Our perceptions and decisions are not always objectively correct as they are featured by a bias related to our self. What are the behavioral, neural, and computational mechanisms of such cognitive bias? Addressing this yet unresolved question, we here investigate whether the cognitive bias is related to temporal integration and segregation as mediated by the brain’s Intrinsic neural timescales (INT). Using Signal Detection Theory (SDT), we operationalize the cognitive bias by the Criterion C as distinguished from the sensitivity index d’. This was probed in a self-task based on morphed self- and other faces. Behavioral data demonstrate clear cognitive bias, i.e., Criterion C. This was related to the EEG-based INT as measured by the autocorrelation window (ACW) in especially the cortical midline structures (CMS) of the transmodal default-mode network (DMN) (as distinct from unimodal visual cortex). Finally, simulation of the same paradigm in a large-scale network model shows high degrees of temporal integration of temporally distinct inputs in CMS/DMN while temporal segregation predominates in visual cortex. Together, we demonstrate a key role of INT-based temporal integration in DMN including its relation to the brain’s uni-transmodal topographical organization in mediating the cognitive bias of our self.
Article
Intrinsic neural timescales mediate the cognitive bias of self – Temporal integration in default-mode network
https://doi.org/10.21203/rs.3.rs-1761381/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
The self is a core feature of our mental life and plays an important role in perception, action, and cognition. The mediation of our personal preferences and the cognitive strategies adopted under uncertainty are key sources by which the self exerts a bias on our cognitive processes (Amodeo et al., 2021; Barton et al., 2021; Nijhof et al., 2020; Sparks et al., 2016; Sui & Humphreys, 2015). Humans are subjected to such cognitive bias in the sense that self-specific stimuli are differently treated compared to non-self-specific information (Christoff et al., 2011; Northoff, 2011; Northoff et al., 2006; Sui & Humphreys, 2015). In particular, self-specific stimuli induce a benefit in response time and precision in diverse tasks (Jiang et al., 2019; Sui et al., 2012) and are more easily remembered (for an overview, see Cunningham & Turk, 2017). On the neural side, several studies have linked self-referential processing to the cortical midline structures (CMS, Frewen et al., 2020; Kelley et al., 2002; Murray et al., 2012; Northoff & Bermpohl, 2004; Qin et al., 2020) and the default mode network (DMN, Frewen et al., 2020; Kolvoort et al., 2020; Qin & Northoff, 2011). What are the behavioral, neural and computational mechanisms by which the CMS/DMN mediate the cognitive bias? Addressing this yet unresolved question is the goal of our study.
Decision making is a complex process. When we have to make a decision, we base our judgment on objectively accurate external criteria but also on more subjective internal criteria, that is, our personal preferences (Nakao et al., 2013, 2019; Wolff et al., 2019). Signal detection theory (SDT) distinguishes between objective and more subjective criteria influencing the decision in forced choice tasks (Macmillan & Creelman, 2005). The general assumption of SDT is that decisions are made against a background of uncertainty, e.g., noise (Anderson, 2015); the goal of the decision-maker is to extract the decision signal from such background noise which may lead the subjects to introduce a bias in their decisions. This bias is operationalized in SDT by the Criterion C reflecting subjective, individual factors like personal preference. In contrast, the sensitivity measure d’ in SDT is an indicator of the external, objective criteria that indicate the discriminability between the two stimuli by the subject, thus reflecting a measure of accuracy.
Little is known about the neuro-temporal mechanisms underlying Criterion C and Sensitivity d’. Recent studies suggest a link between the Criterion C and spontaneous or ongoing activity like in the pre-stimulus neural activity of monkeys (van Vugt et al., 2018) and humans (Iemi et al., 2017; Iemi & Busch, 2018; Limbach & Corballis, 2016; Samaha et al., 2020). The apparent role of the spontaneous activity in mediating Criterion C suggests that the former’s Intrinsic Neural Timescales (INT) (Wolff et al., 2022) may be involved in mediating the cognitive bias. The INT are particularly long in the spontaneous activity of the cortical midline regions of the default-mode network (DMN) which, interestingly, are key regions in mediating self-specificity (Kolvoort et al., 2020; Smith et al., 2022; Wolff et al., 2019). Together, these observations suggest relationship of INT in CMS/DMN with the cognitive bias, that is, the Criterion C.
The INT are measured by the autocorrelation window (ACW) through determining the degree to which the brain’s activity correlates with itself over time (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Golesorkhi, Gomez-Pilar, Zilio, et al., 2021; Hasson et al., 2015; Honey et al., 2012; Wolff et al., 2022; Yeshurun et al., 2021). Constituting temporal windows, the INT play an important role in input processing and cognition driven through temporal integration and segregation (Himberger et al., 2018; Wolff et al., 2022; Yeshurun et al., 2021). Temporal integration occurs by temporally smoothing different inputs at different timepoints into one coherent source of information, e.g., summing different words to one sentence through a long INT (Himberger et al., 2018; Wolff et al., 2022). On the other hand, temporal segregation of the external stimuli features is mediated primarily by short ACW in unimodal sensory regions (i.e. primary visual cortex, etc.; Golesorkhi, Gomez-Pilar, Tumati, et al., 2021). Hence, the length of the ACW can be conceived as a proxy of the balance of temporal segregation and integration (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Golesorkhi, Gomez-Pilar, Zilio, et al., 2021; Himberger et al., 2018; Wolff et al., 2022). How such balance of temporal integration-segregation through long-short INT/ACW is related to the cognitive bias as measured by the SDT indices (and specifically Criterion C) remains yet unclear, though.
To clarify this relationship, we conducted a combined behavioral, neural, and computational study. Probing the cognitive bias, we used a modified version of a well-established task (Tsakiris, 2008) on visual perception of faces presented in a self-other continuum of different morphed pictures. Given the key importance of timescales with their supposed role in temporal integration-segregation, we presented the task in two temporally distinct ways (Huk et al., 2018). Once we presented it in a discontinuous or discrete way as typical event-related paradigm (separate single faces with different degrees of morphing in order to calculate Criterion C and sensitivity d’ based on the SDT). Additionally, we presented the same task also in a continuous way with an ongoing sequence of morphed faces where the participant had to indicate the transition between the two faces; this allowed us to specifically test for temporal integration-segregation within the paradigm itself, that is on the psychological level. EEG was recorded during both resting state and task-related activity and analysed on both channel and source space regional level. Finally, applying a data-driven network model (Chaudhuri et al., 2015), we included computational modelling to further support the assumption of different degrees of temporal integration and segregation in CMS/DMN and sensory regions respectively.
