Out and about: Subsequent memory effect captured in a natural outdoor environment with smartphone EEG

Abstract

Spatiotemporal context plays an important role in episodic memory. While temporal context effects have been frequently studied in the laboratory, ecologically valid spatial context manipulations are difficult to implement in stationary conditions. We investigated whether the neural correlates of successful encoding (subsequent memory effect) can be captured in a real‐world environment. An off‐the‐shelf Android smartphone was used for wireless mobile EEG acquisition and stimulus presentation. Participants encoded single words, each of which was presented at a different location on a university campus. Locations were approximately 10–12 m away from each other, half of them with striking features (landmarks) nearby. We predicted landmarks would improve recall performance. After a first free recall task of verbal stimuli indoors, participants performed a subsequent recall outdoors, in which words and locations were recalled. As predicted, significantly more words presented at landmark locations as well as significantly more landmark than nonlandmark locations were recalled. ERP analysis yielded a larger posterior positive deflection during encoding for hits compared to misses in the 400–800 ms interval. Likewise, time‐frequency analysis revealed a significant difference during encoding for hits compared to misses in the form of stronger alpha (200–300 ms) and theta (300–400 ms) power increases. Our results confirm that a vibrant spatial context is beneficial in episodic memory processing and that the underlying neural correlates can be captured with unobtrusive smartphone EEG technology. The advent of mobile EEG technology promises to unveil the relevance of natural physical activity and natural environments on memory.

Mobile EEG in research on neurodevelopmental disorders: Opportunities and challenges

“…Although mobile EEG research is in its infancy, this technology has already been used by cognitivists to study a variety of processes such as attention ( Jungnickel and Gramann, 2016 ), memory ( Griffiths et al, 2016 ; Piñeyro Salvidegoitia et al, 2019 ), spatial cognition ( Ehinger et al, 2014 ), speech/auditory processing ( Callan et al, 2015 ; Mirkovic et al, 2016 ) and motor processing ( Lin et al, 2016 ; Lo et al, 2016 ; Wong et al, 2014 ). Mobile EEG has also been used for everyday applications in sports ( Park et al, 2015 ), urban behaviours ( Mavros et al, 2016 ), emotion recognition ( Aspinall et al, 2015 ; Bercik et al, 2016 ; Li et al, 2015 ), neurofeedback ( Stopczynski et al, 2014b ), motor rehabilitation ( Kranczioch et al, 2014 ; Wagner et al, 2012 ), epilepsy ( Askamp and van Putten, 2014 ), and cognitive impairment ( Kashefpoor et al, 2016 ).…”

“…2 Illustration of innovative EEG-based paradigms facilitated by ongoing mobile EEG developments. Examples include a set-up a) for mobile brain/body imaging (MoBI) allowing for body movements merging EEG with other motion-based sensors, adapted from Gramann et al (2010) ; b) using mobile EEG while walking with smartphone-based stimuli presentation (of word stimuli) in an outdoor setting (i.e., a pre-specified route), adapted from Piñeyro Salvidegoitia et al (2019) ; c) using mobile EEG simultaneously in multiple individuals within an indoor social setting (i.e., in the classroom), adapted from Dikker et al (2017) . …”

“…One type of study shows that mobile systems are able to replicate findings derived from traditional EEG (providing support for construct validity), such as capturing several expected ERPs (N2, P3 and reward-positivity) during well-established lab-based stationary cognitive paradigms ( Krigolson et al, 2017 ). Expected frequency-based features can also be extracted, for example, reduction in theta power and increase in alpha power from eyes-closed to eyes-open resting conditions ( Debener et al, 2015 ), even in a single-channel dry system ( Johnstone et al, 2012 ), as well as event-related decrease in beta power for successful memory retrieval in a outdoor walking-based paradigm ( Griffiths et al, 2016 ; Piñeyro Salvidegoitia et al, 2019 ). Nevertheless, in a recent study only a research-grade gel-based system (e.g., compared to dry EEG) was able to simultaneously capture several expected EEG-based patterns ( Radüntz, 2018 ; see Fig.…”

Challenge Accepted? Individual Performance Gains for Motor Imagery Practice with Humanoid Robotic EEG Neurofeedback

“…After the training block was finished, EEG data were high-pass filtered at 8 Hz (finite impulse response, filter order 826) and subsequently low-pass filtered at 30 Hz (finite impulse response, filter order 220) using EEGLAB toolbox Version 12.0.2.6b for MATLAB (Version: 2011B). This filter range was set to encompass the sensorimotor rhythms mu (8-12 Hz) and beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), to which the neural correlate of interest, the event-related desynchronization (ERD), is highly specific [18,19]. Epochs were extracted from -5 to 9 s relative to MI onset, and segments containing artifacts were rejected (EEGLAB function pop_jointprob.m, SD = 3).…”

“…power changes of rhythmic brain activity captured over sensorimotor areas within the mu (8)(9)(10)(11)(12) and beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) frequency range, are prominent neural correlates of MI to control EEG-based NF and BCI paradigms [17][18][19]. EEG is currently the target technique of choice of many BCIs and NFs, because aside from well-established laboratory setups, it can also be recorded using small, low-cost, unobtrusive, and wireless hardware in everyday life [20,21]. This mobile EEG technology allows, for instance, to apply NF and BCI long-term at the user's home [3,22].…”