scispace - formally typeset
Search or ask a question
Posted ContentDOI

A causal test of the mechanisms by which affect state biases affective perception

15 Jan 2021-bioRxiv (Cold Spring Harbor Laboratory)-
TL;DR: In this article, the role of self-induced affect states in biasing the affective perception of subsequent image stimuli was investigated. And they found that self-induction of a positive affective valence state significantly positively biased the perceived valence of subsequent stimuli.
Abstract: In this study, we merged methods from machine learning and human neuroimaging to causally test the role of self-induced affect states in biasing the affective perception of subsequent image stimuli. To test this causal relationship, we developed a novel paradigm in which (n=40) healthy adult participants observed multivariate neural decodings of their real-time functional magnetic resonance image (rtfMRI) responses as feedback to guide explicit regulation of their brain (and corollary affect processing) state towards a positive valence goal state. By this method, individual differences in affect regulation ability were controlled. Attaining this brain-affect goal state triggered the presentation of pseudo-randomly selected affectively congruent (positive valence) or incongruent (negative valence) image stimuli drawn from the International Affective Picture Set. Separately, subjects passively viewed randomly triggered positively and negatively valenced image stimuli during fMRI acquisition. Multivariate neural decodings of the affect processing induced by these stimuli were modeled using the task trial type (state-versus randomly-triggered) as the fixed-effect of a general linear mixed effects model. Random effects were modeled subject-wise. We found that self-induction of a positive affective valence state significantly positively biased the perceived valence of subsequent stimuli. As a manipulation check, we validated affective state induction achieved by the image stimuli using independent psychophysiological response measures of hedonic valence and autonomic arousal. We also validated the predictive fidelity of the trained neural decoding models for brain states induced by an out-of-sample set of image stimuli. Beyond its contribution to our understanding of the neural mechanisms that bias affect processing, this work demonstrated the viability of novel experimental paradigms triggered by pre-defined affective cognitive states. This line of individual differences experimentation potentially provides scientists with a valuable tool for causal exploration of the roles and identities of intrinsic cognitive processing mechanisms that shape our perceptual processing of sensory stimuli.

Summary (2 min read)

Jump to: [Introduction][Methods][Results][Discussion] and [Conclusion]

Introduction

  • The authors capacity to process and regulate emotions is central to their ability to optimize psychosocial functioning and quality of life1.
  • As a corollary, disruptions in emotion processing and regulation are broadly ascribed to psychiatric illnesses including borderline personality disorder, depression, anxiety disorders, PTSD, and substance-use disorders2 which negatively impact quality of life and functioning3,4.
  • Within BCI studies, neurophysiological and psychological variables (e.g., self-confidence and concentration) were shown to significantly predict performance variation15–17.
  • Very little is known about the source of individual differences in the ability to volitionally regulate affective states.
  • The authors then compared image stimulus-cued brain and affective responses arising from explicitly self-induced feedback-facilitated positive valence states versus random affective states (passive viewing) and causally tested the ability of self-induced positive valence states to bias the affect processing of subsequent image stimuli.

Methods

  • All participants provided written informed consent after receiving written and verbal descriptions of the study procedures, risks, and benefits.
  • Identification (Id) task acquisition consisted of 2 x 9.4 min fMRI scans during which the participant was presented with 120 images drawn from the International Affective Picture System18 (IAPS) to support one of two trial types : 90 passive stimulus (PS) trials and 30 cued-recall (CR) trials.
  • The initial hyperplane distance threshold was fixed for 20 seconds.
  • The authors performed this encoding process separately for each dimension of affect processing (valence and arousal).
  • Psychophysiology Data Acquisition and Preprocessing All MRI acquisitions included concurrent psychophysiological recordings conducted using the BIOPAC MP150 Data Acquisition System and AcqKnowledge software combined with the EDA100C-MRI module (skin conductance), TSD200-MRI pulse plethysmogram (heart rate), TSD221-MRI belt , and EMG100C-MRI module (facial electromyography).

Results

  • Psychophysiological Response Validation of Affect Processing Induction via Image Stimuli.
  • Slope and intercept random-effects were modeled subject-wise.
  • These results support the validity of their neural decoding models as brain representations of affective valence and arousal.
  • The authors modeled the neurally decoded valence of the final volume of the selfinduce step of Mod-FS trials (see Fig 2) as a function of the individual subjects’ explicit affect regulation performance parameters (slope and intercept, respectively, for the valence and arousal properties of affect processing) controlling for the subjects’ age and sex.