First, we will analyse the behavioral data by extracting the SDT indices and the transitions between self-other faces in both the discontinuous and continuous version of our task. We hypothesize a relation between the discontinuous and the continuous paradigm, that is a correlation between subjects’ bias (Criterion C) in the discontinuous paradigm with the subjects’ behavior in the continuous one. Given that the continuous version explicitly probes temporal integration versus segregation, it provides behavioral evidence for the involvement of temporal integration and segregation in the cognitive bias, i.e., Criterion C, and accuracy, i.e., sensitivity d’.
Next, we extract the ACW of EEG resting and task state activity in order to establish its link to the SDT indices. We expect the longer ACW (ACW0) to be implicated in the cognitive bias (Criterion C) while the shorter ACW (ACW5) supposedly mediates discrimination (sensitivity d’). We then also conducted source space reconstruction in EEG. This served to probe whether the cognitive bias, namely Criterion C, is related to the long ACW (ACW0) in the CMS/DMN (as being closely related to the self; Frewen et al., 2020; Northoff et al., 2006; Qin et al., 2020), while the sensitivity d’ supposedly is more associated with shorter timescales (ACW5) in sensory regions like primary visual cortex. Is CMS/DMN activity associated with a high degree of temporal integration of temporal input sequences? Applying the same temporal input structure as in our empirical-behavioral study, this will be probed in a computational realistic neural network model relying on Chaudhuri et al. (2015). This model reinforces the idea that the CMS/DMN’s neural activity show high degrees of temporal integration of distinct inputs across time, while primary visual cortex favors temporal segregation of temporally distinct inputs. That supports the assumption that the long ACW in CMS/DMN with its relation to the cognitive bias (Criterion C) are indeed related to temporal integration, that is, the temporal smoothing of temporally parsed inputs or stimuli across time.
Participants: Twenty-seven participants (15 women, 12 men, age: 22.18, SD: 4.28, range: 19 to 36) were recruited. None of them had a current or history of psychiatric or neurological disorders and normal or corrected-to-normal vision. We excluded one participant due to technical problems with EEG recordings in rest and task state and another subject was excluded only for the task state. Participants performed two morphed face recognition tasks adapted from (Tsakiris, 2008): In the discontinuous task, one subject was excluded. In the continuous task, one subject had no behavioural data due to a technical problem. Written informed consent for each participant was obtained prior to study participation. The study was approved by the local Research Ethics Board (REB # 2018054).
Materials: Prior to the experiment, pictures were taken from each of the participants with a Samsung A50 phone. Participants were asked to show a neutral facial expression, mouth closed. The pictures were then prepared in the free and open-source graphics editor GIMP (2.10.12). All faces were turned to black and white (grayscale) and cut in an oval circle, with only the eyes, nose, and mouth areas remaining on a black background. Then, each face was morphed with a same-sex and same-race face from the NimStim facial pictures set (Tottenham et al., 2009). Morphings were realized with the Abrosoft Fantamorph 5 Software in 1% steps (from 0% self to 100% self-face).
Continuous morphing task (Fig. 1A left): This task was constructed to be sensitive to the process of temporal integration and segregation. Recorded activity of this task reflect temporal integration on a psychological and neural level. A hundred and one morphed pictures in 1% morphing steps were extracted from the continuum between the own and the other’s face. A control continuum between a famous and an unknown face has also been created, leaving us with four morphing conditions based on the order and the identity depicted on the pictures. The ‘self’ conditions have the morphing order (1) from the ‘self’ face to other face (SO), or from (2) other face to ‘self’ face (OS). The control conditions have a morphing order from a (3) famous face to an unknown face (FU), or from the (4) unknown face to the famous face (UF).
The order of the conditions was randomised. Pictures were presented in a continuous manner to create the experience of a morphed movie between the two faces. Movie duration was between 10-15sec and the interstimulus interval varied between 4-6 sec. A total of 200 trials, 50 trials per condition in 8 blocks was presented. Participants were instructed to press a key when they stop to see the first face and to give another key when they started to see the second face. Of particular interest for us are the two self-related keypresses which indicate the emergence and the loss of the self. The emergence keypress is the second keypress when the pictures are presented in the order other-self, while the loosing keypress is the first keypress in the order self-other. In order to record neuronal activity without motor activity noise due to keypress (Tsuchiya et al., 2015, no-report paradigms), 5 trials in each condition were visualized and recorded while no keypress was given. The no-report trials took place before the report-trials for the half of participants, and afterwards for the other ones. To ensure that the moment of the keypress was time-independent, we randomly varied the movie length between 10 to 15 seconds. In addition, the intertrial interval was jittered between 4 to 6 sec (jittered by 1sec). This task has been recorded in EEG.
Discontinuous signal detection task): This task purely served to extract the SDT indices. From the morphed pictures, 13 pictures were selected at a morphing degree from 0% to 100% self in 10% steps. The middle morphing values (40 to 60%) are selected in 5% morphing steps to gather a more fine-grained behavior in the ambiguous range. Each picture was presented 12 times in a random order for 200ms, with a total 156 trials per subject (Fig. 1A right). Participants were instructed to classify each of the pictures in ‘self’ pictures (more self than the other) or in ‘other’ pictures (more other than self). No correct response was possible for the 50% morphing. We analysed the data with Signal Detection Theory (SDT) which offers cognitive bias (Criterion C) and sensitivity (d’) measures based on the relation of hits and false alarms.
Resting-state: Eyes open resting-state was recorded in EEG for 7min. The participant were fixating a white cross on a grey background.
Data analysis:
Signal detection theory (SDT): The discontinuous task was analysed with SDT (Macmillan & Creelman, 2005) because it yields separate indicators for the accuracy in form of the cognitive bias Criterion C and sensitivity d’ (Fig. 1B).
Criterion C: The criterion C of the signal detection theory (Macmillan & Creelman, 2005) indicates the preference of the subject to report rather one than the other of the two faces when committing an error (false alarms or misses). It is calculated based on the Hit and False alarm rate as follows:
Some subjects tend to perceive more often their own face then the other’s face (liberal response style), whereas other subjects rather see the ‘other’ face than themselves (conservative response style). Some participants show no bias. A C > 0 designates a preference to press the ‘other’ button (referred to as ‘other’ bias) whereas a C < 0 indicates a preference click the ‘self’ face button (here called ‘self’ bias). A C = 0 indicates an equal probability of pressing either of the buttons and therefore an absence of any bias.
Sensitivity d’: d' is a measure of discriminability and indicates the ability of a subject to distinguish between two stimuli. In our case, it indicates how well the subject distinguished between their own and someone else’s face. It is calculated based on the hit and false alarm rate.
If the hit rate equals the false alarm rate, d’ = 0 where the subject is unable to discriminate between the two stimuli. A higher d’ corresponds to less errors and therefore indicates a higher sensitivity of the subject.
