Fnhum 10 00694

REVIEW
published: 12 January 2017

doi: 10.3389/fnhum.2016.00694
Understanding Minds in Real-World

Environments: Toward a Mobile
Cognition Approach
Simon Ladouce, David I. Donaldson, Paul A. Dudchenko and Magdalena Ietswaart *
Psychology, Faculty of Natural Sciences, University of Stirling, Stirling, UK
There is a growing body of evidence that important aspects of human cognition have
been marginalized, or overlooked, by traditional cognitive science. In particular, the use
of laboratory-based experiments in which stimuli are artificial, and response options are
fixed, inevitably results in findings that are less ecologically valid in relation to real-world
behavior. In the present review we highlight the opportunities provided by a range of
new mobile technologies that allow traditionally lab-bound measurements to now be
collected during natural interactions with the world. We begin by outlining the theoretical
support that mobile approaches receive from the development of embodied accounts
of cognition, and we review the widening evidence that illustrates the importance of
examining cognitive processes in their context. As we acknowledge, in practice, the
development of mobile approaches brings with it fresh challenges, and will undoubtedly
require innovation in paradigm design and analysis. If successful, however, the mobile
Edited by:
cognition approach will offer novel insights in a range of areas, including understanding
Klaus Gramann, the cognitive processes underlying navigation through space and the role of attention
Technical University of Berlin, during natural behavior. We argue that the development of real-world mobile cognition
Germany
offers both increased ecological validity, and the opportunity to examine the interactions
Reviewed by:
Martin Georg Bleichner, between perception, cognition and action—rather than examining each in isolation.
University of Oldenburg, Germany
Keywords: mobile brain imaging, cognitive neuroscience, ecological validity, EEG, embodiment, situated cognition
Rob Zink,
KU Leuven, Belgium
Bojana Mirkovic contributed to the A RATIONALE FOR MOBILE REAL-WORLD COGNITION

review of Martin Georg Bleichner
*Correspondence: The human mind is a dynamic predictor that perceives, understands and acts within complex
Magdalena Ietswaart and ever-changing environments. To produce flexible and adaptive reactions that are relevant and
magdalena.ietswaart@stir.ac.uk appropriate to the individual’s goals, the brain must integrate concurrent multi-modal sensory and
motor signals, using continuous real-time feedback to guide the execution of on-going behavior.
Received: 15 September 2016 Despite this dynamic reality, however, the traditional approach to understanding human cognition
Accepted: 29 December 2016 has been the collection of empirical findings from experiments taking place in relatively static,
Published: 12 January 2017 often simulated, laboratory settings. Typically, participants sit or lie down, are given explicit and
Citation: highly constrained instructions, and are required to attend to artificial stimuli whilst performing
Ladouce S, Donaldson DI, deliberately stereotyped responses. The strength of such an approach is the experimental control
Dudchenko PA and Ietswaart M
it affords; the cost, however, is a loss of real-world dimensionality, and perhaps, relevance. In
(2017) Understanding Minds in
Real-World Environments: Toward a
the current article we present an alternative approach that capitalizes on recent technological
Mobile Cognition Approach. developments to allow experimental work to be situated in the real world. This mobile cognition
Front. Hum. Neurosci. 10:694. approach capitalizes on the ability to record brain activity (e.g., EEG) and body dynamics
doi: 10.3389/fnhum.2016.00694 (e.g., eye movements) concurrent to natural behaviors. We believe that the emerging field of
Frontiers in Human Neuroscience | www.frontiersin.org 1 January 2017 | Volume 10 | Article 694

Ladouce et al. Toward Mobile Cognition
mobile cognition offers significant added value to traditional pursuit of internal validity. Attempts to ensure that cognitive
laboratory science, with particular implications for the phenomena under investigation are being measured accurately
translation of theoretical knowledge into impact. Before and precisely (i.e., with high internal validity) have resulted
outlining this view, we first highlight the rationale for mobile in an ever-increasing drive toward greater resolution of data
cognition, which stems at least in part from dissatisfaction acquisition (e.g., greater number of electrodes during EEG
with assumptions underlying laboratory experiments: namely, recording, or finer-grained assessment of where a participant
that behavioral and neurobiological measures recorded under looks during eye-tracking). Furthermore, the concern for internal
strictly controlled laboratory conditions accurately reflect the validity has contributed to the desire for ever improved
complexity of cognitive processing. signal to noise ratios during measurement and the removal of
potential confounding variables via more and more artificial
experimental set ups (Schmuckler, 2001). Equally, individual
THE ISSUE OF ECOLOGICAL VALIDITY cognitive functions have gradually been studied at a greater and
greater level of detail (i.e., specificity), with distinct processes
Psychology has made substantial progress through the use of being further and further subdivided in to sub processes of sub
laboratory based experimentation; such has been the success processes (e.g., declarative memory dividing into episodic and
of this largely reductionist approach that concerns have been semantic memory, episodic memory dividing into recollection
expressed about it crowding out other fields within Cognitive and familiarity, recollection into rate and precision, etc.). Whilst
Science (Gentner, 2010). However, despite its acknowledged the push for greater internal validity is warranted, the increasing
success, right from the very beginning of the psychological abstraction, isolation and focussing of measurements have,
investigation of cognition, concerns were raised about how inevitably perhaps, contributed to an unintended reduction in
ecologically valid many of the findings were. For example, ecological validity.
as early as 1943, Brunswik expressed concern that cognitive Of course, mobile cognition is not to replace laboratory work
psychology was heading toward the study of narrow and and the two must work in parallel and actively inform each other.
artificially isolated conditions that were not representative of While mobile techniques offer the unprecedented opportunity
the actual functioning of cognition (Brunswik, 1943). Similarly, to investigate cognition in real-world context, they however
significant debate around ecological validity was raised in the do not currently compete with lab-based counterparts in terms
1970s. Perhaps the most well-known advocate of this concern is of qualitative and quantitative features. The spatial resolution
Ulric Neisser, who argued (e.g., see Neisser, 1976) that assessing offered by a fMRI scanner will never be matched by a mobile
cognitive processes in an artificial environment would only fNIRS or EEG system. Furthermore, some research questions are
enhance our understanding of those specific circumstances—but better served by the experimental control afforded in laboratory
not necessarily generalize to real-world cognition. research. Our view is that the mobile cognition approach will
More pointedly, Bronfenbrenner (1977) added that measuring add considerable value to existing laboratory based cognitive
restricted responses in artificial setups would generate behaviors neuroscience: indeed, there are many research questions that can
that are in fact different from the behaviors displayed in a only really be sensibly addressed in real-world contexts.
natural context. By this view, having participants sit at a
computer looking at pixelated images or scenes may not allow
researchers to fully characterize cognition—because the processes COGNITION IS EMBODIED
being engaged, or representations being accessed, occur in far
more complex forms in real life. For example, researchers The lack of ecological validity in cognitive science has become
interested in recognition (as per witness identification scenarios even more of an issue with the emergence of evidence that
in real life) typically present photographic images of people on cognition is embodied (for reviews, see Gallagher, 2005; Barsalou,
computer screens—which are inevitably less rich than interacting 2008). Over the past 20 years it has become apparent that
with a real person. A striking example of the implications of cognition is inherently reliant on its situated position in the
this restriction can be found in the clinical case of agnosic environment. Chiel and Beer (1997) were one of the first to
patients (e.g., Goodale et al., 1991), who despite being unable argue that understanding the interactions between brain, body
to recognize or even describe the features (size, color, shape) of and environment is crucial. The significance of this view has
objects presented visually, are nonetheless able to appropriately become more apparent through recent research that shows
adjust their movement toward these objects when interacting that bodily experience in fact shapes the way we process the
with them. This finding provided evidence that the internal environment. To be clear, the broad idea of interdependence
representations associated with objects are accessed differently between perception/action and the environment had already
depending on the purpose of the output (vision for recognition been emphasized by Gibson in his theory of affordances (1979),
vs. vision for action). A significant advantage of the mobile which states that the external information available to us
cognition approach is that it encourages researchers to investigate is processed in relation to the opportunities for action that
cognition in context, in relation to our natural interactions within they provide. Although it was developed largely independently,
the environment, rather than abstracted away from it. embodiment theory also argues that cognition is for action, and
Paradoxically, to our reading, one significant contribution to furthermore that cognition is actually dependent on the bodily
the problem of ecological validity has been the over enthusiastic experience (e.g., Clark, 1999). Understanding cognition through

