topographical N170 to allocentric and egocentric reference frames: evidence
from allocentric system violation study
the current trends in neuroscience of focusing on the navigational aspect of
the cognition, especially after recent discovery of grid cells in the parahippocampal region (Moser, Kropff, and Moser, 2008), processing of the
topographical information in the brain is not well studied in the field of
event-related potentials (ERP) research, unlike processing of fairly well
studied auditory, visual and other modalities.
Recently discovered topographical
N170 component (Baker and Holroyd, 2008; Baker and Holroyd, 2013) is an
extremely promising tool for bringing EEG research to this very vibrant area of
scientific investigations. The discovery of this novel component has wide
ranging implications for understanding of the fundamental principle of the neural
computations, which are underlying navigation. The ERP research could
bridge the gap between the animal
studies and human cognition research, allowing for a much needed
transdisciplinary research (Baker, Lam, Lan, Uban, and Weinberg, 2015), which
would eventually lead to a complete understanding of the processing of the
N170 component was first described as a negative deflexion almost around 170 ms
after the onset of the stimulus, which is related to spatial navigation. It was
demonstrated in a series of experiments, in which stimuli signifying reward,
were placed in different spatial locations in the T-maze (Baker and Holroyd,
2008; Baker and Holroyd, 2013). What is remarkable about the topographical N170
is that the latency of this component in response for the image of the right
turn was earlier than the one for the image of the left turn, which suggests
that spatial navigation is definitely underlying this phenomenon. The extensive
studies of the aforementioned component
have revealed that is can be precisely localized to the right parahippocampal
region, with distinct subdivision into functionally different anterior and
posterior clusters (Baker, Umemoto, Krawitz and Holroyd, 2015), which is in
agreement with with the previous research done in animal studies.
advance this line of research, profound considerations about the nature of the
processing of the topographical information are needed. In particular, it is
important to take into account the theory of the spatial reference frames,
which distinguishes between egocentric and allocentric points of view (Klatzky,
1998). For the egocentric point of view, the starting point of the coordinates
or reference point is the person who is perceiving the environment and a
spatial location of an object will be identified with respect to the person
identifying it; for example, in this reference frame a person would perceive an
object A being in front of them. In
contract, in allocentric reference system, spatial localisation of an object is
processed in a reference frame that is independent from the person receiving it
and external to her or him; thus, for example, a person might perceive that an
object A is located between objects B and C, and no matter how the perceiving
person would relocate himself or herself, this identification of spatial
location will remain unchanged because it is based on the external reference
point. It is worth noting that there is an ongoing debate about this matter,
including an opinion that allocentric spatial representation may be based on
egocentric reference frame (Filimon, 2015). This makes the question of the
nature of the referencing in spatial navigation even more worth studying, and
neuroimaging can greatly contribute to this line of research.
a better understanding of the nature of the topographical N170 component, there
is an urgent need to elucidate the question whether this component is sensitive
to the egocentric reference point and / or to the allocentric one. This is a
crucial question that should be addressed in order to advance our understanding
of topographical information processing in general. An experiment described
below is a valuable undertaking to achieve this goal.
For this experiment a paradigm of virtual T-Maze task
was used, which was proven to be an efficient setting in many experiments
cutting across many different subjects (for example: Baker and Holroyd, 2011; Baker,
Stockwell, Barnes and Holroyd, 2011). In our case, classical virtual
T-maze was complemented with an additional element, namely the different
mountains behind the walls of the branches of the T-maze. In case of the left
branch of the maze, there was a small dark mountain behind it, and for the
right branch, there was a high other mountain. In the beginning of the
experiment, the view from above of the T-maze was shown from different angles
so that the participant would learn about the spatial relationships between the
maze and the surrounding mountains. The views from above are presented on the
Figure 1. The views of the T-Maze from above from
In the beginning of each trial, a entry of maze was
presented (fig. 2a) for 500 ms, then the same screen with arrows appeared (fig.
triggering the choice to turn to the left or to the right. The goal was
to choose a turn, which would bring a reward. The assignment of the reward and
non-reward conditions was random across all trials. After a participant would
choose a turn, the
image would return to the start of the maze without arrows (fig. 2a) for 250 ms
to separate the expected ERP component from motor movement (the right hand was
used to press the button to choose the turn).