EEG data acquisition and preprocessing: EEG data was recorded using Ag/AgCl electrodes through a 64-channel Brain Vision Easycap (according to the International 10-20 System) referenced to the right mastoid. The data was sampled at 1,000 Hz with DC recording. The EEG data preprocessing was performed using the EEGLAB toolbox for MATLAB (R2017b; (Delorme & Makeig, 2004), RRID:SCR_007292). The data was downsampled to 500 Hz and filtered with a FIR zero-phase low-pass filter at 50 Hz and a high-pass filter at 1 Hz. With a custom script, noisy channels (defined as 5 interquartile above or below each channels mean) have been spherically interpolated before re-referencing to the average. Further, artifacts were identified using independent component analysis (ICA) via the EEGLAB software (infomax) creating 63 independent components. Next, we used MARA implementation to automatically reject noisy components (Winkler et al., 2011).
Autocorrelation window (ACW): ACW5 is defined as full-width-at-half-maximum of the temporal autocorrelation function of each electrode. Golesorkhi et al. (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021) introduced a novel measure, the ACW0. This is the moment when the autocorrelation, instead of reaching a correlation of 0.5 as in the standard ACW (Honey et al., 2012; J. D. Murray et al., 2014), reaches the point where the autocorrelation equals zero. We calculated both ACW0 and ACW5 to see the impact of longer (ACW0) and shorter (ACW5) windows on the different behavioral indices, i.e., d’ and C. In addition, we investigated the gradient between ACW5 and ACW0 via the autocorrelation measures taken at different thresholds of autocorrelation (ACW4 for a correlation of 0.4, ACW3 for a correlation of 0.3, ACW2 for a correlation of 0.2 and ACW1 for a correlation of 0.1). All autocorrelations were calculated for each electrode or region of interest (ROI) before being averaged across all electrodes/ROIs to extract one ACW value per subject.
Source localisation with eLORETA: Regions were defined according to the Glasser parcellations (Glasser et al., 2016). Exact Low Resolution Electromagnetic Tomography (eLORETA, Pascual-Marqui et al., 1994) was performed by estimating the virtual-channel source for each region of the Glasser atlas in FieldTrip (Oostenveld et al., 2011). The output is a timeseries of estimated source level activity for each region. We investigated the core DMN (Andrews-Hanna et al., 2014) consisting of the posterior cingulate cortex (PCC) and the pregenual anterior cingulate cortex (pACC) because of its known implication in self-related functioning. As a control region that should not be implicated in self-related processing, we choose the visual primary cortex (V1). The precise Glasser-parcels used were the LV1, RV1, LRSC, RRSC, L23d, R23d, Ld23ab, Rd23ab,L23c, R23c, Rp24, Lp24, Ra24, La24, L31a, R31a, L31pv and the R31pv.
Statistical analyses: Comparison of the ACW durations based on the thresholds were performed via a repeated measures ANOVA with a Greenhouse-Geisser correction to correct for sphericity violations. The link between the behavioral and neuronal measures were realized via simple Pearson’s correlations. Inferential analyses were run in JASP. ACW was extracted in Python (3.8). SDT indices were calculated in Excel following the instructions in (Macmillan & Creelman, 2005).
Computational model: To further probe the effect of different regions on temporal segregation and integration, we used a computational model based on (Chaudhuri et al., 2015). In essence, this model is two differential equations:
is the firing rate, 
is the intrinsic time constant, IExt is the external input to the system governed by the slope b of the f-I curve. w values are coupling parameters. μ is a fixed parameter that controls the strength of long-range excitatory input. FLN (Fraction of Labeled Neurons) is the structural connectivity matrix based on a macaque study (Markov et al., 2014). E and I correspond to excitatory and inhibitory, respectively; i and j denotes different regions. Solving these differential equations give us 29 time series, one per region in the model. We used the exact same parameters as Chaudhuri et al. (2015). The hierarchy index 
is a sigmoid function scaled between 0 and 1, increasing from periphery to core (see Fig. 6). First, we used a continuous input of 1 second and plotted activity the same way (Fig. 6 left column). We also simulated the input as close as the real discontinuous paradigm in this study. We gave 200 ms inputs with amplitude 1 to the outermost region (V1) and randomized the inter-stimulus interval (ISI) between mean of the ISI in the study (0.528 s) +- 1 SD (0.253 s) and plotted the resulting neural activity changes in both periphery (V1) and core (24c) regions (Fig. 6A, left column). To probe the effect of the hierarchical uni-transmodal topography on temporal segregation and integration, we eliminated the hierarchy by setting the hierarhical index to 0.5 for all regions (Fig. 6, right column).
Behavioral results
We first conducted extensive behavioral testing using two morphed self-other face tasks. In one task, the morphed pictures were presented in a discontinuous event-related manner where subjects had to give response after each randomized presented picture; this enabled us to calculate SDT with discriminability (d’) and the cognitive bias (Criterion C). Additionally, we also presented the same paradigm in a more continuous movie-like version over 10-15s (jittered, see Methods) – this allows probing for temporal integration and segregation of the temporally continuous stimuli.
Extracting SDT indices (Discontinuous task): In the discontinuous self-task (Fig. 1A right, see Methods), subjects indicated the predominant facial identity (self or other) of single randomly presented morphed faces reflecting an event-related paradigm structure. The psychometric function (Fig. 2B) indicating the proportion of correct classifications for each morphing degree shows more variability in the middle morphings (where the subjective uncertainty is higher) than in the extreme ones (where the subjective uncertainty is lower). The Sensitivity index d’ had a mean (SD) of 2.65(0.52) (range: 1.6 to 3.79). The cognitive bias Criterion C had a mean (SD) of 0.25 (0.5) (range: -0.51 to 1.34). A higher d’ indicates a better distinction between the two faces (discriminability). A criterion C greater than 0 indicates a bias with preference to report ‘other’, while a low value (less than 0) indicates the subject’s bias or preference to report the ‘own’ face. A value equal to 0 indicates no bias (Fig. 2B).
Behavioral temporal integration (continuous task): While morphed faces with different morphing degrees were presented in a random order in the discontinuous task, the continuous version presents the morphing between self and other faces in a temporally continuous way, one face morphing slowly into the other face (see methods). Contrarily to the discontinuous task, such a continuous paradigm structure serves the purpose of probing for subjects’ capacity to integrate or segregate temporally disjointed but sequential inputs, i.e., the morphed self-other faces. Temporally adjacent stimuli like two morphed faces may be temporally integrated/separated entailing separate judgement as self or other; alternatively, they may be integrated entailing subsumption of both under either self or other. In the continuous paradigm, participants had to give a keypress when the first face disappeared and another keypress when the second face emerged. Two keypresses are of particular interest as they are the most self-related: In the morphing order from self to other (SO), the first keypress indicates the loss of the own face, while the second keypress in the order from other to self (OS) indicates the emergence or extraction of the own face.
