Abstract
In the event-related potential (ERP) of scalp electroencephalography (EEG) studies, the vertex reference (Cz), linked mastoids or ears (LM), and average reference (AVG) are popular reference methods, and the reference electrode standardization technique (REST) is increasingly applied. Because scalp EEG recordings are considered as spatially degraded signals, independent component analysis (ICA) is a widely used data-driven method for obtaining ERPs by decomposing EEG data. However, the accurate estimation of the differences in ERP components extracted by ICA with different references remains unclear. In this study, we first provided formal descriptions of the above reference methods (Cz, LM, AVG, and REST) and ICA decomposition in ERP and then investigated the influences of different reference techniques on simulation and real EEG datasets. The results revealed that (1) the reference method did not change the peak amplitudes and latencies of relative ERPs corresponding to some IC time courses; (2) there were non-negligible effects of different reference methods on both temporal ERPs and spatial topographies of some ICs; and (3) compared to Cz, LM, and AR, considering both the performances of temporal ERPs and spatial topographies, the REST reference had overall superiority. These findings provide a recommended choice of REST for ICA analysis at the trial level and contribute to empirical investigations regarding the use of reference methods in ERP domains with ICA analysis.
Highlights
Since the first report in 1929 (Berger, 1929), scalp electroencephalography (EEG) has been a costeffective and non-invasive technique that directly quantifies the mean electrical activity of the brain at scalp sites with excellent temporal resolution (∼milliseconds) (Cohen, 2017)
ERP1 and ERP2 could be detected by independent component analysis (ICA) with different choices of references, and the spatial distribution patterns of the spatial components were similar for different references to some degree
One-way repeated ANOVA and post-hoc paired t-test were performed to quantify the performances of ICA time courses (RMSEs between ICA-extracted event-related potential (ERP) with average reference (AVG)/linked mastoids or ears (LM)/Cz and reference electrode standardization technique (REST) references) and topographies
Summary
Since the first report in 1929 (Berger, 1929), scalp electroencephalography (EEG) has been a costeffective and non-invasive technique that directly quantifies the mean electrical activity of the brain at scalp sites with excellent temporal resolution (∼milliseconds) (Cohen, 2017). A number of different reference methods have been proposed, including the vertex reference (Cz) (Lehmann et al, 1998), linked mastoids or ears (LM) (Gevins and Smith, 2000), average reference (AVG) (Offner, 1950) and reference electrode standardization technique (REST) (Yao, 2001; Dong et al, 2017). Under their certain assumptions, reference methods including Cz, LM, and AVG have been commonly used while ignoring that they are not zero references. At present, the choice of appropriate EEG references in EEG studies across the world remains an open issue, especially in the ERP domain (Luck, 2014)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
