Late ERP amplitude for self-face perception positively associated with heartbeat perception accuracy

Perception of yourself involves the integration of information from various sources. In a number of studies, it was found that the perception of one’s own face is accompanied by an increase in the accuracy of perception of heartbeats and the amplitude of brain potentials caused by heart beats. In this study, subjects had to do a heartbeat count test to determine the accuracy of the interception. Then, the subjects were presented with the faces of an unknown person, a friend and the subject’s own face. The simultaneous registration of EEG was organized. We analyzed the relationship between the amplitude of the evoked potentials when viewing these faces and the accuracy of interoception. It was found that the amplitude of the late EP component (850 - 1106 ms) has a positive correlation with IAcc in the central and right parietal and occipital areas when perceiving one’s own face. According to the localization of distributed sources of activity, it was found that the connection is localized in the right anterior upper temporal cortex. Thus, the association between exteroceptive perception of one’s own face and IAcc occurs in the late period of EP. Moreover it is localized in the right temporal region of the cortex, associated with multisensory integration and recognition of personal information.

Increasing and decreasing sequences were excluded and the remaining sequences variants were randomised between subjects. Participants were asked do not move during the registration, keep their eyes open and count their own heartbeats silently. The beginning and the end of the counting intervals were signalled acoustically. After the stop signals participants were asked to verbally report the number of counted heartbeats. The participants were not informed about the length of the counting phase nor about the quality of their performance. Participants were not permitted to take their pulse or to attempt any other manipulations that could facilitate the detection of heartbeats.
Interoceptive accuracy (IAcc) was estimated as the mean heartbeat perception score according to the following transformation: n -number of periods, RH -Recorded Heartbeats, CH -Counted Heartbeats.
Maximum is equal 1 and correspond to absolute accuracy of heartbeat perception.

Stimuli
Each subject sent photos of his or her own face and the photo of his or her friend's face of the same sex. The faces on the photoes should be emotionless and should be clearly identified (without gross shadows or hairs covering the face). The faces from Karolinska Directed Emotional Faces (KDEF) with neutral expression, four females (AF01NES, AF11NES, AF19NES, AF34NES) and four males (AM10NES, AM17NES, AM23NES, AM26NES), were taken as control condition. All photos were cut off on the upper edge of the hairstyle and the lower part of the neck. Finally, the photos were discolored and aligned in contrast and brightness. The size of the images was 23 X 16 cm, which from a distance of 1 meter was 13°7'x 9°9'. All photos were shown against a black background.

Procedure
Registration took place in the room with controlled lighting and each participants was sitting on a chair at a distance of 1 meter in front of the monitor. Four faces of unknown persons of the same sex as the participants were given to the subject with the instructed to choose one that does not cause positive or negative emotional associations and does not attract attention and will not be remembered by chance meeting. The choice of one person is due to the fact that the participant's and the friend's photo were also in the singular. Then the instruction followed, during which all three faces (the participant's own face, participant's friend face and the face of unknown person) after appropriate processing were presented to the participants of the experiment were presented and asked them to watch these faces during the task.
In the course of the experiment the faces were presented in the random order, each face been presented 32 times. The sequence of the events in trial was as follows: a white cross in the center of the screen for 1 second, then a face presentation for 5 seconds and an interstimulus interval for 1.5 seconds. Paradigm programming and stimuli presentation were carried out in PsychoPy 1.82.
Photosensor in the upper left corner of the screen was used for precise synchronization of stimuli presentation with the EEG.
Electrode resistance was below 10 kOhms for all electrodes. The EEG was recorded against the averaged values of the ear lobes reference electrodes. Online high-pass filter was 0.05 Hz and a low-pass filter was 100 Hz.
Horizontal eye movements were recorded with electrodes placed at the outer canthus of each eye. Vertical eye movements were recorded with electrodes placed above and below left eye.
The band-pass filter for EOG recording was 1-20 Hz.
ECG was recorded using the two external channels with a bipolar ECG lead II configuration.
ECG data were offline band-pass filtered between 1 and 40 Hz.

Event Related Potentials
For the ERP analysis, the continuous EEG data were filtered between 0.1 and 15 Hz, rereferenced to common average and epoched from −200 to 1500 ms from the beginning of the stimulus presentation. Segments were baseline corrected for a pre-stimulus interval of −200 to 0 ms. Epochs were visually inspected and in the case of presence of pronounced artifacts were removed. Next, the individual components were extracted using the Infomax algorithm. The removal of the components containing EOG and myographic artifacts was carried out semiautomatically. Further, the epochs were again visually inspected and, exceeding the threshold of 80 mV, were removed. Finally, EEG data were averaged for each category.
The analysis was carried out in T5/T6 for the N170 in the period from 140 to 200 ms, in the Cz electrode for N200 from 200 to 320 ms, in PO7/PO8 for the N250 in the period from 230 to 330 ms. The analysis of the late components was carried out in the CPz electrode for P300 in the period 400 -600 ms and for LPP in the period 600 -1000 ms. The procedure was as follows. First, t-test differences (p < 0.05) were calculated between the groups in each electrode and at each time point in the period from -0.05 to 0.6 seconds relative to the timing of the motor response. Based on spatiotemporal adjacency (minimum of 2 neighbouring units) clusters were created and a cluster with a maximum sum of t-values was used in further statistics. Next, the CSD data was permuted between the groups and the statistics were calculated again. A total of 2000 randomizations was used and a permutation distribution of the maximum differences in the clusters was created. Further, the probability of deviating the data obtained in the study from the parameters of this distribution using the Monte Carlo method was done to calculate the significance level (p < 0.025). This approach allows us to exclude the acceptance of false positive results in multiple comparisons and to allocate clusters by time and space, based on the experimental data.

Source localization.
Based on the distribution of electric potential recorded on the scalp, the standardized low resolution brain electromagnetic tomography (sLORETA) software (publicly available free academic software at http://www.uzh.ch/keyinst/loreta.htm) was used to compute the cortical three-dimensional distribution of current density. sLORETA uses a distributed source localization algorithm to solve the inverse problem of brain electric activity regardless of the This study has limitations. First, the subjects' attention during the study was not controlled.
Using a motor response would provide additional information. Secondly, stimuli were presented in the nested design of the study, which increased the individualization of effects. This can be considered both a disadvantage and a merit of research. Thirdly, the calculation of the localization of distributed activity sources is more accurate with a larger number of electrodes.
Fourth, the significance of the face can also be controlled by recording autonomous reactivity, for example, a galvanic skin reaction or by changing the heart rate. Fifth, the study did not control the proximity and duration of friendships.

Conclusion
In this study, we studied the relationship between brain reactions in the perception of the subject's own face and the accuracy of perception of interception. Two main results were found.            Table 3. Correlations (Pearson product-moment correlation coefficients) between heartbeat perception (HBP) scores, within-cluster mean ERP amplitude for the Self-face perception and psychometric scales (p-value in parenthesis).