For each subject, we calculated the average of the morphing degree at the moment a keypress occurred. The descriptive results are as follows: Loss of self: m = 50.77, SD = 7.88, range = 30.27, max = 60.33, min = 30.06; Emergence of other: m = 75.789, SD = 7.001, range = 28.31, max = 88.74, min = 60.43; Loss of other: m = 46.59, SD = 7.19, range = 24.38, max = 60.49, min = 36.11; Emergence of self: m = 71.8, SD = 8.33, range = 31.75, min = 54.86, max = 86.61; Fig. S1). The simultaneous effect revealed by a repeated ANOVA is significant (F(3,75) = 104.467, p < .001, h²patial = 0.807; eGG = 0.697), post hoc differences are reported in Fig. S1. We observe no effect specific to the order of morphing presentation (SO or OS).
Due to our particular interest in the self, we will specifically focus on its loss and its emergence. Hence, we correlated the keypresses (as obtained by the continuous task) with C and d’ (as obtained in the discontinuous task). We see that only the keypress related to the emergence of the self correlated with the Criterion C (r = 0.486, p = 0.014, Loss: r = -.275; p = .215; Fig. 2C) but not with the sensitivity d’ (Emergence: r = .150, p = .474; Loss: r = -.299, p = .147). This indicates that an early detection of the own face is linked to a liberal self-bias. A later self-emergence is linked to a conservative ‘other’-bias. Together, these findings strongly support that temporal integration is a key factor in mediating the bias Criterion C. We will next demonstrate the key role of temporal integration in the cognitive bias by investigating the ACW.
Intrinsic neural time scales in rest and task states
We assessed the intrinsic neural timescales (INT) by the autocorrelation window (ACW; Fig. 3A). The autocorrelation measures the similarity of a time series with a time-lagged version of itself. We distinguished, following Golesorkhi et al. 2021, ACW5 and ACW0. We defined ACW5 and ACW0 as the first lags where autocorrelation function (ACF) reaches half of its maximum value (AC = 0.5), and zero (AC = 0.0), respectively. Similarly, we further defined ACW4, ACW3, ACW2 and ACW1.
Employing these different ACWs enable us to investigate a more fine-grained temporal resolution. We expect shorter temporal windows (i.e. ACW5) to represent stronger degrees of temporal segregation while the long temporal windows of ACW (i.e. ACW0) are more related to high degrees of temporal integration.
The ACW was calculated on 7min EEG activity in each channel recorded during rest (eyes open) and the continuous task. In the following, we will describe the ACW by (i) investigating its topographical distribution and (ii) comparing rest and task state.
We investigated the topographic maps of ACW (Fig. 3B) which show the expected rostro-caudal gradient (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Golesorkhi, Gomez-Pilar, Zilio, et al., 2021; Kiebel et al., 2008). A shorter ACW can be observed in primary sensory areas like the temporal and occipital channels while a longer ACW is mostly observed in parietal and frontal channels.
We observe an overall significantly longer ACW during task compared to rest (Fig. 3C), indicating a sensitivity of the ACW to external stimuli. This is well in accordance with its supposed role in input processing (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Golesorkhi, Gomez-Pilar, Zilio, et al., 2021; Smith et al., 2022; Zilio et al., 2021). A longer ACW in task than in rest seems to be related to continuous paradigms which demand stronger temporal integration, while shorter ACW in task than in rest are found in event related paradigms which are demanding in temporal segregation (Smith et al., 2022; Wolff et al., 2022). Further, rest and task correlate highly (r > .85) with each other (Table S1), indicating a strongly shared or common neural basis between rest and task confirming previous data (see Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Smith et al., 2022).
Relationship of intrinsic neural timescales with SDT measures
After having described the ACW extracted from rest and task state, we will now investigate the link between the behavioral measures (self-related keypresses, d’ and Criterion C) and ACW. We expect the autocorrelation as indicator of temporal integration and segregation on the neural level (Wolff et al., 2022) to be linked to the behavioral indices of temporal integration. Given that both, perception-based Criterion C (Iemi et al., 2017; Iemi & Busch, 2018; Limbach & Corballis, 2016; Samaha et al., 2020; van Vugt et al., 2018) and self-consciousness (Davey et al., 2016; Huang et al., 2016; Qin & Northoff, 2011; Wolff et al., 2019), have been associated with the brain’s spontaneous activity, we correlated the Criterion C and Sensitivity d’ with both, resting state and task-related ACW.
First, we will focus on the link between the self-related keypresses of the continuous paradigm and the ACW measured in rest (see simple Pearson’s correlations in Table S2a) and task states (Table S2b). Particularly interesting is the keypress indicating the emergence of self, which shows a gradient with ACW (Fig. 4A). This is one of our key findings as it shows that the emergence of the self is stronger correlated with the longer ACW, which are indicative of stronger temporal integration. Hence, this results (i) confirms the integrative nature of the continuous task, and (ii) shows that it is the emergence, not the loss of the self, that is specifically linked to temporal integration and long ACW (Kolvoort et al., 2020; Smith et al., 2022; Wolff et al., 2019). If the emergence of the self is linked to temporal integration, we expect the cognitive bias in our self-task being linked to temporal integration (long ACW) as well.
We see the confirmation of this hypothesis in Fig. 4B, which depicts the link between d’ and Criterion C with the different ACW in table and graphical form. For d’, this figure clearly shows a range of weak correlations for longer ACWs (e.g., ACW0) and stronger correlations for shorter ones. In contrast, the opposite holds for the cognitive bias Criterion C, which more strongly correlated with the longer ACWs. While these opposite gradients of d’ and Criterion C were observed in resting state, the gradients are less pronounced in task state as here only a gradient between the ACW and Criterion C exists.
Together, these findings support the assumption that the Criterion C is related to higher degrees of temporal integration due to its link to longer ACW. The longer ACW increases the tendency of temporal smoothing and thus lumping together temporally disjointed inputs like the morphed faces. Due to its link to shorter ACW, sensitivity d’ is more linked to temporal segregation and the increased capacity to discriminate between sequential faces in a temporally precise way.
We conclude that for the Criterion C as well as the self-emergence keypress, a long ACW0 leads to a stronger other-perception, while shorter ACW leads to an earlier perception and extraction of the own face. On the other hand, the ACW5 plays a predominant role in the Sensitivity d’. Hence, the cognitive bias seems strongly linked to temporal integration while the discriminability is strongly linked to temporal segregation.
Source space reconstruction eLORETA
The self is known to be closely associated with the cortical midline structures (CMS) like anterior and posterior cingulate cortex of the default-mode network (DMN)(Frewen et al., 2020; Northoff, 2016; Northoff et al., 2006; Qin et al., 2020). Moreover, the CMS and DMN in general are known to exhibit the longest ACW compared to primary sensory regions (Golesorkhi, Gomez-Pilar, Tumati, et al., 2021; Golesorkhi, Gomez-Pilar, Zilio, et al., 2021; Ito et al., 2020; Raut et al., 2020; Wolff et al., 2022). How is our main finding (ACW0 linked to Criterion C and ACW5 linked to Sensitivity d’) represented in the DMN and the primary sensory regions? In other words, are the Criterion C and the d’ modulated by the long and short ACW in DMN and primary visual area?