Discussion

  • Therefore, their use of real-time fMRI-based affective decodings to guide (or focus) this innate process enabled us to test (using unguided explicit affect regulation ability as a baseline) the association between innate affect regulation performance and the performance achievable using their real-time fMRI feedback approach.
  • Moreover, their small study sample did not permit sufficient piloting of parameters prior to selecting the processing design and testing.
  • Further, their analysis included all rtfMRI-guided self-induction trials, even those that required emergency triggering due to a failure to meet the design criteria of the goal state.
  • Therefore, the authors believe the performance of the system, and its effect sizes, are understated, which suggests the potential to further refine this technology for larger-scaled deployment of brain-state driven experiment designs to causally test interactions between internal cognitions and external stimuli.

Conclusion

  • The authors combined established neural decoding methods with real-time fMRI to construct a dynamic experimental design in which the brain representation of a subject’s self-induced positive affect state triggered the randomized presentation of affectively congruent or incongruent image stimuli.
  • The authors first validated the experiment’s ability to induce affect processing with independent measures of psychophysiology as well as the decoding models’ ability to predict affect processing in novel task domains.
  • The authors then demonstrated that self-induced positive affective states positively bias the affect processing of subsequent image stimuli and thereby furnish a causal mechanism by which positive thinking influences how they perceive their environment.

Did you find this useful? Give us your feedback

Figures (7)

Content maybe subject to copyright    Report

A causal test of affect processing bias in response to affect regulation.
Authors
Keith A. Bush*, Clinton D. Kilts
Affiliation
Brain Imaging Research Center, Department of Psychiatry, University of Arkansas for Medical
Sciences, Little Rock, AR 72205
Address correspondence to:
Keith A. Bush, Ph.D.
Brain Imaging Research Center
Department of Psychiatry
University of Arkansas for Medical Sciences
4301 W. Markham St. #554
Little Rock, AR 72205
Email: kabush@uams.edu
Abstract
In this study we merged methods from machine learning and human neuroimaging to causally
test the role of self-induced affect processing states in biasing the affect processing of subsequent
image stimuli. To test this causal relationship we developed a novel paradigm in which (n=40)
healthy adult participants observed affective neural decodings of their real-time functional
magnetic resonance image (rtfMRI) responses as feedback to guide explicit regulation of their
brain (and corollary affect processing) state towards a positive valence goal state. By this method
individual differences in affect regulation ability were controlled. Attaining this brain-affect goal
state triggered the presentation of pseudo-randomly selected affectively congruent (positive
valence) or incongruent (negative valence) image stimuli drawn from the International Affective
Picture Set. Separately, subjects passively viewed randomly triggered positively and negatively
valent image stimuli during fMRI acquisition. Multivariate neural decodings of the affect
processing induced by these stimuli were modeled using the task trial type (state- versus
randomly-triggered) as the fixed-effect of a general linear mixed-effects model. Random effects
were modeled subject-wise. We found that self-induction of a positive valence brain state
significantly positively biased valence processing of subsequent stimuli. As a manipulation check,
we validated affect processing state induction achieved by the image stimuli using independent
psychophysiological response measures of hedonic valence and autonomic arousal. We also
validated the predictive fidelity of the trained neural decoding models using brain states induced
by an out-of-sample set of image stimuli. Beyond its contribution to our understanding of the
neural mechanisms that bias affect processing this work demonstrated the viability of novel
experimental paradigms triggered by pre-defined cognitive states. This line of individual
differences research potentially provides neuroimaging scientists with a valuable tool for causal
exploration of the roles and identities of intrinsic cognitive processing mechanisms that shape our
perceptual processing of sensory stimuli.
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted May 20, 2021. ; https://doi.org/10.1101/2021.01.13.426569doi: bioRxiv preprint