abstract experimental paradigms that lack interaction with the act on (output) the environment (see Figure 1B). By assessing
environment, due to their artificial nature, therefore makes little only one aspect in isolation, as previously done in cognitive
sense from an embodied perspective. research, the dynamic interplay between cognitive functions
Support for the embodiment of cognition can also be found in cannot be captured. Ultimately, therefore, it will be necessary
findings from modern day neuroscience. For example, evidence to integrate concurrent measurement of input and output if we
regarding neural plasticity collected over the last 20 years has wish to account for the dynamic interplay of cognitive functions
made clear that brain and behavior are constantly shaped by in the face of a rich and dynamic real-world environment (see
our experience in interacting with the world. Even at the Figure 1C).
level of the functional organization of vision, it is clear that
representations reflect top-down action potentialities rather than
bottom-up sensory inputs (Bracci et al., 2012). One compelling BRAIN STATES DIFFER DURING
example is provided by Thaler et al. (2011), who have shown MOVEMENT
that blind people can use echolocation to navigate in space
(listening to the echo of clicks-sound to locate the reflection A growing animal literature demonstrates the fundamental
point), giving rise to functional restructuring of brain regions importance of interactions between brain, body and
typically involved in vision. Perhaps the most famous example of environment. For example, in the mouse hippocampus,
how our behavior shapes our brain is the work on taxi drivers 75% of place cells (neurons which encode specific locations
presenting with increased hippocampal size, which is thought in the animal’s environment) show a significant decrease in
to be due to their experience of navigating the maze of London firing when the mouse is prevented from moving (Chen et al.,
streets (Maguire et al., 2000). Taken together, these finding 2013). Further, in rats, changes in theta power associated with
highlight the fact that interaction with the surrounding world ambulation (McFarland et al., 1975; Long et al., 2014) are also
shapes brain structure and function. From this perspective, recent modulated by the anticipation and initiation of goal-directed
acceptance of cognition as being embodied emphasizes a problem instrumental behavior (Wyble et al., 2004; Sinnamon, 2006).
inherent to traditional cognitive experiments—highly controlled In monkeys, the hippocampus is essential for a task where the
and artificial experimental testing tends to separate cognition animal must walk to a to-be-remembered location (Hampton
from natural bodily experiences. Furthermore, by separating et al., 2004), but it is not needed when the same type of memory
cognition from the bodily experience we remove their intrinsic task is performed while the animal is seated (Malkova and
interaction that defines the human mind as an active agent. Mishkin, 2003). Importantly, the inter-dependence between
The concept of situated cognition, stemming from the locomotion and cognition is bi-directional: brain dynamics in
embodiment framework, postulates that our cognitive experience Drosophila suggest that the processing of visual information
is dependent on the body’s position in the environment. This is different in flight compared to resting state (Maimon et al.,
interdependence implies that when embodied agents actualize 2010). To be clear, existing data reveals an inter-dependence
their intentions they have to ensure their behavior accommodates between cognition and the exploration of the environment that
the contingencies of the environment. In contrast, by looking emphasizes the importance of understanding cognition in real-
at cognition through an ever-narrower lens, cognitive science world contexts (suggesting that even studies in virtual reality will
has focused on the study of specific cognitive processes in be insufficient to characterize cognitive processes as they support
isolation. Consequently, the complex interplay of perception, everyday behavior). Strikingly, emerging findings support the
cognition and action has not generally been investigated (Beer, claims made by Bronfenbrenner (1977) that cognitive responses
2000). The isolated cognitive account of classic input-output measured in artificial experimental conditions are different from
models is clearly represented by the idea of minds as machines, those in natural exploration, and furthermore that ecological
decoding sensory inputs (perception) to then deliver output validity is not just desirable but essential if we are to fully
commands (behavior) to the body (see Figure 1A). Whilst understanding cognition.
the unilateral direction from input to output implies a major Whilst existing evidence clearly demonstrates that the neural
bottom-up influence (and questions the very existence of self- correlates of visual perception or locomotion can be altered
motivated behavior), there is now abundant evidence of top- by context, one reading of these data is that they only argue
down and dynamic influences on the selection and processing for sensitivity in sensory or motor systems, rather than in
of features relevant to the ongoing task (Henderson, 2007). association cortex linked to higher order cognitive processing.
For example, the execution of naturalistic goal-oriented tasks However, the dual-task literature demonstrates that introduction
induces anticipatory eye movements toward relevant affordances of gait and balance control has a significant interfering effect on
(e.g., Pelz and Canosa, 2001; Mennie et al., 2007; Hayhoe et al., higher-order cognitive processes such as executive functions (i.e.,
2012). Another example of this dynamic interplay between inhibition, divided attention), verbal fluency, decision-making
input and output is the integration of visuomotor feedback, and working memory (for a review, see Al-Yahya et al., 2011).
allowing for online correction of movements to sudden changes As well as linking bodily changes to cognitive performance, these
of the environment (e.g., adaptive reaching movements to avoid data also question the validity of static single-task experiments in
obstacles; Chapman and Goodale, 2010). This body of evidence particular, because everyday life necessarily involves considerable
emphasizes that perception and action are interdependent and cognitive-motor multitasking. In short, embodied cognition
that bodily experience influences the way we process (input) and theory argues that sensory, motor and cognitive processing are

FIGURE 1 | Development of psychological models about the interactions between environment, body and central nervous system. Early simple
input-output models (A) ignore the environment and represent sensory inputs as being processed discretely to produce motor outputs. More sophisticated
dynamic-interactive models (B) include environmental factors. By this account, direct feedback from motor output can interact with, and act on, the
environment—resulting in a change to future sensory inputs. In embodied-situated models (C) the nervous system is embedded within the environment through the
body. From this perspective input and output systems are integrated rather than discrete separable elements, and the nervous system is inherently linked with the
environment—as parts of a dynamic system. Adapted from Chiel and Beer (1997).
interdependent, and evidence supports this claim, providing key developments that illustrate the changing landscape of tools
a strong rationale for studying cognition in motion or while with which cognition can be measured.
engaged in natural motor tasks (Schaefer, 2014). The last decade has seen the emergence of compact,
lightweight, non-invasive and wireless brain imaging hardware
that do not hinder everyday movements, and yet still provides
THE EMERGENCE OF MOBILE METHODS accurate recordings of brain cortical dynamics. Indeed, mobile
iterations of electroencephalography (EEG) and near-infrared
Despite long standing awareness of the issue of ecological spectroscopy (fNIRS) systems have evolved rapidly to closely
validity, and the recent accumulation of evidence highlighting the match the standards of high-density laboratory versions in both
need for a real-world approach in cognitive research (e.g., Clark, spatial and temporal resolution (Gramann et al., 2014b). More
1997; Smilek et al., 2006; Williams and Long, 2015), the issue has than just recording of brain activity in motion, mobile techniques
not been widely addressed in practice. Indeed, on the whole, these allow brain processes to be captured “on the go,” in relation to
concerns are only incidentally acknowledged in the literature natural behaviors in real-world environments (such as navigating
(Sbordone, 1996; Burgess et al., 1998; Chaytor and Schmitter- the streets of London either on foot or as a taxi driver, choosing
Edgecombe, 2003; Spooner and Pachana, 2006; Williams and products in a shop, putting a golf ball, or an elderly person
Long, 2015). At least in part, this inertia can be explained by a getting up and moving about). Importantly, early proof-of-
lack of satisfactory solutions to conduct ecologically valid studies concept studies have reported successful recording of classic
while maintaining scientific rigor and high levels of data quality. EEG components during motion (Gramann et al., 2010; Debener
Studying cognition in the real-world demands a combination of et al., 2012; Severens et al., 2012) and the detection of task-
technical and methodological requirements (Makeig et al., 2009; related changes in hemoglobin concentration using wearable
Reis et al., 2014), including portable devices that can operate with fNIRS (Koenraadt et al., 2014; Piper et al., 2014).
minimal noise, whilst also developing paradigms that retain an While a range of mobile imaging techniques exists, it is
adequate degree of experimental control (see Figure 2). important to recognize that they each offer particular strengths.
Recent technological developments have led to the Multi-channel fNIRS offers a good spatial resolution over a
advancement of portability of traditional brain imaging delimited cortical surface, describing neuro-vascular changes
and behavioral measurement techniques. Research techniques at a cortical level, but this technique lacks the temporal
that were previously restricted to laboratory settings due to resolution required to investigate fast cognitive processes. In
hardware limitations (e.g., weight, size, battery life) have become contrast, the high temporal resolution of EEG can reveal rapid
fully portable. Whilst it is beyond the scope of the current article changes in electro-cortical activity related to the ever-changing
to exhaustively review available equipment (for a comprehensive demands of real-world cognition. An additional advantage
review see Gramann et al., 2011), here we highlight a number of of mobile EEG as a neuroimaging tool is that EEG has