This was probed by extracting ACW in source space of our EEG data. Hence, we extracted regional activity via eLORETA for the DMN focusing on CMS (PCC and pACC) and the primary visual area (V1; see Methods). ACW was calculated on the extracted DMN regions and correlated with d’ and Criterion C (Table 1). The Criterion C is only associated with the ACW0 in the DMN, measured during task-state (DMN-ACW0: r = 0.502, p = 0.012, Fig. 5A). Neither the V1 nor the ACW5 elicited a significant correlation with the Criterion C (p > .05). The Sensitivity index d’ is only correlated with the ACW5 in the resting state V1 activity (V1-ACW5: r = .469, p = .018, Fig. 5A). Neither the DMN nor the ACW0 showed significant correlation with the d’ (p > .05).
Together, these results confirm the specific link of the longer ACW0 in the CMS/DMN and Criterion C while the shorter ACW5 in V1 relates to Sensitivity d’. Further, the eLORETA results show that the CMS as core of the DMN are strongly implicated in the link between the cognitive bias, i.e., Criterion C and the longer ACW. This supports a special role of CMS/DMN in mediating specifically higher degrees of temporal integration underlying the cognitive bias in visual perception and subsequent decision making. The Sensitivity d’ instead is linked to the shorter timescales (ACW5) of the primary visual area favouring temporal segregation over integration.
|
State |
Crit C |
|
Sens d’ |
|||||
|
|
|
r |
|
p |
|
r |
|
p |
|
DMN_0 |
rest |
0.342 |
|
0.094 |
|
0.207 |
|
0.321 |
|
task |
0.502* |
|
0.012 |
|
0.188 |
|
0.38 |
|
|
DMN_5 |
rest |
0.308 |
|
0.143 |
|
0.339 |
|
0.106 |
|
task |
0.251 |
|
0.238 |
|
0.242 |
|
0.254 |
|
|
V1_0 |
rest |
-0.043 |
|
0.838 |
|
0.007 |
|
0.973 |
|
task |
0.387 |
|
0.062 |
|
0.11 |
|
0.608 |
|
|
V1_5 |
rest |
0.071 |
|
0.735 |
|
0.469* |
|
0.018 |
|
task |
0.283 |
|
0.191 |
|
0.306 |
|
0.155 |
|
Table 1. Correlations between Criterion C (Crit C) and Sensitivity d’ (Sens d’) in the DMN and the primary visual area V1 via the ACW0 (_0) and ACW5 (_5).
Computational Modeling of Temporal Integration and Segregation
The large-scale computational model we used in this study gave us the ability to probe the effect of different timescales on input processing (Chaudhuri et al., 2015). The hierarchy of different regions in the model was defined by a sigmoid function scaled between 0 and 1 (see methods). First, we applied the continuous task paradigm on the model to probe temporal segregation in hierarchically low (periphery / V1) and integration in hierarchically high (core / 24c) regions. As can be seen in A, our findings emphasize the importance of temporal segregation in sensory regions like V1 for processing disjoint external inputs in a temporally (and objectively) accurate way. On the other hand, the cortical midline region 24c integrates different temporally disjoint inputs, thus showing a high degree of temporal integration and low temporal segregation. These findings are well compatible with the empirical data from eLORETA.
In order to investigate the effect of the differential hierarchical positions of visual cortex and CMS/DMN in their short and long timescales, we eliminated the hierarchy in our computational model by assigning the same hierarchical index to all regions. We now observed similar activity in both V1 and midline 24c (Fig. 6, right column). This suggests that the different degrees of temporal integration and segregation as well as extension during input processing in unimodal regions like V1 and transmodal midline regions like 24c are, at least in part, related to their differential hierarchical positions in the brain’s overall uni-transmodal topography. This further underlines the key difference of transmodal CMS/DMN and unimodal V1 in their different timescales and ultimately their differential degrees of temporal integration and segregation of temporally disjoint inputs.
Further, we would expect that the neuronal activity during the discontinuous task shows temporal integration and segregation as we consider temporal integration as an intrinsic feature of the brain. This is confirmed by Chaudhuri’s model when simulating discontinuous input (Fig. 6B, left). As in the continuous task, the neural activity in response to an input is temporally extended in 24c whereas it has the same duration in V1. The temporal extension of the input is compromised once the hierarchical information is removed. Together, these results show that the underlying uni-transmodal structure of the brain is the base of temporal integration and segregation in both continuous and discontinuous versions of the paradigm – this suggests temporal integration-segregation to be an intrinsic feature of the brain (rather than being primarily related to the extrinsic or external input of the paradigm). Moreover, our findings highlight the important role of topography as without the uni-transmodal topography, temporal integration and subsequently the cognitive bias can no longer be observed in our model.
In the present study, we investigated whether temporal integration mediates the cognitive bias of self in perception and decision making. For this, we combined behavioral, neural and computational levels of analyses. Using Signal detection theory (SDT), we observed the cognitive bias on the behavioral level (Criterion C) with different subjects showing bias towards either the own or other face. Given the nature of our paradigm, this implies high temporal integration on the psychological level. That was further extended to the neural level by showing that the cognitive bias is specifically associated with longer INT (ACW0) in CMS/DMN, thus suggesting high temporal integration on the neural level. This is complemented by computational neural network modelling showing high degrees of temporal integration of temporally distinct inputs in specifically CMS/DMN as distinct from visual cortex, suggesting a strong implication of the uni-transmodal hierarchical topography in the cognitive bias.
We observe opposing correlational gradients between the bias (Criterion C) and the discrimination (d’) with the ACW. The ACW0 represents best the bias while the ACW5 better captures the Sensitivity d’. Hence, the bias in our cognition compared to a non-biased cognition occurs gradually with longer ACW and higher degrees of temporal integration. In contrast shorter ACW5 are more related to the sensitivity d’ reflecting high discrimination between the distinct stimuli, e.g., temporal segregation.
The differentiated implications of the ACW in the cognitive indices is further supported and extended on neuroanatomical grounds. Using eLORETA source space reconstruction, we observed a similar link of the cognitive bias (Criterion C) with the ACW0 and of the ACW5 with the d’. The topographical specifications through eLORETA show that the cognitive bias, that is Criterion C is processed in the DMN, while the discriminability d’ is more related to the primary visual area.
Our findings are consistent with the literature. The cortical midline structures (CMS) including the pACC and the PCC, are well known to be implicated in self-related processing (Davey et al., 2016; Frewen et al., 2020; Qin & Northoff, 2011). Spontaneous activity as for instance pre-stimulus activity in these regions predict the subsequent judgement of stimuli as self-related (Qin et al., 2016). Our data extend these results by showing that the spontaneous and task-related activity in these CMS regions mediate the cognitive bias through long INT (ACW0) and temporal integration.
