAB (including Neuroscan, EGI, and Biosemi; see the EEGLAB documentation for the supported file types). You should now be able to see the EEG in the plotting region of the EEG tab.
2. Add an EventList structure to the current dataset using the EventList panel. This structure stores information about ERPLAB's event processing and bin assignments. Important note: ERPLAB currently requires numeric event codes, and if your system uses text labels (e.g., 'S23') you must convert them into numeric codes at this point.
3. Preprocess the continuous EEG. There are many relevant panels for preprocessing, including Filtering, Channel Operations (for re-referencing the data), Interpolate Channels, Edit/Delete Channels & Locations (to add channel location coordinates), and Shift Event Codes (to account for stimulus delays).
4. Perform ICA-based artifact correction using the EEGLAB ICA panel. See the Tutorial page on Artifact Correction for details.
5. Assign events to bins using BINLISTER. This routine takes each event code in your EEG data, determine which bin (if any) that event code belongs to, and store this information in the EventList structure. For example, you can specify that Bin 3 will consist of targets that were preceded by a nontarget and followed 200-1500 ms later by a left-hand button-press response. This step can also be configured to extract reaction times (see our Behavioral Analyses documentation for details).
6. Convert the continuous EEG data in the dataset into a set of fixed-length epochs (e.g., from 200 ms before stimulus onset until 800 ms after stimulus onset) using the Extract Bin-Based Epochs panel. This creates an epoched dataset, in which time zero is the event code. Baseline correction is usually applied at this point, but you can re-baseline the data later using the Baseline Correction & Linear Detrend panel.
7. Determine which epochs contain artifacts using the Artifact Detection panel. Epochs with artifacts are not deleted; they are simply "marked" so that they can be excluded during the averaging process. Epochs can be marked using either a set of automated algorithms or by visual inspection.
8. Average together the epochs that have been assigned to each bin using the Compute Averaged ERPs panel. The averaged data are stored in a new structure called an ERPset, which will be visible in the ERP tab. Note that the ERPset also contains a copy of the EVENTLIST structure from the dataset that was used to create the ERP structure. This step also creates a set of data quality metrics.
9. Note that a history of processing steps is available in the History panel. This panel stores the script equivalents of the processing steps, which can help you create a script that recreates the steps you carried out in the GUI.

9. The ERP tab allows you to view and process ERPsets. In addition to the plotting region within the main ERP tab, you can also launch the Advanced Waveform Viewer (from the Plotting Options popup menu at the bottom right corner of the plotting region). This tool gives you much more control over plotting, allowing you to make publication-quality plots, which can be exported in a variety of formats.
10. You can also plot scalp maps using the Plot Scalp Maps panel, and you can view the data quality metrics with the View Data Quality Metrics panel.
11. Common processing steps can be carried out using the Filtering panel, the Channel Operations panel (to re-reference the data or create new channels that average across a cluster of the original channels), and the Bin Operations panel (to average bins together or make difference waves).
12. You can make a grand average of the data in multiple ERPsets (typically coming from different participants) by loading the ERPsets and using the Average Across ERPsets panel.
13. Once you have processed the data from all the participants, you can use the Measurement Tool panel to quantify the amplitudes and latencies from the individual ERPsets. These can then be saved as text files in formats suitable for importing into spreadsheets and statistics programs. Alternatively, you can use the Mass Univariate Toolbox, which can directly read the ERPsets created by ERPLAB.
14. Note that a history of processing steps is available in the History panel. This panel stores the script equivalents of the processing steps, which can help you create a script that recreates the steps you carried out in the GUI.