FIGURE 2 | An example of the differences between a laboratory-based (left) and a mobile real-world (right) experimental setup using EEG. (1) EEG
sensors, (2) Amplifier and data storage unit, (3) Stimulus presentation. Using the example of a classic face recognition paradigm, this figure illustrates the typical
laboratory setup (left) in contrast to recording of real-world face recognition (right). In the latter, faces are presented in-context, while the participants are behaving
naturally experiencing a real-world environment. Note that event triggers are also implemented differently, i.e., based on computer controlled timing of stimulus
presentation or on behavioral response in the laboratory, in contrast to event registration based on natural behavior in relation to stimuli in the scene (e.g., as assessed
by fixation points recorded with a head mounted eye tracking device). Mobile brain imaging of neural activity with behavioral measurements permits the study of
cognition underlying everyday life.
been used extensively in laboratory-based settings, providing a to upright walking through the integration of simultaneously
background of information against which mobile cognition data recorded multi-modal data streams (Bulea et al., 2013). By using
can be benchmarked. Nevertheless, as discussed below under gait dynamics such as heel strikes to time-lock continuous EEG
“current challenges,” there are considerable methodological recordings, Gwin et al. (2011) have reported increased power
issues associated with the application of mobile brain imaging— spectral activity in the left and right sensorimotor cortex during
and an important element of current research is to overcome contralateral foot suspension in subjects walking at a steady pace
these remaining issues and demonstrate the viability of mobile on a treadmill, suggesting increased cortical involvement related
neuroimaging. to visuo-motor integration and error monitoring. More recently,
Methodologies used to capture behavioral responses have Wagner et al. (2016) reported different patterns of power spectral
evolved from simple movement measures (button presses, activity reflecting movement initiation and execution (Mu and
singular body part acceleration) to the measurement of natural Beta desynchronization in sensorimotor and parietal cortex) and
whole body kinematics (Aminian and Najafi, 2004). In particular, motor control and inhibition (increased frontal Beta power)
motion capture systems have become increasingly portable, during a gait adaptation task. These important early findings
allowing the positioning of multiple independent wireless demonstrate the feasibility of characterizing modulations of EEG
sensors on a single participant, such that movement can activity in relation to body dynamics through the integration
be recorded unconstrained (Lim et al., 2011; Marin-Perianu of brain and body measurements. Nonetheless, whilst treadmill
et al., 2013). While most current motion capture solutions studies are undeniably important for establishing the feasibility
still require external cameras to track the position of sensors of recording brain activity in motion, they inherently remain
(thereby restricting the recording area), camera-less motion- lab-based demonstrations rather than real-world applications.
capture alternatives using networks of inertial and magnetic Mobile technological development has also occurred for eye-
sensors methods (exoskeleton suits composed of accelerometers, trackers, which have developed into wearable devices that can
gyroscopes and magnetic sensors) may be used to detect changes track gaze dynamics during head rotations and full body motion
in position, orientation and acceleration of body parts, allowing (Pelz et al., 2000; Babcock and Pelz, 2004). As a result, mobile
kinematics to be tracked in complete autonomy of external eye-trackers are now able to provide insight on the deployment
stationary devices (Zhu and Zhou, 2004; Luinge and Veltink, of attention in real environments. Additionally, eye-tracking
2005; Roetenberg et al., 2007). Similar to mobile EEG, wireless glasses are now typically equipped with a high-resolution camera
electromyography (EMG) systems have also been developed to enabling synchronous audio-visual recording. An increasing
record muscular activity in complete freedom of movement (Roy number of studies are using head-mounted eye-tracking devices
et al., 2013). From a mobile cognition perspective, the integration to record natural visual exploration during natural behavior
of high resolution behavioral measurements with real-world (e.g., see Hayhoe and Ballard, 2005). Mobile eye tracking has
mobile brain imaging make it possible to study the cognitive been applied to the investigation of visual memory and motor
markers related to natural behavior in relevant environments. planning of everyday-life behaviors (e.g., Pelz and Canosa, 2001;
Indeed, in this context, one important role of body kinematics Mennie et al., 2007), predictive eye movements in sports (e.g.,
and EMG data may be to define the onset of behavioral responses in squash, Hayhoe et al., 2012), in developmental research (e.g.,
and motor outputs (see Figure 3). see Franchak et al., 2010), and has been adopted in the context
Recent studies have applied this multi-modal brain and of marketing research (e.g., see Gidlöf et al., 2013). Here we
body imaging approach to characterize brain dynamics related highlight one important consequence of the development of

FIGURE 3 | An illustration of one potential implementation of the mobile cognition approach to real-world brain imaging. Natural behavior provides
multiple sources of data, recorded concurrently, allowing the integration of mobile eye-tracking and body dynamics measurement with mobile electroencephalography
(EEG). A concrete example of the application of this integrated approach can be found in a shopping situation where fixations on target objects will be use to
timestamp the EEG and proceed to the classification of brain responses. Conversely, in a sport scenario, the onset of a specific goal-oriented sporting behavior will be
used to extract meaningful information from the continuous EEG trace. We note one significant technical challenge associated with this multi-methods approach: in
practice the simultaneous synchronization of data acquisition across devices is non-trivial because each individual measure has typically been developed and used in
isolation.
mobile EEG and eye-tracking: in future the combination of One critical concern is the optimization of the signal-to-noise
gaze dynamics and first-person scene capture will enable the ratio, and the fact that mobile participation inevitably produces
timestamping of visual events either based on fixations (Baccino more noise (Gwin et al., 2010). Considering the complexity of
and Manunta, 2005) or saccades (Jagla et al., 2007) in event- disentangling neural signal from noise, traditionally researchers
related real-world brain imaging (see Figure 3). Taken together, have opted to act pre-emptively by minimizing the potential
these techniques are beginning to allow us to extract brain for “artifacts” (data unrelated to the cognitive process under
and gaze dynamics related to real-time, every-day, real-world investigation). For example, during EEG data acquisition, eye
cognitive processing. movements, along with facial and neck muscle activity, are
prevalent sources of noise. Attempts to minimize the impact
CURRENT CHALLENGES of artifacts has typically translated into avoidance techniques,
instructing participants to remain still and suppress any
For the first time these mobile devices are enabling researchers movement not directly related to the performance of the
to record behavioral, neural and physiological markers that experimental task, and requiring participants to inhibit natural
reflect cognitive processing as it occurs in natural contexts, reflexes such as blinks and swallowing (Picton et al., 2000).
while subjects are freely exploring and interacting with their Within the mobile cognition approach, however, the aspiration
environment (Makeig et al., 2009). These new technologies is to allow natural unconstrained behavior—precluding the use
allow researchers to investigate cognition in an ecologically of avoidance techniques.
valid and integrated manner that is more representative of the The introduction of motion in itself requires the wholesale
intrinsic interdependence of perception, cognition and action. re-evaluation of established practices in laboratory-based
But although mobile equipment is typically smaller and lighter cognitive research. Success may therefore require innovation
in weight (and often wireless) than the static equivalent, they in experimental design, data processing methods, and analysis
nonetheless remain subject to similar constraints in terms of techniques to reveal the patterns hidden within real world
data acquisition and analysis and associated methodological brain dynamics. Advanced processing methods to deal with
challenges. Furthermore, there are a number of additional the inevitable motion-related artifacts are in development.
methodological challenges associated with mobile cognition that Independent Components Analysis (ICA) (Makeig et al., 1996),
require innovation. which involves the statistical linear decomposition of EEG data

into maximally independent components, can be applied to the EEG systems still do not match electrodes with conductive
identification and dissociation of non-brain signals (e.g., line gel applied in terms of data acquisition quality, and are less
noise, mechanical artifacts) from cognitive brain activity, eye comfortable for the subjects (Oliveira et al., 2016). From an
movements and muscular activity (Delorme et al., 2007). aesthetic point of view, recent studies have proposed in ear EEG
As noted above, mobile EEG studies have successfully sensors (Looney et al., 2012; Kidmose et al., 2013; Mikkelsen et al.,
addressed motion-related artifacts present in data recorded 2015; Goverdovsky et al., 2016), around-the-ear electrodes grids
during high physical activities such as running on a treadmill (Debener et al., 2015; Bleichner et al., 2016; Mirkovic et al., 2016),
(Gwin et al., 2010). Equally impressively, recent study or a baseball cap fitted with electrodes (Bleichner et al., 2015).
successfully demonstrated the feasibility of parsing non-brain Future developments in mobile brain imaging should therefore
from brain signals in subjects cycling in a natural environment aim to increase the ease of use, discretion and comfort of the
(Zink et al., 2016). Moreover, advanced EEG data analyses sensors while maximizing data quality in order to be successfully
can be applied toward signal source localization based on the applicable in real-world settings.
reconstruction of equivalent dipoles of independent components Given their high temporal resolution, ERPs have been
(Gramann et al., 2010; Wagner et al., 2016). However, EEG invaluable in the investigation of the time-course of cognitive
source modeling methods are essentially based on computational processes involved in the integration of sensory inputs and
derivations and therefore require a feed of high-dimensional motor output in the face of a dynamic reality. The major
EEG data (i.e., 120+ channels) in order to reach sufficient practical issue with mobile ERPs lies in the acquisition of accurate
approximations of signals’ origins. While offering an interesting timings of such events, time-stamping the EEG trace based
option for brain signal source estimation, high density EEG is on stimulus presentation and behavioral responses. Whilst the
still impractical for the use in real-world environments. The time-locking of events of interest has been facilitated through
high-density set-up required for this approach are not truly the use of computerized paradigms in laboratories settings,
mobile at present: for example the studies by Gramann et al. acquisition of the precise timing of events of interest is much
(2010) and Gwin et al. (2010) took place on a treadmill with EEG more complex in a natural environment. This issue is of critical
cables suspended from the ceiling. Therefore, although such importance since event-related components are investigated at a
solutions allow recording of brain activity in motion, they are millisecond scale. A recent study by Jungnickel and Gramann
not yet useable in truly real-world settings. (2016) demonstrates the feasibility of recording brain activity
Even if high-density EEG recording was developed to time-locked to physical interaction with dynamically moving
allow free movement through natural spaces, advanced signal objects. In this case the definition of movement onset was
processing methods would still require sufficient amount of data based on velocity features of behavioral responses, recorded
points. The same is true for the most popular approach of through motion capture. The results revealed faster behavioral
isolating brain signals and their dynamics from the ambient noise response times and increased neural response (P300 following
in the raw data: time-locked averaging of many trials linked target stimuli) during physical pointing, in comparison to a
to hypothetical cognitive processes. This approach yields Event- classic button press condition. Jungnickel and Gramann interpret
Related Potentials (ERP), time-locked deflections in the EEG these results as suggesting dynamic integration of perceptual
trace reflecting sensory, cognitive and motor processes in the inputs, along with the execution of complex motor outputs, lead
time domain at a milliseconds scale (Luck, 2005). Brain signals of to higher computational efforts related to embodied cognitive
interest (those related to events such stimulus presentation and processes.
behavioral responses) are generally uncovered by canceling out Another classic approach in EEG research is the investigation
unrelated signals through summation and averaging of multiple of changes in the frequency domain. Through spectral density
trials. A real concern when moving into mobile data settings estimation methods, it is possible to characterize frequency
is that it may be impractical (and in some cases unnatural) to bands contribution to recorded data. Power Spectral Density
record the required numbers of repetitions of a specific event. (PSD) estimation methods (e.g., variants of Fourier Transform)
On the other hand, a distinct advantage of real-world mobile characterize frequency bands contributions to whole epochs
methods is the ability to record data over longer periods (e.g., signals. However, this stationary approach poorly represents
during home-based monitoring of patients), making collection dynamic changes over time, central to mobile EEG. More
and characterization of larger scale single participant data sets a recently however, time/frequency analyses have given insights
real innovation. into Event-Related Spectral Perturbations (ERSP) allowing the
The example of high-density recording as a solution to characterization of power spectral modulations in the time
current challenges also flags up the issue of fully equipped domain (Makeig et al., 2004). In addition, the characterization
participants’ appearance potentially defeating the purpose of of interactions between remote cell assemblies could provide
increasing ecological validity—by rendering subjects more self- insight as to how different parts of the brain work together to
conscious about the experiment and affecting their real-world bind multimodal information to create a coherent perceptual
interactions. Mobile EEG systems using dry sensors may be experience. Traveling waves analyses assess the propagation
a convenient and user-friendly solution to reduce preparation of brain signals in terms of mode and velocity, to uncover
time, which could make mobile EEG more accessible to patient local networks connections from global fields activity (Nunez
and consumer based applications (Zander et al., 2011; Chi et al., and Srinivasan, 2006). For example, recent evidence from
2012; Dias et al., 2012). However, state of the art dry electrodes human electrocortigraphy (ECoG) supports the idea that theta