Finally, we demonstrate that, on computational grounds, CMS (24c) regions do indeed show higher temporal integration in their response to both discontinuous and continuous input sequences. Specifically, CMS/DMN regions’ task-related responses are longer extending beyond the input duration as they temporally smooth and lump temporally disjoint inputs into one longer lasting activity change. In contrast, the primary visual cortex exhibits the opposite pattern with neural activity responses representing the single temporally parsed inputs thus suggesting a high degree of temporal segregation. These computational findings are well converging with our empirical observations and therefore further underline the key role of temporal integration in CMS/DMN mediating the cognitive bias (Criterion C). Given that these differential effects in uni- and transmodal regions when abolishing the hierarchy in our model, we suppose, albeit tentatively, a key role of the brain’s uni-transmodal topographic organization in mediating the cognitive bias (Criterion C).
Limitations
One potential limitation of our study is the novel use of ACW0 as distinct from ACW5 including their association with distinct cognitive features, namely sensitivity d’ and Criterion C. However, two previous studies (Golesorkhi et al. 2021a, Smith et al. 2022) also demonstrated differential association of ACW5 and ACW0 with neural, e.g., topographic distinction of transmodal DMN vs unimodal sensory regions, and psychological, e.g., self-continuity, features. This is well in line with our observation of especially the longer variant (ACW0) in trans-modal CMS/DMN being associated with Criterion C.
Yet another potential issue in our study concerns temporal integration. While especially the longer variant, ACW0, strongly suggests high degrees of temporal integration (Wolff et al. 2022), we did not show it in a more direct way in our empirical EEG data. However, various indirect lines strongly suggest temporal integration to be the main mechanism. First, especially the continuous version of our paradigm with the sequential self-other face morphing required temporal integration and segregation: if two sequential faces are associated with one and the same person, e.g., self or other, the two temporally distinct inputs (or stimuli/faces) must be integrated across their distinct time points. If, in contrast, they are attributed to distinct persons, e.g., self and other, they must be segregated. Temporal integration is thus required on the psychological level of the paradigm itself. Second, we demonstrate temporal integration on the computational level in our simulation where CMS/DMN, unlike visual cortex, do indeed integrate temporally parsed inputs into one activity change.
Finally, we demonstrate the importance of topographic hierarchy for temporal integration. That, however, was shown only in an indirect way, that is, by abolishing the uni-transmodal hierarchy in our computational model. Future studies in for instance patients with CMS/DMN lesions are warranted that relate different degrees of uni-transmodal topographic hierarchy with different degrees of both temporal integration and cognitive bias.
In conclusion, we demonstrate the key role of temporal integration for our cognitive bias in perception and decision making. We demonstrate temporal integration on the behavioral level using signal detection theory that allows distinguishing Criterion C (bias) and sensitivity d’ (accuracy). This is complemented on the neural level by showing that the cognitive bias is related to longer INT of especially CMS/DMN (ACW0). The involvement of temporal integration and the uni-transmodal topography is substantiated by computational modelling showing high degrees of temporal integration of temporally parsed inputs in the neural response of specifically CMS/DMN (as distinct from primary visual cortex). Together, we demonstrate that the cognitive bias in perception and decision making is intimately linked to the intrinsic neural timescales as based on the brain’s uni-transmodal topographical organisation.
Amodeo, L., Wiersema, J. R., Brass, M., & Nijhof, A. D. (2021). A comparison of self-bias measures across cognitive domains. BMC Psychology, 9(1), 132. https://doi.org/10.1186/s40359-021-00639-x
Anderson, N. D. (2015). Teaching signal detection theory with pseudoscience. Frontiers in Psychology, 6. https://doi.org/10.3389/fpsyg.2015.00762
Andrews-Hanna, J. R., Smallwood, J., & Spreng, R. N. (2014). The default network and self-generated thought: Component processes, dynamic control, and clinical relevance: The brain’s default network. Annals of the New York Academy of Sciences, 1316(1), 29–52. https://doi.org/10.1111/nyas.12360
Barton, T., Constable, M. D., Sparks, S., & Kritikos, A. (2021). Self-bias effect: Movement initiation to self-owned property is speeded for both approach and avoidance actions. Psychological Research, 85(4), 1391–1406. https://doi.org/10.1007/s00426-020-01325-0
Chaudhuri, R., Knoblauch, K., Gariel, M.-A., Kennedy, H., & Wang, X.-J. (2015). A Large-Scale Circuit Mechanism for Hierarchical Dynamical Processing in the Primate Cortex. Neuron, 88(2), 419–431. https://doi.org/10.1016/j.neuron.2015.09.008
Christoff, K., Cosmelli, D., Legrand, D., & Thompson, E. (2011). Specifying the self for cognitive neuroscience. Trends in Cognitive Sciences, 15(3), 104–112. https://doi.org/10.1016/j.tics.2011.01.001
Cunningham, S. J., & Turk, D. J. (2017). Editorial: A Review of Self-Processing Biases in Cognition. Quarterly Journal of Experimental Psychology, 70(6), 987–995. https://doi.org/10.1080/17470218.2016.1276609
Davey, C. G., Pujol, J., & Harrison, B. J. (2016). Mapping the self in the brain’s default mode network. NeuroImage, 132, 390–397. https://doi.org/10.1016/j.neuroimage.2016.02.022
Delorme, A., & Makeig, S. (2004). EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods, 134(1), 9–21. https://doi.org/10.1016/j.jneumeth.2003.10.009