Introduction
Our capacity to process and regulate emotions is central to our ability to optimize psychosocial
functioning and quality of life
1
. As a corollary, disruptions in emotion processing and regulation
are broadly ascribed to psychiatric illnesses including borderline personality disorder, depression,
anxiety disorders, PTSD, and substance-use disorders
2
which negatively impact quality of life and
functioning
3,4
. In light of this, a primary focus of cognitive behavioral therapy (CBT), an efficacious
treatment for disorders involving emotion dysregulation
5
, is the development of mental strategies
for identifying and volitionally reducing negatively biased emotional states that are the product of
maladaptive emotion processing and regulation. Neuroimaging has provided insight into the
functional neurocircuits involved in CBT-based emotion regulation strategies
6
; however, the
causal neurobiological mechanisms by which these strategies induce adaptive emotion
processing over time remain elusive.
Research into the effects of temporal context on affect and emotion processing may have
implications for increasing our understanding of the neural bases of emotion regulation. Prior work
has demonstrated that changing affective context prior to an emotional target shapes the
processing of that target. Such priming effects both accelerate and weaken the emotional
response to affectively congruent target stimuli
7
. Manipulations of affect processing state impact
the temporal structure of the neural responses to subsequent affective image stimuli
8
as well as
the corollary psychophysiological responses to those stimuli
9,10
. Further, stimulus-cued emotion
processing states bias the self-reported perception of successive emotional stimuli
11
.
These findings are consistent with effects that would be predicted by the deployment of
situational and attentional modification strategies according to the process model of emotion
regulation
12
and point to potential underlying mechanisms driving CBT-related changes to
emotion processing and thus its therapeutic efficacy. However, the neural representation of the
observed ability of affective cognitions related to these strategies to bias subsequent emotional
responses has not yet been causally tested. Thus, the primary aim of this work was to contribute
to our knowledge of the mechanisms underlying emotion regulation (operationalized as affect
regulation) by experimentally demonstrating that self-induced and verified affect processing
states causally bias the affect processing of subsequent image stimuli.
Real-time functional magnetic resonance imaging (rtfMRI), when used to generate brain
activation feedback
13
(i.e., rtfMRI-guided neuromodulation or neurofeedback), reflects a
promising methodology that has not to our knowledge been applied for mechanistic testing of how
the neural correlates of such feedback-induced affect processing states causally bias subsequent
affect processing. Here, the applied advantage of rtfMRI is that self-induced neurocognitive states
(achieved via rtfMRI guidance) can be verified and used as independent experimental variables
to trigger subsequent affective stimulus-response characterizations. Yet, a challenge to rtfMRI-
guided neuromodulation studies, and brain computer interface (BCI) research in general, is the
large individual variation observed in subjects’ ability to volitionally modulate their cognitive states
the well-known “BCI-illiteracy phenomenon”
14
.
Within BCI studies, neurophysiological and psychological variables (e.g., self-confidence
and concentration) were shown to significantly predict performance variation
1517
. However, very
little is known about the source of individual differences in the ability to volitionally regulate
affective states. Therefore, the secondary aim of this project was to characterize individual
variation in the ability to self-induce affective states using neurofeedback according to the
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted May 20, 2021. ; https://doi.org/10.1101/2021.01.13.426569doi: bioRxiv preprint