Figure 2. (a) An entry to the maze; (b) An entry to
the maze with arrows, triggering the choice by the participant.
a screen showing
the chosen turn (fig. 3a-c) was presented for 500 ms without reward. In 30% of
the trials the mountain that should have been behind the wall of the chosen
branch was replaced with the one for the other turn, as exemplified on fig. 3
a-c; the remaining 70% had the correct mountain corresponding to the top view
picture shown in the beginning of the experiment. This was a central
manipulation of the experiment, as it aimed at differentiating the allocentric
reference system from the egocentric one. The processing of the topographical
information related to the right or left turn is a purely based on the
egocentric reference point, whereas the processing of the information about which
mountain is behind which branch of the maze is based on the allocentric
reference point. The current experiment aims at assessing sensitivity of the topographical N170
component to the violation of information, which is based on the allocentric
point of view. Finally, a reward or non-reward stimulus was shown 500 ms, which
was represented by an apple for reward and an orange for no reward (fig. 3 b,
c). The presentation of the reward was an onset for the stimuli for the ERP
component. Then, it was followed by an empty deadend of the maze again. There
were total of approximately 300 trials.
A B C
Figure 3. (a) Left turn, congruent condition without
reward-related stimulus;(b) Left turn, incongruent condition with an apple
signifying reward; (c) Right turn, incongruent with a stimulus signifying no
Participants were recruited using resources of
Department of Psychology and among personal acquaintances of the researchers.
Total number of participants was 8, 5 of them were female, mean age was 21.75
(standard deviation was 4.2). 6 of the
participants have taken part in the experiment in order to receive credits for
their courses in Psychology department. Also, 2 of the participants were
volunteers from Behavioral and Neural Sciences program. 2 participants were
excluded from the analysis due to their performance on the test (described
later). All of the participants were right handed and had no history of
neurological disease and brain damage.
All participants had normal or corrected to normal vision. All the work with
the participants was done in accordance with the ethical standards, to which
Rutgers University adheres.
Prior to the experiment participants were informed
about the experimental procedure and asked to give an informed consent. Data
about age, gender, health status and handedness were collected. Before the
experiment, the participants were instructed to sit in a relaxed and
comfortable position and perform the task. The distance from the screen was
approximately 80 cm. Instructions for the experimental task were presented on
the screen. The participants were informed that this study is about how people
are navigating. The participants were asked to minimize their movements and eye
blinks during the presentation of the stimuli. The participants were instructed
to choose between going left and going right in a virtual T-maze environment in
order to have a reward, and that it is them who determine the choice. To choose
turn to the left participants were instructed to press button ‘1’ on the
keyboard, to choose the right one – button ‘2’. They were further instructed
that for the correct choice there will be a reward, which will be signified by
an apple, whereas for the incorrect choice there will be no reward and it will
be signified by an orange.
The experimental session was subdivided into 6 blocks,
each consisting of 50 stimuli presentations, each block lasted approximately 4
minutes in time. After each block a participant had a
period of rest, duration of which was determined by herself or himself. During
the breaks participants were allowed to move to relieve the muscular tension
and to ask questions about the experimental procedures.
the end of the experiment, each participant except the first one had a short
test to evaluate whether they have remembered, which mountains were on the
right and and which were on the left. The test consisted of presenting each of
the possible combinations of mountain and turn (congruent and incongruent for each
one, summing to the total of 4) and asking whether a particular picture
corresponded to the left or to the right turn. The question presentation is
exemplified on the fig. 4. On the basis of this test, people who gave incorrect
answers in more than 25% of the cases were excluded from the analysis because
such performance suggested that the participants didn’t pay attention to the
the experiment, the participants were debriefed.
The experiment was programmed with E-Prime Software
3.0. The total duration of the experiment from the arrival of participant to
the end of the task was approximately 1.5 hours.
Figure 4. An example of the question asked after the
acquisition and analysis
The EEG was recorded using 23 Ag/AgCl electrodes. The
EEG recording was amplified by QuickAmp-40 amplifier produced by BrainVision.
EasyCap DE-82211 Herrsching was used to place the electrodes. The sampling rate
1000 Hz was applied during recording. Recording electrodes were grounded to the
ground electrode, which was located between Fz and Fpz electrodes. The
impedance for all the electrodes was kept below 15 kOm.
were recording EOG bipolarly (i. e., from both eyes). One pair of the
electrodes for ocular recordings were placed on the eye outer canthi, and
additional electrode was fixed on zygomaticus muscles on the cheek. Scalp EEG
was recorded from 23 electrodes mounted in an electrode cap with respect to a
left and right mastoid reference. The electrodes used for this recording are the following:
C3 C4 Cz F3 F4 FC1 FC2 FCz Fp1 Fp2 Fpz Fz LHZ M1 M2 Oz P3 P4 PO7 PO8 Pz RHZ VEOG
For the ERP
analysis Brain Vision Analyser (version 2.1) was used to process the EEG
data. Raw data were filtered using Zero phase shift Butterworth filters with
the low cutoff of 0.1 Hz (order 1) and the high cutoff of 30 Hz (order 2). Eye
movement correction based on Gratton and Coles (Gratton et al, 1983) algorithm was performed using
an electrode above and below the right eye (Fp2 and RHZ) as a reference.
