A validation of Emotiv EPOC Flex saline for EEG and ERP research

Abstract

Background Previous work has validated consumer-grade electroencephalography (EEG) systems for use in research. Systems in this class are cost-effective and easy to set up and can facilitate neuroscience outside of the laboratory. The aim of the current study was to determine if a new consumer-grade system, the Emotiv EPOC Saline Flex, was capable of capturing research-quality data. Method The Emotiv system was used simultaneously with a research-grade EEG system, Neuroscan Synamps2, to collect EEG data across 16 channels during five well-established paradigms: (1) a mismatch negativity (MMN) paradigm that involved a passive listening task in which rare deviant (1,500 Hz) tones were interspersed amongst frequent standard tones (1,000 Hz), with instructions to ignore the tones while watching a silent movie; (2) a P300 paradigm that involved an active listening task in which participants were asked to count rare deviant tones presented amongst frequent standard tones; (3) an N170 paradigm in which participants were shown images of faces and watches and asked to indicate whether the images were upright or inverted; (4) a steady-state visual evoked potential (SSVEP) paradigm in which participants passively viewed a flickering screen (15 Hz) for 2 min; and (5) a resting state paradigm in which participants sat quietly with their eyes open and then closed for 3 min each. Results The MMN components and P300 peaks were equivalent between the two systems (BF10 = 0.25 and BF10 = 0.26, respectively), with high intraclass correlations (ICCs) between the ERP waveforms (>0.81). Although the N170 peak values recorded by the two systems were different (BF10 = 35.88), ICCs demonstrated that the N170 ERP waveforms were strongly correlated over the right hemisphere (P8; 0.87–0.97), and moderately-to-strongly correlated over the left hemisphere (P7; 0.52–0.84). For the SSVEP, the signal-to-noise ratio (SNR) was larger for Neuroscan than Emotiv EPOC Flex (19.94 vs. 8.98, BF10 = 51,764), but SNR z-scores indicated a significant brain response at the stimulus frequency for both Neuroscan (z = 12.47) and Flex (z = 11.22). In the resting state task, both systems measured similar alpha power (BF10 = 0.28) and higher alpha power when the eyes were closed than open (BF10 = 32.27). Conclusions The saline version of the Emotiv EPOC Flex captures data similar to that of a research-grade EEG system. It can be used to measure reliable auditory and visual research-quality ERPs. In addition, it can index SSVEP signatures and is sensitive to changes in alpha oscillations.

It’s all about time: precision and accuracy of Emotiv event-marking for ERP research

“…Similarly, Emotiv’s EPOC system was found to measure research-grade auditory ERPs in adults ( Badcock et al, 2013 ) and children ( Badcock et al, 2015 ); as well as visual ERPs in response to faces ( De Lissa et al, 2015 ). Recently, Williams et al (2020) found analogous results for the Emotiv EPOC Flex system. The fact that EEG systems in this class cost a fraction of the price of research systems makes them an appealing alternative to researchers for ‘acquiring research-grade ERPs on a shoestring budget’ ( Barham et al, 2017 ).…”

Comparison of Smoothing Filters’ Influence on Quality of Data Recorded with the Emotiv EPOC Flex Brain–Computer Interface Headset during Audio Stimulation

“…This customer-grade EEG with “Ag” sensors allows recording signals and is mainly dedicated to academic research, however, these signals are not totally raw, but preprocessed and do not contain full information [ 7 , 15 ]. Some sources, however, claimed Emotiv Flex to provide clinical quality of data [ 36 ].…”

CNN‐Based Personal Identification System Using Resting State Electroencephalography

“…In order to evaluate whether the proposed system works well on these commercial EEG sensors, a series of experiments are conducted using 14, 32, and 64 channels, respectively. The selected 14-channel and 32-channel are based on the EMOTIV EPOC X 14 Channel Mobile Brainwear® [ 10 ] and EMOTIV EPOC Flex EEG Brainwear® system [ 36 ]. Note that the 14 channels of EMOTIV EPOC X (AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, and AF4) are all contained in the original 64 channels of PhysioNet dataset.…”