subjects’ unguided self-induction ability. This research has direct clinical relevance to informing
our understanding of the neuroregulation capabilities of psychiatric patients to identify those most
or least capable of guided affect regulation.
To explore our aims, we developed a novel task in which healthy adult participants utilized
rtfMRI feedback to explicitly regulate their brain response and corollary affect processing states
toward a goal of extreme pleasantness (i.e., positive valence). Attaining this brain-affect state
triggered the presentation of an affectively congruent (positive valence) or incongruent (negative
valence) image stimulus drawn from the International Affective Picture Set
18
(IAPS). Between
regulation trials participants passively viewed (without regulation) IAPS stimuli associated with
either positive or negative valence. We then compared image stimulus-cued brain and affective
responses arising from explicitly self-induced feedback-facilitated positive valence states versus
random affective states (passive viewing) and causally tested the ability of self-induced positive
valence states to bias the affect processing of subsequent image stimuli.
Our results reveal that self-induction of positive valence causally biases affect processing
responses to image stimuli, suggesting a potential mechanism by which CBT-based mental
strategies may work to reduce negatively biased affect processing states. However, we also found
that individual differences in the intrinsic ability to precisely self-induce affect processing states
without guidance did not generalize to the achievement of self-induced positive valence in the
presence of rtfMRI-feedback, suggesting inherent affect regulation ability that is separate from
concentration, e.g., the ability to accurately perceive temporally proximal affect processing states
or temporally distal goal states. Additional research will be necessary to characterize the latent
neurobiological and psychological factors driving these individual differences.
Methods
Ethics Statement
All participants provided written informed consent after receiving written and verbal descriptions
of the study procedures, risks, and benefits. We performed all study procedures and analysis with
approval and oversight of the Institutional Review Board at the University of Arkansas for Medical
Sciences (UAMS) in accordance with the Declaration of Helsinki and relevant institutional
guidelines and policies.
Participants
We enrolled healthy adult participants (n=40) having the following demographic characteristics:
age [mean(s.d.)]: 38.8(13.3), range 20‒65; sex: 22 (55%) female; race/ethnicity: 28 (70.%) self-
reporting as White or Caucasian, 9 (22.5%) as Black or African-American, 1 (2.5%) as Asian, and
2 (5%) self-reporting as other; education [mean(s.d.)]: 16.8(2.2) years, range 12‒23; WAIS-IV IQ
[mean(s.d.)]: 102.5(15.3), range 73‒129. All of the study’s participants were right-handed
(assessed via Edinburgh Handedness Inventory
19
) native-born United States citizens who were
medically healthy and exhibited no current Axis I psychopathology, including mood disorders, as
assessed by the SCID-IV clinical interview
4
. All participants reported no current use of
psychotropic medications and produced a negative urine screen for drugs of abuse (cocaine,
amphetamines, methamphetamines, marijuana, opiates, and benzodiazepines) immediately prior
to both the clinical interview and MRI scan. When indicated, we corrected participants’ vision to
20/20 using an MRI compatible lens system (MediGoggles™, Oxforshire, United Kingdom), and
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted May 20, 2021. ; https://doi.org/10.1101/2021.01.13.426569doi: bioRxiv preprint

we excluded all participants endorsing color blindness.
Experiment Design.
Following the provision of informed consent, subjects visited the Brain Imaging Research Center
of the University of Arkansas for Medical Sciences on two separate days. On Study Day 1 a
trained research assistant assessed all subjects for major medical and psychiatric disorders as
well as administered instruments to collect the following data to be used as either secondary
variables hypothesized to explain individual variance in affect regulation-related neural activity,
covariates of no interest, or to assess inclusion/exclusion criteria. The participant returned to the
BIRC for Study Day 2 within 30 days after Study Day 1 to complete the MRI acquisition. During
this day, the participant received task training and completed the full MRI acquisition protocol,
depicted in Figure 1.
Figure 1: Study Day 2 Experimental tasks: order, number of repetitions, duration, and stimuli.
Tasks are colored by role. Gray depicts task training and application of psychophysiology
recording apparatus. Blue depicts brain structural image acquisition. Orange depicts functional
image acquisition. Identification and Modulation blocks of the fMRI acquisition summarize the
relevant trial types used within that task (see Neuroimaging section for abbreviations). *Training
of real-time multivariate pattern analysis predictive models was performed concurrently with the
Resting State task of the fMRI acquisition.
Training: Each participant received a video-based overview of the experiment to be
performed on that day as well as training on the study’s task variations and trial types. The
participant was offered the opportunity to use the restroom and then was moved to the MRI
scanner room and fully outfitted with psychophysiological recording equipment.
Neuroimaging: For each subject we captured a registration scan and detailed T1-weighted
structural image. We then acquired functional MRI data for three task variations: identification,
resting state, and modulation. Identification (Id) task acquisition consisted of 2 x 9.4 min fMRI
scans during which the participant was presented with 120 images drawn from the International
Affective Picture System
18
(IAPS) to support one of two trial types (see Figure 2): 90 passive
stimulus (PS) trials and 30 cued-recall (CR) trials. Identification task PS trials (abbreviated Id-PS)
presented an image for 2 s (cue) succeeded by a fixation cross for a random inter-trial interval
(ITI) sampled uniformly from the range 26 s. Identification task cued-recall (Id-CR) trials were
multi-part: a cue image was presented for 2 s followed by an active cue response step for 2 s (the
word “FEEL” overlaying the image) followed by the word FEEL alone for 8 s, which signaled the
participant to actively recall and re-experience the affective content of the cue image, followed by
a 26 s ITI. During pre-scan training on the Id-CR task’s recall condition, subjects were instructed
to “Imagine the last picture you saw as best you can. Try to make yourself feel exactly how you
felt when you saw this picture the first time. Hold that feeling the whole time you see the word
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted May 20, 2021. ; https://doi.org/10.1101/2021.01.13.426569doi: bioRxiv preprint