Frewen, P., Schroeter, M. L., Riva, G., Cipresso, P., Fairfield, B., Padulo, C., Kemp, A. H., Palaniyappan, L., Owolabi, M., Kusi-Mensah, K., Polyakova, M., Fehertoi, N., D’Andrea, W., Lowe, L., & Northoff, G. (2020). Neuroimaging the consciousness of self: Review, and conceptual-methodological framework. Neuroscience & Biobehavioral Reviews, 112, 164–212. https://doi.org/10.1016/j.neubiorev.2020.01.023
Glasser, M. F., Coalson, T. S., Robinson, E. C., Hacker, C. D., Harwell, J., Yacoub, E., Ugurbil, K., Andersson, J., Beckmann, C. F., Jenkinson, M., Smith, S. M., & Van Essen, D. C. (2016). A multi-modal parcellation of human cerebral cortex. Nature, 536(7615), 171–178. https://doi.org/10.1038/nature18933
Golesorkhi, M., Gomez-Pilar, J., Tumati, S., Fraser, M., & Northoff, G. (2021). Temporal hierarchy of intrinsic neural timescales converges with spatial core-periphery organization. Communications Biology, 4(1), 277. https://doi.org/10.1038/s42003-021-01785-z
Golesorkhi, M., Gomez-Pilar, J., Zilio, F., Berberian, N., Wolff, A., Yagoub, M. C. E., & Northoff, G. (2021). The brain and its time: Intrinsic neural timescales are key for input processing. Communications Biology, 4(1), 970. https://doi.org/10.1038/s42003-021-02483-6
Gomez-Pilar, J., Gutiérrez-Tobal, G. C., Poza, J., Fogel, S., Doyon, J., Northoff, G., & Hornero, R. (2021). Spectral and temporal characterization of sleep spindles—Methodological implications. Journal of Neural Engineering, 18(3), 036014. https://doi.org/10.1088/1741-2552/abe8ad
Hasson, U., Chen, J., & Honey, C. J. (2015). Hierarchical process memory: Memory as an integral component of information processing. Trends in Cognitive Sciences, 19(6), 304–313. https://doi.org/10.1016/j.tics.2015.04.006
Hasson, U., Yang, E., Vallines, I., Heeger, D. J., & Rubin, N. (2008). A Hierarchy of Temporal Receptive Windows in Human Cortex. Journal of Neuroscience, 28(10), 2539–2550. https://doi.org/10.1523/JNEUROSCI.5487-07.2008
Himberger, K. D., Chien, H.-Y., & Honey, C. J. (2018). Principles of Temporal Processing Across the Cortical Hierarchy. Neuroscience, 389, 161–174. https://doi.org/10.1016/j.neuroscience.2018.04.030
Honey, C. J., Thesen, T., Donner, T. H., Silbert, L. J., Carlson, C. E., Devinsky, O., Doyle, W. K., Rubin, N., Heeger, D. J., & Hasson, U. (2012). Slow Cortical Dynamics and the Accumulation of Information over Long Timescales. Neuron, 76(2), 423–434. https://doi.org/10.1016/j.neuron.2012.08.011
Huang, Z., Obara, N., Davis, H. (Hap), Pokorny, J., & Northoff, G. (2016). The temporal structure of resting-state brain activity in the medial prefrontal cortex predicts self-consciousness. Neuropsychologia, 82, 161–170. https://doi.org/10.1016/j.neuropsychologia.2016.01.025
Huk, A., Bonnen, K., & He, B. J. (2018). Beyond Trial-Based Paradigms: Continuous Behavior, Ongoing Neural Activity, and Natural Stimuli. The Journal of Neuroscience, 38(35), 7551–7558. https://doi.org/10.1523/JNEUROSCI.1920-17.2018
Iemi, L., & Busch, N. A. (2018). Moment-to-Moment Fluctuations in Neuronal Excitability Bias Subjective Perception Rather than Strategic Decision-Making. Eneuro, 5(3), ENEURO.0430-17.2018. https://doi.org/10.1523/ENEURO.0430-17.2018
Iemi, L., Chaumon, M., Crouzet, S. M., & Busch, N. A. (2017). Spontaneous Neural Oscillations Bias Perception by Modulating Baseline Excitability. The Journal of Neuroscience, 37(4), 807–819. https://doi.org/10.1523/JNEUROSCI.1432-16.2016
Ito, T., Hearne, L. J., & Cole, M. W. (2020). A cortical hierarchy of localized and distributed processes revealed via dissociation of task activations, connectivity changes, and intrinsic timescales. NeuroImage, 221, 117141. https://doi.org/10.1016/j.neuroimage.2020.117141
Jiang, M., Wong, S. K. M., Chung, H. K. S., Sun, Y., Hsiao, J. H., Sui, J., & Humphreys, G. W. (2019). Cultural Orientation of Self-Bias in Perceptual Matching. Frontiers in Psychology, 10, 1469. https://doi.org/10.3389/fpsyg.2019.01469
Kelley, W. M., Macrae, C. N., Wyland, C. L., Caglar, S., Inati, S., & Heatherton, T. F. (2002). Finding the Self? An Event-Related fMRI Study. Journal of Cognitive Neuroscience, 14(5), 785–794. https://doi.org/10.1162/08989290260138672
Kiebel, S. J., Daunizeau, J., & Friston, K. J. (2008). A Hierarchy of Time-Scales and the Brain. PLoS Computational Biology, 4(11), e1000209. https://doi.org/10.1371/journal.pcbi.1000209
Kolvoort, I. R., Wainio‐Theberge, S., Wolff, A., & Northoff, G. (2020). Temporal integration as “common currency” of brain and self ‐ scale‐free activity in resting‐state EEG correlates with temporal delay effects on self‐relatedness. Human Brain Mapping, 41(15), 4355–4374. https://doi.org/10.1002/hbm.25129
Limbach, K., & Corballis, P. M. (2016). Prestimulus alpha power influences response criterion in a detection task: Prestimulus alpha power influences response. Psychophysiology, 53(8), 1154–1164. https://doi.org/10.1111/psyp.12666
Macmillan, N. A., & Creelman, C. D. (2005). Detection Theory: A User’s Guide—2nd Editio. https://www.routledge.com/Detection-Theory-A-Users-Guide/Macmillan-Creelman/p/book/9780805842319
Markov, N. T., Vezoli, J., Chameau, P., Falchier, A., Quilodran, R., Huissoud, C., Lamy, C., Misery, P., Giroud, P., Ullman, S., Barone, P., Dehay, C., Knoblauch, K., & Kennedy, H. (2014). Anatomy of hierarchy: Feedforward and feedback pathways in macaque visual cortex. Journal of Comparative Neurology, 522(1), 225–259. https://doi.org/10.1002/cne.23458
Murray, J. D., Bernacchia, A., Freedman, D. J., Romo, R., Wallis, J. D., Cai, X., Padoa-Schioppa, C., Pasternak, T., Seo, H., Lee, D., & Wang, X.-J. (2014). A hierarchy of intrinsic timescales across primate cortex. Nature Neuroscience, 17(12), 1661–1663. https://doi.org/10.1038/nn.3862
Murray, R. J., Schaer, M., & Debbané, M. (2012). Degrees of separation: A quantitative neuroimaging meta-analysis investigating self-specificity and shared neural activation between self- and other-reflection. Neuroscience & Biobehavioral Reviews, 36(3), 1043–1059. https://doi.org/10.1016/j.neubiorev.2011.12.013
Nakao, T., Matsumoto, T., Morita, M., Shimizu, D., Yoshimura, S., Northoff, G., Morinobu, S., Okamoto, Y., & Yamawaki, S. (2013). The Degree of Early Life Stress Predicts Decreased Medial Prefrontal Activations and the Shift from Internally to Externally Guided Decision Making: An Exploratory NIRS Study during Resting State and Self-Oriented Task. Frontiers in Human Neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00339