oscillations related to working memory are traveling waves, state of mobile brain imaging literature. While the initial push
showing a spatial propagation across the hippocampus (Zhang toward mobile neuroimaging sensors was largely driven by brain-
and Jacobs, 2015). Based on a theoretical connectionist model computer interfaces applications aiming to maximize the online
of cognitive functions, traveling waves applied to data acquired classification of signal components at minimal cost in terms of
during natural behaviors may yield insight about brain-wide data acquisition requirements, a considerable number of mobile
cognitive networks underlying everyday life cognition. brain imaging studies have turned toward mobile systems with
Given the range of analytic techniques that could be more sensors, in order to record the high-dimensional data
applied to real-world data an important development will be required to perform advanced signal analyses. Consequently,
the synchronization of concurrent behavioral measurements there is now a large spectrum of hardware and software
to brain dynamics in mobile settings, such as simultaneous solutions, which vary in terms of their ease of setup, quality
recording and integration of mobile EEG and eye-tracking data. of data acquisition and cost. While the on-going competition
Merging of eye-tracking data (along with first-person audio- between manufacturers may be regarded as a healthy drive
visual recording) would provide information about the actual in the improvement of mobile brain imaging technologies, it
timing of engagement with real-world stimuli. For example, also seems likely that discrepancies in terms of conceptual and
initial fixations on an object or person can in principle be used methodological standards (related to both data acquisition and
to generate post-hoc timestamps for the analysis of EEG data. data analysis) may slow down the progression of the field toward
To our knowledge, there is currently no scientific publication standard practices.
reporting such integration of mobile eye-tracking and EEG One important attempt at developing standardized
data in a real-world environment. Even though eye-movements frameworks for mobile technology comes from open-source
related potentials have been used in laboratories settings, specific initiatives designed to support the processing and analysis
technical challenges inherent to mobile eye-tracking and mobile of mobile brain and body imaging data and facilitate their
EEG might still impede the integration of both techniques. integration (e.g., MoBILAB; Ojeda et al., 2014 and Lab Streaming
Eye-tracking data acquired in laboratories setups is usually Layer; Kothe, 2014). While these frameworks allow for the
based on a fixed reference frame (e.g., computer screen) which recording and processing of multimodal data, the exact
allows for the segmentation of this two-dimensional frame synchronization of data streams remains problematic due
in pre-defined Regions of Interests (ROI). This segmentation to current hardware limitations. For example, differences in
facilitates the quantification of gaze dynamics across meaningful terms of refresh rate across mobile techniques can lead to
parts of the visual display. In the case of mobile eye-tracking inconsistencies (or jittering) in the time-stamping of EEG data.
recording, this frame of reference is dynamically affected by the Notably, even though current mobile eye-tracking devices now
subject’s displacement across the three-dimensional planes of offer up to 120 Hz sampling rates, this is still insufficient to define
the environment. Therefore, the definition of ROIs in mobile the onset of visual events in the EEG trace with enough precision
environment is a significant challenge to the analysis of mobile to carry milliseconds scale analyses in the time domain. Thus,
eye-tracking data. Current options to address this issue reside whilst existing data processing schemes offer clear benefits for a
in the use of optical pattern barcodes (i.e., QR codes) or mobile cognition approach, at present, the acquisition of events
infrared-based markers placed in the environment to delimit markers remains a non-trivial challenge to the investigation of
ROI. However, this approach requires ROI to be defined a priori cognition in the real-world.
and any gaze dynamics recorded outside these areas still have Besides the increased noise that inevitably accompanies
to be annotated and processed manually. A potential solution cognition in motion, controlling for confounding variables will
for the processing and analysis of mobile eye-movement data also be a significant challenge in everyday environments. Due
is the use of automatic mapping of the video stream through to the rich and unpredictable nature of the outside world,
object recognition algorithms (Brône et al., 2011). Even with inconsistencies may arise across conditions and between subjects.
these solutions to ROI definition, recording eye movements in Whilst resolution of most of the aforementioned issues will build
a three-dimensional environment also requires accounting for upon future technical improvements of the techniques, inventive
depth, which has not been resolved yet by current mobile eye- experimental designs and methodological compromises will also
tracking systems. Since the calibration procedure is performed be of critical importance to translate cognitive research into the
on arrays of elements presented at a predefined distance, eye real-world.
fixations on elements beyond the range of the calibration are
usually poorly tracked by current systems and represent another
major issue in the study of visual exploration in natural contexts. CURRENT STATE AND FUTURE
As the preceding discussion highlights, there are real practical DIRECTIONS
challenges in developing mobile approaches. We anticipate that
the mobile cognition approach will make such multi-methods Given the novelty of mobile technologies it is perhaps
data collection more attractive across a range of measures, unsurprising to discover that the current state of the literature
allowing the study of natural behaviors (Gramann et al., 2014a) mainly consisting in proof-of-concept experiments assessing
in context and thus providing greater ecological validity in the feasibility of brain imaging in motion. Most published
the process (see Figure 3). The development of mobile brain studies have worked toward the validation of mobile techniques
imaging methods follows the same dichotomy as the current through the replication of paradigms known to reliably elicit