After that, EEG was re-referenced to average
Rejection was performed using semi Automatic Inspection with maximal allowed
voltage step of 20 µV/ms, criteria for noisy data was difference of values in
intervals more than 100 µV, interval length chosen was 200 ms. In addition, the
intervals of 100 ms were checked for low activity, with the lowest allowed
activity being 0.5 µV. ERPs were segmented beginning 200 ms prior to the
onset of the stimuli and ending 1,000 ms after its onset. Trials containing
ocular or motor artifacts were excluded from the analysis.
Time windows were chosen in accordance with the
established practices in the previous studies. To determine the magnitude of
the NT170 component, additional measure of P1 component was used, which was
measured as a peak value at a latency in the time window of 50–150 ms. For N170
component the window of 130-220 ms was used in accordance with the established
literature on the subject matter (example: Blau et al, 2007). To evaluate
topographical N170 component, the value of P1 component was subtracted from
N170 component. The previous studies have shown that the maximal value of
topographical N170 is registered at PO8 electrode (Baker and Holroyd, 2008; Baker and
Holroyd, 2013), which is consistent with fMRI studies on the
localization of topographical N170 component. Therefore, the analysis was
performed for this region of interest, as it is the optimal channel for
detection of the NT170 component.
In order to identify the topographical N170 component,
firstly, grand mean ERP waveforms at electrode PO8 comparing brain responses to
the experimental conditions of the left turn and the right turn stimuli for the
condition without the violation of allocentric reference point frame were
computed, which are represented in a fig. 5. Visual inspection has shown that
the peak latency of the topographical N170 for the right turn is preceding the
peak latency of the same component for the left turn, which is in accordance
with the extensive and profound previous research on the subject matter (Baker and Holroyd,
2008; Baker and Holroyd, 2013). In addition, the amplitude of topographical
N170 was apparently greater for the right turn versus the left turn. In order
to evaluate the actual differences, the peak values for the right turns and the
left turns were compared using t-test. To our deepest regrets, the t-test
failed to find significance in the differences between these latencies (p =
0.63, standard deviation = 12.67). It is well worth mentioning that the
data for individual subjects was not exactly consistent. Namely, for the three
out of six subjects, the latency of the peak of the topographical N170
component for the left turn actually preceded the one for the right turn.
However, this is not a problem for this research. Then, the amplitudes for the
same topographical N170 were compared for the right turn versus the left turn.
And again, the statistics have unfortunately failed to capture the difference
observed on the graph, t-test didn’t show any significance (p=0.91, standard
5. Grand average waveforms for the left and right turns, electrode PO8; the
left turn is represented by the black line, the right one – by the red line.
For the assessment of the sensitivity of topographical
N170 component to allocentric reference frame violations, grand mean ERP
waveforms at the same electrode PO8 comparing brain responses to the
experimental conditions of congruent (correct mountain) and incongruent
(incorrect mountain) stimuli were computed, presented in fig. 6. It can be
easily observed that incongruent stimuli has produced a topographical N170 with
smaller amplitude and later latency of the peak than the congruent stimuli. The
statistical analysis, however, again did not show any significant differences
between these conditions. For the differences in the latencies, t-test results
were not supportive (p=0.67, standard deviation = 23.28). Similarly to that,
t-test for the amplitudes has shown that there is no significant difference as
well (p=0.49, standard deviation = 2.29).
6. Grand average waveforms for the congruent and incongruent conditions,
electrode PO8; congruent is represented
by the black line, the incongruent one – by the red line.
The analysis, which is summarized above, clearly shows
that the data obtained by this particular study is very interesting, but,
unfortunately, it is not enough for any conclusive statement. Recruiting more
participants and finding better parameters for data analysis (choosing another
electrode for comparison, redefining the time windows, etc) may improve the
statistical results. The inconsistencies across subjects can be explained by
the simple observation that people are different in their navigational skill;
some people fail to show the expected response (topographical N170) merely
because they are not performing the task in the correct way. The differences,
observed by visual inspection, are consistent with the previous research and
hold promise for future success of the research.
With respect to the allocentric reference frame,
visual inspection have found that topographical N170 is sensitive to violations
in allocentric system. The smaller amplitude and later onset of the peak is
clearly showing that the processing of allocentric information is reflected by
topographical N170 component. This is a very exciting finding, as it would
allow for a deeper understanding of the cognitive information processing.