Nakao, T., Miyagi, M., Hiramoto, R., Wolff, A., Gomez-Pilar, J., Miyatani, M., & Northoff, G. (2019). From neuronal to psychological noise – Long-range temporal correlations in EEG intrinsic activity reduce noise in internally-guided decision making. NeuroImage, 201, 116015. https://doi.org/10.1016/j.neuroimage.2019.116015
Nijhof, A. D., Shapiro, K. L., Catmur, C., & Bird, G. (2020). No evidence for a common self-bias across cognitive domains. Cognition, 197, 104186. https://doi.org/10.1016/j.cognition.2020.104186
Northoff, G. (2011). Self and brain: What is self-related processing? Trends in Cognitive Sciences, 15(5), 186–187. https://doi.org/10.1016/j.tics.2011.03.001
Northoff, G. (2016). Is the self a higher-order or fundamental function of the brain? The “basis model of self-specificity” and its encoding by the brain’s spontaneous activity. Cognitive Neuroscience, 7(1–4), 203–222. https://doi.org/10.1080/17588928.2015.1111868
Northoff, G., & Bermpohl, F. (2004). Cortical midline structures and the self. Trends in Cognitive Sciences, 8(3), 102–107. https://doi.org/10.1016/j.tics.2004.01.004
Northoff, G., Heinzel, A., de Greck, M., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006). Self-referential processing in our brain—A meta-analysis of imaging studies on the self. NeuroImage, 31(1), 440–457. https://doi.org/10.1016/j.neuroimage.2005.12.002
Oostenveld, R., Fries, P., Maris, E., & Schoffelen, J.-M. (2011). FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience, 2011, 1–9. https://doi.org/10.1155/2011/156869
Pascual-Marqui, R. D., Michel, C. M., & Lehmann, D. (1994). Low resolution electromagnetic tomography: A new method for localizing electrical activity in the brain. International Journal of Psychophysiology, 18(1), 49–65. https://doi.org/10.1016/0167-8760(84)90014-X
Qin, P., Grimm, S., Duncan, N. W., Fan, Y., Huang, Z., Lane, T., Weng, X., Bajbouj, M., & Northoff, G. (2016). Spontaneous activity in default-mode network predicts ascription of self-relatedness to stimuli. Social Cognitive and Affective Neuroscience, 11(4), 693–702. https://doi.org/10.1093/scan/nsw008
Qin, P., & Northoff, G. (2011). How is our self related to midline regions and the default-mode network? NeuroImage, 57(3), 1221–1233. https://doi.org/10.1016/j.neuroimage.2011.05.028
Qin, P., Wang, M., & Northoff, G. (2020). Linking bodily, environmental and mental states in the self—A three-level model based on a meta-analysis. Neuroscience & Biobehavioral Reviews, 115, 77–95. https://doi.org/10.1016/j.neubiorev.2020.05.004
Raut, R. V., Mitra, A., Marek, S., Ortega, M., Snyder, A. Z., Tanenbaum, A., Laumann, T. O., Dosenbach, N. U. F., & Raichle, M. E. (2020). Organization of Propagated Intrinsic Brain Activity in Individual Humans. Cerebral Cortex, 30(3), 1716–1734. https://doi.org/10.1093/cercor/bhz198
Samaha, J., Iemi, L., Haegens, S., & Busch, N. A. (2020). Spontaneous Brain Oscillations and Perceptual Decision-Making. Trends in Cognitive Sciences, 24(8), 639–653. https://doi.org/10.1016/j.tics.2020.05.004
Smith, D., Wolff, A., Wolman, A., Ignaszewski, J., & Northoff, G. (2022). Temporal continuity of self: Long autocorrelation windows mediate self-specificity. NeuroImage, 119305. https://doi.org/10.1016/j.neuroimage.2022.119305
Sparks, S., Cunningham, S. J., & Kritikos, A. (2016). Culture modulates implicit ownership-induced self-bias in memory. Cognition, 153, 89–98. https://doi.org/10.1016/j.cognition.2016.05.003
Sui, J., He, X., & Humphreys, G. W. (2012). Perceptual effects of social salience: Evidence from self-prioritization effects on perceptual matching. Journal of Experimental Psychology: Human Perception and Performance, 38(5), 1105–1117. https://doi.org/10.1037/a0029792
Sui, J., & Humphreys, G. W. (2015). The Integrative Self: How Self-Reference Integrates Perception and Memory. Trends in Cognitive Sciences, 19(12), 719–728. https://doi.org/10.1016/j.tics.2015.08.015
Tsakiris, M. (2008). Looking for Myself: Current Multisensory Input Alters Self-Face Recognition. PLoS ONE, 3(12), e4040. https://doi.org/10.1371/journal.pone.0004040
Tsuchiya, N., Wilke, M., Frässle, S., & Lamme, V. A. F. (2015). No-Report Paradigms: Extracting the True Neural Correlates of Consciousness. Trends in Cognitive Sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002
van Vugt, B., Dagnino, B., Vartak, D., Safaai, H., Panzeri, S., Dehaene, S., & Roelfsema, P. R. (2018). The threshold for conscious report: Signal loss and response bias in visual and frontal cortex. Science, 360(6388), 537–542. https://doi.org/10.1126/science.aar7186
Winkler, I., Haufe, S., & Tangermann, M. (2011). Automatic Classification of Artifactual ICA-Components for Artifact Removal in EEG Signals. Behavioral and Brain Functions, 7(1), 30. https://doi.org/10.1186/1744-9081-7-30
Wolff, A., Berberian, N., Golesorkhi, M., Gomez-Pilar, J., Zilio, F., & Northoff, G. (2022). Intrinsic neural timescales: Temporal integration and segregation. Trends in Cognitive Sciences, 26(2), 159–173. https://doi.org/10.1016/j.tics.2021.11.007
Wolff, A., Di Giovanni, D. A., Gómez‐Pilar, J., Nakao, T., Huang, Z., Longtin, A., & Northoff, G. (2019). The temporal signature of self: Temporal measures of resting‐state EEG predict self‐consciousness. Human Brain Mapping, 40(3), 789–803. https://doi.org/10.1002/hbm.24412
Yeshurun, Y., Nguyen, M., & Hasson, U. (2021). The default mode network: Where the idiosyncratic self meets the shared social world. Nature Reviews Neuroscience, 22(3), 181–192. https://doi.org/10.1038/s41583-020-00420-w
Zilio, F., Gomez-Pilar, J., Cao, S., Zhang, J., Zang, D., Qi, Z., Tan, J., Hiromi, T., Wu, X., Fogel, S., Huang, Z., Hohmann, M. R., Fomina, T., Synofzik, M., Grosse-Wentrup, M., Owen, A. M., & Northoff, G. (2021). Are intrinsic neural timescales related to sensory processing? Evidence from abnormal behavioral states. NeuroImage, 226, 117579. https://doi.org/10.1016/j.neuroimage.2020.117579
There is NO Competing Interest.
- SupplementsWolman.docx
Supplementary information