specific neural signals, such as well-known ERP components. demonstrate the feasibility of truly mobile EEG recording in the
These technically oriented studies have provided evidence that real-world.
portable brain imaging can reach comparable levels of accuracy Whilst there is relatively little published data thus far, existing
as traditional stationary devices within the same stationary studies have aimed primarily at testing and validating mobile
laboratory setup (Gargiulo et al., 2008; Dias et al., 2012; Liao cognition methods using traditional well-established paradigms
et al., 2012; De Vos et al., 2014b, and also during treadmill from the lab setting as benchmarks (Oliveira et al., 2016).
walking (Gramann et al., 2010; Severens et al., 2012; Lin et al., Critically, the evidence to date provides strong proof of concept
2014). This body of research has pinpointed technical (e.g., regarding the basic feasibility of a mobile cognition approach
ensuring the necessary sensor connectivity during whole-body (e.g., Figure 4). It is important to recognize however, that
movements), methodological (e.g., time-stamping of events in whilst the demonstration of proof of concept was necessary and
real-life situations) and mathematical questions (e.g., tackling provides confidence in the mobile methods going forwards, it
motion artifacts) posed by mobile brain imaging (for a review, did not directly contribute to further our understanding of real-
see Reis et al., 2014). Hardware and software solutions have been life cognition, per se. Moreover, Kingstone et al. (2008) argue
developed in response to these issues of concern (e.g., MoBI; that the legacy of laboratory-based practices and experimental
Gramann et al., 2014b), providing a solid framework to progress protocols may induce bias in the capture of the expression
to the next step: addressing actual cognitive questions in natural of human cognition in complex environments. For further
environments. progress to be made, future mobile cognition research must not
Interest in mobile cognition can be found across a number focus solely on mimicking lab-based research, but should also
of different fields in neuroscience—the relevance of real-world investigate human cognition from an embodied, integrated and
cognition has been highlighted for sport (Park et al., 2015; ecological perspective—assessing more naturalistic real-world
Cheron et al., 2016), ergonomics (Mehta and Parasuraman, behavior (e.g., Figure 5).
2013), dual-task paradigms (De Sanctis et al., 2014), spatial
cognition (Mavros et al., 2016) and mental imagery (Kranczioch
et al., 2014). Increasing numbers of studies are investigating IMPLICATIONS
cognitive processes during full-body motion in the real-world.
Perhaps the clearest example to date is provided by Debener Our view is that the mobile cognition approach will add
et al. (2012), who used an auditory oddball task to elicit P300 considerable value to existing laboratory based cognitive
ERP effects. The P300 is a well-characterized and much studied neuroscience: indeed, there are many research questions that
neural marker of attention, found during the presentations of a can only really be sensibly addressed in real-world contexts.
series of frequent distractors vs. infrequent odd-ball targets (for a Take, for example, the study of spatial navigation. To date, in
review see Polich, 2007). Debener et al. (2012) recorded the P300 humans, the examination of neural correlates of spatial cognition
in a seated-indoor condition vs. an outdoor-walking condition. has been limited to a small subset of the questions that are
An attenuation of the P300 ERP amplitude was reported in the examined in non-human animals. Many studies in rodents allow
walking condition in comparison to the sitting condition. While free-movement through space, whereas human studies typically
classification rates of single-trial ERPs were above chance levels do not—for example, they employ fixed location map reading
for both conditions, the signal-to-noise ratio (SNR) was lower in tests, or at best, Virtual Reality (VR) devices to simulate the
the walking condition, suggesting an increased amount of noise exploration of an environment while offering the experimental
in that condition. Whether these differences were a consequence control of a laboratory setting. However, we already know that
of residual noise or due to a reallocation of cognitive processing the act of moving is central to navigation. For example Ehinger
resources in the outdoor-walking condition remained open for et al. (2014) have shown that the integration of vestibular and
future investigations. A follow-up study compared outdoor- kinesthetic information (provided through the navigation of the
walking with being seated outdoors, finding equivalent P300 physical body in the environment) modulates brain activity in the
effects in each case (De Vos et al., 2014a). Importantly, a similar alpha frequency band. These findings demonstrate that sensory
degree of noise was found in walking and seated conditions, and vestibular feedback are essential parts of spatial navigation
suggesting that the muscular activity involved in walking did not that are neglected in lab-based navigation experiments. As a
result in increased movement-related noise. Furthermore, Zink result, an obvious application of the mobile cognition approach
et al. (2016) have reported a decrease in P300 amplitude during is the study of the how we explore and navigate in real-
an outdoor cycling condition in comparison to a fixed bike world environments. We envisage participants navigating around
conditions. The increased cognitive load related to natural real- complex environments (maze like corridors of large buildings, or
life behaviors appeared to be a major factor contributing around parks or city centers), performing route-finding tasks in
to the difference observed in ERP waveforms between the real world, whilst wearing a host of mobile cognition sensors.
conditions. Such an approach would allow researchers to see whether there
Why these natural behaviors (i.e., walking, cycling) should are human analogs of the phenomena seen in rats (e.g., place,
reduce attention (as indexed by changes in the magnitude of the head direction and speed signals), which to this point have only
P300) compared to being seated indoors remains unexplained— been assessed in virtual navigation tasks (Maguire et al., 1998;
and an important question for future research. Regardless, and Ekstrom et al., 2003; Doeller et al., 2010; Jacobs et al., 2010; see
more relevant here, the studies by Debener and colleagues Taube et al., 2013 for a critical review).

FIGURE 4 | Illustrative single-subject ERP example recorded while the participant walked around the corridors of Stirling University performing an
auditory oddball task (eliciting the classic P300 Event-Related Potential). Top: Average ERP waveforms across 32 channels, the P300 amplitude is most
prominent at mid-parietal electrode sites, showing the classic P300 scalp distribution. Bottom: 36 single-trial Event-Related Potentials of target stimuli classically
recorded at Pz electrode displaying consistent amplitude peaks 300 ms after stimulus onset. Examples of mobile EEG findings can be found in the literature (e.g.,
Debener et al., 2012; Zink et al., 2016). This figure of single-subject raw data provides a visual demonstration to show that ERPs can be reliably recorded across trials
and electrodes during locomotion in the real-world.
The mobile cognition approach is also particularly suited context, the interdependence between perception, cognition and
to investigating attention. Current understanding of attention action is clear, and forces attention to be considered alongside
is mainly based on visual exploration studies that have used the integration of sensorimotor information, during interaction
static scenes (or at best moving images), while participants with our environment. Whether existing theoretical accounts of
are stationary themselves. These studies necessarily place attention produced within laboratory settings can accommodate
participants in a relatively passive spectating perspective, the varieties of attention found in real-world settings remains to
potentially over-emphasizing top-down influences on visual be seen.
exploration. Indeed, the artificial nature of the stimuli or the We believe that moving toward a mobile cognition framework
task, and the restriction of participants’ behavior, all inevitably will also lead to changes in the way that problems are approached.
lead to a very specific context that does not involve the same For example, in the context of motor cognition, we predict
interaction between perception and action that can be found in a move away from stereotyped, relatively narrow, response
complex and dynamic environments. The deployment of (visual) options, toward more complex, self-generated and spontaneous
attention in the real-world may therefore be more sensitive to movement. On this basis, the investigation of sporting behavior
bottom-up influences, emphasizing the dynamic integration of can move away from examining the impact of sporting expertise
information coming from multiple external and internal sources. on the performance of abstract laboratory based tests (e.g.,
Consider, for example, the role of attention when shopping: a demonstrating that the P300 elicited by auditory oddballs is
large amount of information must be attended to in real-time— larger in elite athletes). Instead athletes can be examined whilst
providing feedback that allows us to reorient our attention on performing real sports behavior (cf. Park et al., 2015), in
elements of our surroundings that matter at that very moment, real-world environments. Such an approach is more likely to
and allows us to adjust our movements to satisfy our goals. In this deliver correlates of predictive value. Laboratory performance

FIGURE 5 | Single-subject real-world example of mobile EEG recording while taking a penalty kick. The red line marks the timing at which the ball was hit.
The green underlined time interval represents where the participant is mobile, stepping up to kick the ball. The topographic scalp map shows the averaged alpha
activity during the pre-shot interval before kicking the ball. This example recording illustrates how the integration of behavioral markers with mobile brain imaging could
allow insight into cognitive processing related to natural behaviors. The quantification of changes in power spectral activity related to the execution of goal-oriented
actions would provide information about the cognitive aspects related to real-world sporting behaviors.
may not produce effective predictors of sporting performance, in the complex, dynamic, modality integrated reality of real-life
while mobile cognition should capture the highly adaptive settings.
and integrated complexity of sporting behavior, producing The potential of a mobile cognition approach in terms of
models with far greater applied relevance. Similarly, for health clinical applications is particularly likely to be far-reaching. One
science, a mobile cognition approach could add considerable obvious first step is to examine problem behaviors, such as
value by producing evidence-based interventions of real societal falls in the elderly, using real-world monitoring to capture the
application. For example, we may be able to better predict (and physiological and neural pre-cursors of relatively rare but critical
therefore help prevent) falls in the elderly if models fully capture behavioral errors. Monitoring brain states of patients at home
the multi-modal, integrative and environment-based nature of may yield crucial information to devise and adjust informed
the problem—evident in the act of getting up from the chair to medical decisions (e.g., stroke and epilepsy patients; Askamp and
answer the door (for example, Nieuwboer et al., 2007). Equally, van Putten, 2014). The possibility to record brain activity during
for rehabilitation following brain injury such as stroke, or when whole body motion and the related processing methods to handle
considering the consequences of dementia, it seems particularly motion artifacts allow to study populations that experience
important to have an understanding of the cognitive processes difficulties to remain still such as children and patients suffering

from motor impairments (e.g., ALS, Parkinson’s disease). In design, processing and analyses methods will be required to
the future, mobile techniques may be integrated to cognitive address the major challenges of the study of human cognition in
rehabilitation strategies under the form of neurofeedback and the complex environments. More important, perhaps, moving to a
online acquisition of cognitive biomarkers metrics could be used mobile cognition approach requires an intellectual adjustment,
by medical practitioners as tailored and ecological assessment away from an atomized examination of individual cognitive
tools to assist in the diagnosis and rehabilitation processes of sub-processes, letting go of a degree of experimental control
patients affected by various neurological etiologies. and ownership of isolated cognitive domains. Fundamentally,
however, our view is that the greatest motivation for adopting
CONCLUSION a mobile cognition approach is simply the exciting prospect
of developing a more relevant, ecologically valid, cognitive
Venturing away from highly controlled laboratories-based science.
experiments opens up a range of new research questions,
adding value to the traditional cognitive neuroscience approach. AUTHOR CONTRIBUTIONS
At the very least, the exploration of real-life cognition is
likely to lead to the refinement or correction of previous All authors listed, have made substantial, direct and intellectual
models, while achieving greater ecological validity. At best, contribution to the work, and approved it for publication.
mobile cognition will remove barriers, allowing naturalistic
behavior to be studied in situ, narrowing the explanatory ACKNOWLEDGMENTS
gap between what is measured experimentally and what it
means for understanding the human mind. To deliver on this This work is supported by a scholarship from the University of
promise inevitably implies tackling a number of methodological, Stirling and a research grant from SINAPSE (Scottish Imaging
technical and conceptual issues. Innovations in experimental Network: A Platform for Scientific Excellence).
REFERENCES Brunswik, E. (1943). Organismic achievement and environmental probability.