One of the objections would be that dominant handedness
played a role in creating the differences in the latencies of the peaks of the
topographical N170 component, which would have been easily assessed by the Edinburgh
Handedness Questionnaire inventory for the assessment and analysis of handedness
(Oldfield 1971), which allows for a refined determination of the degree of
handedness, including assessment of ocular dominance; including familial left
handedness might have also been beneficial, as a huge body of the ERP research
is showing that genetic component of left handedness, which is not always
reflected in the left handed phenotype, plays a role in lateralization of the
functions in the brain (Kutas and Hillyard, 1980; Polich, John, and Lisa D.
Hoffman, 1998; Thompson, Cannon, Narr, Van Erp, Poutanen, Huttunen, Lönnqvist,
Standertskjöld – Nordenstam, Kaprio, Khaledy and Dail, 2001). However, this
considerations are should not be taken into account, as the huge amount of the
experimental work (Baker and Holroyd, 2008; Baker and Holroyd, 2013) shows
that the topographical N170 component is beyond any doubt related to the
possible criticism that the onset of the stimuli of an empty turn (which
appears before the reward stimuli and lasts for 500 ms) should have been
chosen, as the topographical component might be considered to be related to the
topography as opposed to reward presentation and that there should have been
jitter before the stimuli.
general, the current research is not marred by the above possible criticisms,
as the scientific value of the research is exceptionally high. It opens endless
possibilities to transdisciplinary research, which would allow to gain holistic
understanding of the neural computations, which are underlying the spatial
navigation. Localization of the topographical N170 component can provide a way
to conduct research, which would allow for a breakthrough in our understanding
of how cognition works.
T.E. and Holroyd, C.B., 2011. Dissociated roles of the anterior cingulate
cortex in reward and conflict processing as revealed by the feedback
error-related negativity and N200. Biological
psychology, 87(1), pp.25-34.
T.E. and Holroyd, C.B., 2008. Which way do I go? Neural activation in response
to feedback and spatial processing in a virtual T-maze. Cerebral Cortex, 19(8),
T.E. and Holroyd, C.B., 2013. The topographical N170: electrophysiological
evidence of a neural mechanism for human spatial navigation. Biological psychology, 94(1), pp.90-105.
T. E., Lam, V., Lan, N., Uban, K. A., & Weinberg, J., 2015. Of mice and
wo/men: transdisciplinarity in the laboratory. Transforming Addiction: Gender, Trauma, Transdisciplinarity, p. 50.
T.E., Stockwell, T., Barnes, G. and Holroyd, C.B., 2011. Individual differences
in substance dependence: at the intersection of brain, behaviour and cognition.
Addiction biology, 16(3), pp.458-466.
T.E., Umemoto, A., Krawitz, A. and Holroyd, C.B., 2015. Rightward-biased
hemodynamic response of the parahippocampal system during virtual navigation. Scientific reports, 5.
V.C., Maurer, U., Tottenham, N. and McCandliss, B.D., 2007. The face-specific
N170 component is modulated by emotional facial expression. Behavioral and brain functions, 3(1), p.7.
F., 2015. Are all spatial reference frames egocentric? Reinterpreting evidence
for allocentric, object-centered, or world-centered reference frames. Frontiers in human neuroscience, 9.
G., Coles, M.G. and Donchin, E., 1983. A new method for off-line removal of
ocular artifact. Electroencephalography
and clinical neurophysiology, 55(4),
L.D. and Polich, J., 1999. P300, handedness, and corpus callosal size: gender,
modality, and task. International Journal
of Psychophysiology, 31(2),
R.L., 1998. Allocentric and egocentric spatial representations: Definitions,
distinctions, and interconnections. In Spatial
cognition (pp. 1-17). Springer, Berlin, Heidelberg.
M. and Hillyard, S.A., 1980. Event-related brain potentials to semantically
inappropriate and surprisingly large words. Biological
psychology, 11(2), pp.99-116.
E.I., Kropff, E. and Moser, M.B., 2008. Place cells, grid cells, and the
brain’s spatial representation system. Annual
review of neuroscience, 31.
R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory.
Neuropsychologia, 9(1), pp.97-113.
P.M., Cannon, T.D., Narr, K.L., Van Erp, T., Poutanen, V.P., Huttunen, M.,
Lönnqvist, J., Standertskjöld-Nordenstam, C.G., Kaprio, J., Khaledy, M. and
Dail, R., 2001. Genetic influences on brain structure. Nature neuroscience, 4(12),