Psychol. Rev. 50, 255–272. doi: 10.1037/h0060889
Al-Yahya, E., Dawes, H., Smith, L., Dennis, A., Howells, K., and Cockburn, Bulea, T. C., Kilicarslan, A., Ozdemir, R., Paloski, W. H., and Contreras-Vidal, J. L.
J. (2011). Cognitive motor interference while walking: a systematic (2013). Simultaneous scalp electroencephalography (EEG), electromyography
review and meta-analysis. Neurosci. Biobehav. Rev. 35, 715–728. (EMG), and whole-body segmental inertial recording for multi-modal neural
doi: 10.1016/j.neubiorev.2010.08.008 decoding. J. Visual. Exp. 77:e50602. doi: 10.3791/50602
Aminian, K., and Najafi, B. (2004). Capturing human motion using body-fixed Burgess, P. W., Alderman, N., Evans, J., Emslie, H., and Wilson, B., A. (1998). The
sensors: outdoor measurement and clinical applications. Comput. Animat. ecological validity of tests of executive function. J. Int. Neuropsychol. Soc. 4,
Virtual Worlds 15, 79–94. doi: 10.1002/cav.2 547–558. doi: 10.1017/S1355617798466037
Askamp, J., and van Putten, M. J. (2014). Mobile EEG in epilepsy. Int. J. Chapman, C. S., and Goodale, M. A. (2010). Obstacle avoidance during online
Psychophysiol. 91, 30–35. doi: 10.1016/j.ijpsycho.2013.09.002 corrections. J. Vis. 10, 1–14. doi: 10.1167/10.11.17
Babcock, J. S., and Pelz, J. B. (2004). “Building a lightweight eyetracking headgear,” Chaytor, N., and Schmitter-Edgecombe, M. (2003). The ecological validity of
in Proceedings of the Eye Tracking Research and Applications Symposium on Eye neuropsychological tests: a review of the literature on everyday cognitive skills.
Tracking Research and Applications, ETRA’2004 (San Antonio, TX), 109–114. Neuropsychol. Rev. 13, 181–197. doi: 10.1023/B:NERV.0000009483.91468.fb
doi: 10.1145/968363.968386 Chen, G., King, J. A., Burgess, N., and O’Keefe, J. (2013). How vision and
Baccino, T., and Manunta, Y. (2005). Eye-fixation-related potentials: movement combine in the hippocampal place code. Proc. Natl. Acad. Sci. U.S.A.
Insight into parafoveal processing. J. Psychophysiol. 19, 204–215. 110, 378–383. doi: 10.1073/pnas.1215834110
doi: 10.1027/0269-8803.19.3.204 Cheron, G., Petit, G., Cheron, J., Leroy, A., Cebolla, A., Cevallos, C., et al. (2016).
Barsalou, L. W. (2008). Grounded cognition. Annu. Rev. Psychol. 59, 617–645. Brain oscillations in sport: toward EEG biomakers of performance. Front.
doi: 10.1146/annurev.psych.59.103006.093639 Psychol. 7:246. doi: 10.3389/fpsyg.2016.00246
Beer, R. D. (2000). Dynamical approaches to cognitive science. Trends Cogn. Sci. 4, Chi, Y. M., Wang, Y.-T., Wang, Y., Maier, C., Jung, T.-P., and Cauwenberghs,
91–99. doi: 10.1016/S1364-6613(99)01440-0 G. (2012). Dry and noncontact EEG sensors for mobile brain-
Bleichner, M. G., Lundbeck, M., Selisky, M., Minow, F., Jager, M., Emkes, R., et al. computer interfaces. IEEE Trans. Neural Syst. Rehabil. Eng. 20, 228–235.
(2015). Exploring miniaturized EEG electrodes for brain-computer interfaces. doi: 10.1109/TNSRE.2011.2174652
An EEG you do not see? Physiol. Rep. 3:e12362. doi: 10.14814/phy2.12362 Chiel, H. J., and Beer, R. D. (1997). The brain has a body: adaptive behavior emerges
Bleichner, M. G., Mirkovic, B., and Debener, S. (2016). Identifying auditory from interactions of nervous system, body and environment. Trends Neurosci.
attention with ear-EEG: cEEGrid versus high-density cap-EEG comparison. J. 20, 553–557. doi: 10.1016/S0166-2236(97)01149-1
Neural Eng. 13:66004. doi: 10.14814/phy2.12362 Clark, A. (1997). Being There: Putting Brain, Body, and World Together Again.
Bracci, S., Cavina-Pratesi, C., Ietswaart, M., Caramazza, A., and Peelen, Cambridge, MA: The MIT Press.
M. V. (2012). Closely overlapping responses to tools and hands in Clark, A. (1999). An embodied cognitive science? Trends Cogn. Sci. 3, 345–351.
the left lateral occipitotemporal cortex. J. Neurophysiol. 107, 1443–1456. Debener, S., Emkes, R., De Vos, M., and Bleichner, M. (2015). Unobtrusive
doi: 10.1152/jn.00619.2011 ambulatory EEG using a smartphone and flexible printed electrodes around
Brône, G., Oben, B., and Goedemé, T. (2011). “Towards a more effective method the ear. Sci. Rep. 5:16743. doi: 10.1038/srep16743
for analyzing mobile eye-tracking data,” in Proceedings of the 1st International Debener, S., Minow, F., Emkes, R., Gandras, K., and de Vos, M. (2012). How about
Workshop on Pervasive Eye Tracking and Mobile Eye-Based Interaction - taking a low-cost, small, and wireless EEG for a walk? Psychophysiology 49,
PETMEI’11 (Beijing), 53. 1449–1453. doi: 10.1038/srep16743
Bronfenbrenner, U. (1977). Toward an experimental ecology of human Delorme, A., Sejnowski, T., and Makeig, S. (2007). Enhanced detection of
development. Am. Psychol. 32, 513–531. doi: 10.1037/0003-066X.32.7.513 artifacts in EEG data using higher-order statistics and independent component

analysis. Neuroimaging 34, 1443–1449. doi: 10.1016/j.neuroimage.2006. Henderson, J. M. (2007). Regarding scenes. Curr. Direct. Psychol. Sci. 16, 219–222.
11.004 doi: 10.1111/j.1467-8721.2007.00507.x
De Sanctis, P., Butler, J. S., Malcolm, B. R., and Foxe, J. J. (2014). Recalibration Jacobs, J., Kahana, M. J., Ekstrom, A. D., Mollison, M. V., and Fried, I. (2010). A
of inhibitory control systems during walking-related dual-task interference: sense of direction in human entorhinal cortex. Proc. Natl. Acad. Sci. U.S.A. 107,
a mobile brain-body imaging (MOBI) study. NeuroImage 94, 55–64. 6487–6492. doi: 10.1073/pnas.0911213107
doi: 10.1016/j.neuroimage.2014.03.016 Jagla, F., Jergelová, M., and Riecanský, I. (2007). Saccadic eye movement related
De Vos, M., Gandras, K., and Debener, S. (2014a). Towards a truly mobile auditory potentials. Physiol. Res. 56, 707–713.
brain-computer interface: exploring the P300 to take away. Int. J. Psychophysiol. Jungnickel, E., and Gramann, K. (2016). Mobile brain/body imaging (MoBI) of
91, 46–53. doi: 10.1016/j.ijpsycho.2013.08.010 physical interaction with dynamically moving objects. Front. Hum. Neurosci.
De Vos, M., Kroesen, M., Emkes, R., and Debener, S. (2014b). P300 speller BCI 10:306. doi: 10.3389/fnhum.2016.00306
with a mobile EEG system: comparison to a traditional amplifier. J. Neural Eng. Kidmose, P., Looney, D., Jochumsen, L., and Mandic, D. P. (2013). “Ear-EEG from
11:036008. doi: 10.1088/1741-2560/11/3/036008 generic earpieces: a feasibility study,” in Annual International Conference of the
Dias, N. S., Carmo, J. P., Mendes, P. M., and Correia, J. H. (2012). Wireless IEEE Engineering in Medicine and Biology Society, IEEE Engineering in Medicine
instrumentation system based on dry electrodes for acquiring EEG signals. and Biology Society Annual Conference (Osaka), 543–546.
Med. Eng. Phys. 34, 972–981. doi: 10.1016/j.medengphy.2011.11.002 Kingstone, A., Smilek, D., and Eastwood, J. D. (2008). Cognitive ethology: a new
Doeller, C. F., Barry, C., and Burgess, N. (2010). Evidence for grid cells in a human approach for studying human cognition. Br. J. Psychol. 99(Pt 3), 317–340.
memory network. Nature 463, 657–661. doi: 10.1038/nature08704 doi: 10.1348/000712607X251243
Ehinger, B. V., Fischer, P., Gert, A. L., Kaufhold, L., Weber, F., Pipa, G., Koenraadt, K. L., Roelofsen, E. G., Duysens, J., and Keijsers, N. L. (2014). Cortical
et al. (2014). Kinesthetic and vestibular information modulate alpha activity control of normal gait and precision stepping: An fNIRS study. NeuroImage 85,
during spatial navigation: a mobile EEG study. Front. Hum. Neurosci. 8:71. 415–422. doi: 10.1016/j.neuroimage.2013.04.070
doi: 10.3389/fnhum.2014.00071 Kothe, C. (2014). Lab Streaming Layer (LSL). Available online at:
Ekstrom, A. D., Kahana, M. J., Caplan, J. B., Fields, T. A., Isham, E. A., Newman, E. https://github.com/sccn/labstreaminglayer
L., et al. (2003). Cellular networks underlying human spatial navigation. Nature Kranczioch, C., Zich, C., Schierholz, I., and Sterr, A. (2014). Mobile
425, 184–188. doi: 10.1038/nature01964 EEG and its potential to promote the theory and application of
Franchak, J. M., Kretch, K. S., Soska, K. C., and Adolph, K. E. (2010). Head- imagery-based motor rehabilitation. Int. J. Psychophysiol. 91, 10–15.
mounted eye-tracking: a new method to describe infant looking. Learning 82, doi: 10.1016/j.ijpsycho.2013.10.004
1–9. doi: 10.1111/j.1467-8624.2011.01670.x Liao, L.-De., Lin, C. T., McDowell, K., Wickenden, A. E., Gramann, K.,
Gallagher, S. (2005). How the Body Shapes the Mind. New York, NY: Oxford Chang, J. Y. et al. (2012). Biosensor technologies for augmented brain-
University Press. doi: 10.1093/0199271941.001.0001 computer interfaces in the next decades. Proc. IEEE 100, 1553–1566.
Gargiulo, G., Bifulco, P., Calvo, R. A., Cesarelli, M., Jin, C., and van Schaik, A. doi: 10.1109/JPROC.2012.2184829
(2008). “A mobile EEG system with dry electrodes,” in IEEE Biomedical Circuits Lim, C. K., Luo, Z., Chen, I. M., and Yeo, S. H. (2011). Wearable wireless sensing
and Systems (Baltimore, MD), 273–276. doi: 10.1109/biocas.2008.4696927 system for capturing human arm motion. Sens. Actuat. Phys. 166, 125–132.
Gentner, D. (2010). Psychology in cognitive science: 1978–2038. Top. Cogn. Sci. 2, doi: 10.1016/j.sna.2010.10.015
328–344. doi: 10.1111/j.1756-8765.2010.01103.x Lin, Y.-P., Wang, Y., and Jung, T.-P. (2014). Assessing the feasibility of online
Gidlöf, K., Wallin, A., Dewhurst, R., and Holmqvist, K. (2013). Using eye tracking SSVEP decoding in human walking using a consumer EEG headset. J.
to trace a cognitive process: gaze behaviour during decision making in a natural Neuroeng. Rehabil. 11:119. doi: 10.1186/1743-0003-11-119
environment. J. Eye Movem. Res. 6, 1–14. doi: 10.16910/jemr.6.1.3 Long, L. L., Hinman, J. R., Chen, C. M., Escabi, M. A., and Chrobak,
Goodale, M. A., Milner, A. D., Jakobson, L. S., and Carey, D. P. (1991). A J. J. (2014). Theta dynamics in rat: speed and acceleration across
neurological dissociation between perceiving objects and grasping them. Nature the septotemporal axis. PLoS ONE 9:e97987. doi: 10.1371/journal.pone.00
349, 154–156. doi: 10.1038/349154a0 97987
Goverdovsky, V., Looney, D., Kidmose, P., and Mandic, D. P. (2016). In-ear EEG Looney, D., Kidmose, P., Park, C., Ungstrup, M., Rank, M., Rosenkranz, K., et al.
from viscoelastic generic earpieces: robust and unobtrusive 24/7 monitoring. (2012). The in-the-ear recording concept: User-centered and wearable brain
IEEE Sensors J. 16, 271–277. doi: 10.1109/JSEN.2015.2471183 monitoring. IEEE Pulse 3, 32–42. doi: 10.1109/MPUL.2012.2216717
Gramann, K., Ferris, D. P., Gwin, J., and Makeig, S. (2014a). Imaging Luck, S. J. (2005). An Introduction to the Event-Related Potential Technique.
natural cognition in action. Int. J. Psychophysiol. 91, 22–29. Cambridge: MIT Press.
doi: 10.1016/j.ijpsycho.2013.09.003 Luinge, H. J., and Veltink, P. H. (2005). Measuring orientation of human body
Gramann, K., Gwin, J. T., Bigdely-Shamlo, N., Ferris, D. P., and Makeig, S. (2010). segments using miniature gyroscopes and accelerometers. Med. Biol. Eng.
Visual evoked responses during standing and walking. Front. Hum. Neurosci. Comput. 43, 273–282. doi: 10.1007/BF02345966
4:202. doi: 10.3389/fnhum.2010.00202 Maguire, E. A., Burgess, N., Donnett, J. G., Frackowiak, R. S. J., Maguire, E. A.,
Gramann, K., Gwin, J. T., Ferris, D. P., Oie, K., Jung, T. P., Lin, C. T., et al. (2011). Frith, C. D., O’Keefe, J., et al. (1998). Knowing where and getting there: a human
Cognition in action: imaging brain/body dynamics in mobile humans. Rev. navigation network. Science 280, 921–924.
Neurosci. 22, 593–608. doi: 10.1515/RNS.2011.047 Maguire, E. A., Gadian, D. G., Johnsrude, I. S., Good, C. D., Ashburner, J.,
Gramann, K., Jung, T., Ferris, D. P., Lin, C., and Makeig, S. (2014b). Towards a Frackowiak, R. S., et al. (2000). Navigation-related structural change in the
new cognitive neuroscience: modeling natural brain dynamics. Front. Hum. hippocampi of taxi drivers. Proc. Natl. Acad. Sci. U.S.A. 97, 4398–4403.
Neurosci. 8:444. doi: 10.3389/fnhum.2014.00444 doi: 10.1073/pnas.070039597
Gwin, J. T., Gramann, K., Makeig, S., and Ferris, D. P. (2010). Removal of Maimon, G., Straw, A. D., and Dickinson, M. H. (2010). Active flight increases
movement artifact from high-density EEG recorded during walking and the gain of visual motion processing in Drosophila. Nat. Neurosci. 13, 393–399.
running. J. Neurophysiol. 103, 3526–3534. doi: 10.1152/jn.00105.2010 doi: 10.1038/nn.2492
Gwin, J. T., Gramann, K., Makeig, S., and Ferris, D. P. (2011). Electrocortical Makeig, S., Debener, S., Onton, J., and Delorme, A. (2004). Mining event-related
activity is coupled to gait cycle phase during treadmill walking. Neuroimage 54, brain dynamics. Trends Cogn. Sci. 8, 204–210. doi: 10.1016/j.tics.2004.03.008
1289–1296. doi: 10.1016/j.neuroimage.2010.08.066 Makeig, S., Gramann, K., Jung, T. P., Sejnowski, T. J., and Poizner, H.
Hampton, R. R., Hampstead, B. M., and Murray, E. A. (2004). Selective (2009). Linking brain, mind and behavior. Int. J. Psychophysiol. 73, 95–100.
hippocampal damage in rhesus monkeys impairs spatial memory in an open- doi: 10.1016/j.ijpsycho.2008.11.008
field test. Hippocampus 14, 808–818. doi: 10.1002/hipo.10217 Makeig, S., Bell, A., Jung, T.-P., and Sejnowski, T. J. (1996). Independent
Hayhoe, M., and Ballard, D. (2005). Eye movements in natural behavior. Trends component analysis of electroencephalographic data. Adv. Neural Inf. Process.
Cogn. Sci. 9, 188–194. doi: 10.1016/j.tics.2005.02.009 Syst. 8, 145–151.
Hayhoe, M. M., McKinney, T., Chajka, K., and Pelz, J. B. (2012). Predictive Malkova, L., and Mishkin, M. (2003). One-trial memory for object-
eye movements in natural vision. Exp. Brain Res. 217, 125–136. place associations after separate lesions of hippocampus and posterior
doi: 10.1007/s00221-011-2979-2 parahippocampal region in the monkey. J. Neurosci. 23, 1956–1965.

Marin-Perianu, R., Marin-Perianu, M., Havinga, P., Taylor, S., Begg, R., in Parkinson’s disease during unconstrained activity. Movem. Disord. 28,
Palaniswami, M., et al. (2013). A performance analysis of a wireless body-area 1080–1087. doi: 10.1002/mds.25391
network monitoring system for professional cycling. Pers. Ubiquitous Comput. Sbordone, R. J., and Long, C. J. (eds.). (1996). “Ecological validity: some critical
17, 197–209. doi: 10.1007/s00779-011-0486-x issues for the neuropsychologist,” in Ecological Validity of Neuropsychological
Mavros, P., Austwick, M. Z., and Smith, A. H. (2016). Geo-EEG: towards the use Testing (Delray Beach, FL: GR Press/St. Lucie Press), 15–41.
of EEG in the study of urban behaviour. Appl. Spat. Anal. Policy 9, 191–212. Schaefer, S. (2014). The ecological approach to cognitive-motor dual-tasking:
doi: 10.1007/s12061-015-9181-z findings on the effects of expertise and age. Front. Psychol. 5:1167.
McFarland, W. L., Teitelbaum, H., and Hedges, E. K. (1975). Relationship between doi: 10.3389/fpsyg.2014.01167
hippocampal theta activity and running speed in the rat. J. Comp. Physiol. Schmuckler, M. A. (2001). What is ecological validity? A dimensional analysis.
Psychol. 88, 324–328. doi: 10.1037/h0076177 Infancy 2, 419–436. doi: 10.1207/S15327078IN0204_02
Mehta, R. K., and Parasuraman, R. (2013). Neuroergonomics: a review of Severens, M., Nienhuis, B., Desain, P., and Duysens, J. (2012). “Feasibility
applications to physical and cognitive work. Front. Hum. Neurosci. 7:889. of measuring event related desynchronization with electroencephalography
doi: 10.3389/fnhum.2013.00889 during walking,” in Annual International Conference of the IEEE Engineering
Mennie, N., Hayhoe, M., and Sullivan, B. (2007). Look-ahead fixations: in Medicine and Biology Society, Vol. 8 (San Diego, CA), 2764–2767.
anticipatory eye movements in natural tasks. Exp. Brain Res. 179, 427–442. Sinnamon, H. M. (2006). Decline in hippocampal theta activity during
doi: 10.1007/s00221-006-0804-0 cessation of locomotor approach sequences: amplitude leads frequency
Mikkelsen, K. B., Kappel, S. L., Mandic, D. P., and Kidmose, P. (2015). EEG and relates to instrumental behavior. Neuroscience 140, 779–790.
recorded from the ear: characterizing the Ear-EEG method. Front. Neurosci. doi: 10.1016/j.neuroscience.2006.02.058
9:438. doi: 10.3389/fnins.2015.00438 Smilek, D., Birmingham, E., Cameron, D., Bischof, W., and Kingstone, A. (2006).
Mirkovic, B., Bleichner, M. G., De Vos, M., and Debener, S. (2016). Target Cognitive Ethology and exploring attention in real-world scenes. Brain Res.
speaker detection with concealed EEG around the ear. Front. Neurosci. 10:349. 1080, 101–119. doi: 10.1016/j.brainres.2005.12.090
doi: 10.3389/fnins.2016.00349 Spooner, D. M., and Pachana, N. A. (2006). Ecological validity in
Neisser, U. (1976). Cognition and Reality. San Francisco, CA: Freeman and Co. neuropsychological assessment: a case for greater consideration in research
Nieuwboer, A., Kwakkel, G., Rochester, L., Jones, D., van Wegen, E., Willems, A. with neurologically intact populations. Arch. Clin. Neuropsychol. 21, 327–337.
M., et al. (2007). Cueing training in the home improves gait-related mobility doi: 10.1016/j.acn.2006.04.004
in Parkinson’s disease: the RESCUE trial. J. Neurol. Neurosurg. Psychiatr. 78, Taube, J. S., Valerio, S., and Ryan M. Y. (2013). Is navigation in virtual
134–140. doi: 10.1136/jnnp.200X.097923 reality with fMRI really navigation? J. Cogn. Neurosci. 25, 1008–1019.
Nunez, P. L., and Srinivasan, R. (2006). A theoretical basis for standing doi: 10.1162/jocn_a_00386
and traveling brain waves measured with human EEG with implications Thaler, L., Arnott, S. R., and Goodale, M. A. (2011). Neural correlates of natural
for an integrated consciousness. Clin. Neurophysiol. 117, 2424–2435. human echolocation in early and late blind echolocation experts. PLoS ONE
doi: 10.1016/j.clinph.2006.06.754 6:e20162. doi: 10.1371/journal.pone.0020162
Ojeda, A., Bigdely-Shamlo, N., and Makeig, S. (2014). MoBILAB: an open source Wagner, J., Makeig, S., Gola, M., Neuper, C., and Müller-Putz, G. (2016). Distinct
toolbox for analysis and visualization of mobile brain/body imaging data. Front. band oscillatory networks subserving motor and cognitive control during gait
Hum. Neurosci. 8:121. doi: 10.3389/fnhum.2014.00121 adaptation. J. Neurosci. 36, 2212–2226. doi: 10.1523/JNEUROSCI.3543-15.2016
Oliveira, A. S., Schlink, B. R., Hairston, W. D., König, P., and Ferris, D. P. Williams, M. J., and Long, C. J. (2015). “Everyday cognition and the ecological
(2016). Proposing metrics for benchmarking Novel EEG technologies validity of intellectual and neuropsychological tests,” in Cognitive Approaches
towards real-world measurements. Front. Hum. Neurosci. 10:188. to Neuropsychology, eds J. M. Williams and C. J. Long (New York, NY: Plenum
doi: 10.3389/fnhum.2016.00188 Press), 123–141.
Park, J. L., Fairweather, M. M., and Donaldson, D. I. (2015). Making the case for Wyble, B. P., Hyman, J. M., Rossi, C. A., and Hasselmo, M. E. (2004). Analysis
mobile cognition: EEG and sports performance. Neurosci. Biobehav. Rev. 52, of theta power in hippocampal EEG during bar pressing and running
117–130. doi: 10.1016/j.neubiorev.2015.02.014 behavior in rats during distinct behavioral contexts. Hippocampus 14, 662–674.
Pelz, J. B., and Canosa, R. (2001). Oculomotor behavior and perceptual strategies doi: 10.1002/hipo.20012
in complex tasks. Vision Res. 41, 3587–3596. doi: 10.1016/S0042-6989(01) Zander, T. O., Lehne, M., Ihme, K., Jatzev, S., Correia, J., Kothe, C., et al. (2011).
00245-0 A dry EEG-system for scientific research and brain-computer interfaces. Front.
Pelz, J. B., Canosa, R. L., Kucharczyk, D., Babcock, J., Silver, A., and Konno, D. Neurosci. 5:53. doi: 10.3389/fnins.2011.00053
(2000). Portable eyetracking: a study of natural eye movements. Proc. SPIE Zhang, H., and Jacobs, J. (2015). Traveling theta waves in the human hippocampus.
3959, 566–582. doi: 10.1117/12.387190 J. Neurosci. 35, 12477–12487. doi: 10.1523/JNEUROSCI.5102-14.2015
Picton, T. W., Bentin, S., Berg, P., Donchin, E., Hillyard, S. A., Johnson, R. Zhu, R., and Zhou, Z. (2004). A real-time articulated human motion tracking using
Jr., et al. (2000). Guidelines for using human event-relate d potentials to tri-axis inertial/magnetic sensors package. IEEE Trans. Neural Syst. Rehabil.
study cognition: recording standards and publication criteria. Psychophysiol. 37, Eng. 12, 295–302. doi: 10.1109/TNSRE.2004.827825
127–152. doi: 10.1111/1469-8986.3720127 Zink, R., Hunyadi, B., Huffel, S., Van Huffel, S., and Vos, M. D. (2016). Mobile EEG
Piper, S. K., Krueger, A., Koch, S. P., Mehnert, J., Habermehl, C., Steinbrink, J., on the bike: disentangling attentional and physical contributions to auditory
et al. (2014). A wearable multi-channel fNIRS system for brain imaging in attention tasks. J. Neural Eng. 13:46017. doi: 10.1088/1741-2560/13/4/046017
freely moving subjects. Neuroimage 85, 64–71. doi: 10.1016/j.neuroimage.2013.
06.062 Conflict of Interest Statement: The authors declare that the research was
Polich, J. (2007). Updating P300: An Integrative theory of P3a and P3b. Clin. conducted in the absence of any commercial or financial relationships that could
Neurophysiol. 118, 2128–2148. doi: 10.1016/j.clinph.2007.04.019 be construed as a potential conflict of interest.
Reis, P. M., Hebenstreit, F., Gabsteiger, F., von Tscharner, V., and Lochmann,
M. (2014). Methodological aspects of EEG and body dynamics measurements Copyright © 2017 Ladouce, Donaldson, Dudchenko and Ietswaart. This is an open-
during motion. Front. Hum. Neurosci. 8:156. doi: 10.3389/fnhum.2014.00156 access article distributed under the terms of the Creative Commons Attribution
Roetenberg, D., Slycke, P. J., and Veltink, P. H. (2007). Ambulatory position and License (CC BY). The use, distribution or reproduction in other forums is permitted,
orientation tracking fusing magnetic and inertial sensing. IEEE Trans. Biomed. provided the original author(s) or licensor are credited and that the original
Eng. 54, 883–890. doi: 10.1109/TBME.2006.889184 publication in this journal is cited, in accordance with accepted academic practice.
Roy, S. H., Cole, B. T., Gilmore, L. D., De Luca, C. J., Thomas, C. A., Saint- No use, distribution or reproduction is permitted which does not comply with these
Hilaire, M. M., et al. (2013). High-resolution tracking of motor disorders terms.

Fnhum 10 00694

Uploaded by

Copyright:

Available Formats

Fnhum 10 00694

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fnhum 10 00694

Uploaded by

Copyright:

Available Formats

REVIEW

published: 12 January 2017

Understanding Minds in Real-World

Psychology, Faculty of Natural Sciences, University of Stirling, Stirling, UK

Bojana Mirkovic contributed to the A RATIONALE FOR MOBILE REAL-WORLD COGNITION

Frontiers in Human Neuroscience | www.frontiersin.org 1 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 2 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 3 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 4 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 5 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 6 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 7 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 8 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 9 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 10 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 11 January 2017 | Volume 10 | Article 694

REFERENCES Brunswik, E. (1943). Organismic achievement and environmental probability.

Frontiers in Human Neuroscience | www.frontiersin.org 12 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 13 January 2017 | Volume 10 | Article 694

Frontiers in Human Neuroscience | www.frontiersin.org 14 January 2017 | Volume 10 | Article 694

You might also like