FEEL.” Within each scan, Id-PS and Id-CR trials were pseudo-randomly sequentially ordered to
minimize correlations between the hemodynamic response function (HRF)-derived regressors of
the tasks. This order was fixed for all subjects.
During resting state acquisition, we acquired 7.5 min of fMRI data in which the subject
performed mind-wandering with eyes open while observing a fixation cross. During training,
subjects were instructed to “Keep your eyes open, look at the cross in front of you, and let your
brain think whatever it wants to.” Concurrently with the resting state task, the real-time variant of
the multivoxel pattern analysis (MVPA) prediction model (see below) was fit using data drawn
from the Identification task fMRI data to define individual brain state representations of the affect
processing goal.
Modulation (Mod) task acquisition consisted of 2 x 10.5 min fMRI scans during which the
participant was presented with 60 IAPS images according to two trial types (see Fig 2): 40 passive
stimulus (Mod-PS) trials, which were identically formatted to the Id-PS trials, and 20 feedback-
triggered stimulus (Mod-FS) trials. Mod-FS trials used real-time fMRI feedback of the subject’s
decoded affective state to guide them in self-inducing affective brain states associated with their
individualized representation of extreme positive valence. The computer system monitored the
subject’s decoded valence processing level at each acquisition volume of fMRI data and if that
decoding met pre-defined criteria (i.e., the goal state, which we defined as hyperplane distance
0.8 for 4 consecutive EPI volumes) then a positively (congruent) or negatively (incongruent) valent
image stimulus was triggered as the test stimulus. The brain state criteria representing the affect
processing goal state were determined by the results of an initial pilot of the experiment to identify
acquisition parameters that were challenging but consistently reachable. Within each scan, Mod-
PS and Mod-FS trials were pseudo-randomly sequentially ordered to minimize correlations
between the hemodynamic response function (HRF)-derived regressors of the tasks. This order
was fixed for all subjects.
We provided real-time visual feedback during Mod-FS trials by manipulating the level of
transparency of the word FEEL, which was the cue to volitionally regulate affect to an extreme
positive valence. The transparency of the text was scaled to reflect real-time estimates of subject’s
represented valence processing with respect to the desired hyperplane distance threshold. This
was achieved by mapping MVPA prediction model hyperplane distances (see below) from their
base range [-1.25,1.25] to the range of possible transparencies, α ϵ [0,1]. Fully transparent text
(α=0) appeared as a black screen and denoted poor affect regulation performance, i.e., highly
negative valence. Fully opaque text (α=1) appeared bright yellow and denoted good performance.
The transparency of the text was reset every 2 s (reflecting the momentary hyperplane distance
prediction based upon each EPI volume, TR=2000 ms). The transparency was adjusted
(approximately 20 frames-per-second) to present smooth transitions toward the brain-affect goal
state. The initial hyperplane distance threshold was fixed for 20 seconds. If the subject had not
attained the threshold (i.e. triggered the test stimulus) by this time then the threshold was linearly
and continuously lowered to 0 over the subsequent 18 s at which point the stimulus was
automatically triggered even if the threshold had not been attained (Fig. 2).
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted May 20, 2021. ; https://doi.org/10.1101/2021.01.13.426569doi: bioRxiv preprint

References
More filters
Journal ArticleDOI
TL;DR: An inventory of 20 items with a set of instructions and response- and computational-conventions is proposed and the results obtained from a young adult population numbering some 1100 individuals are reported.

33,268 citations

01 Jan 1970
TL;DR: The STAI as mentioned in this paper is an indicator of two types of anxiety, the state and trait anxiety, and measure the severity of the overall anxiety level, which is appropriate for those who have at least a sixth grade reading level.
Abstract: The STAI serves as an indicator of two types of anxiety, the state and trait anxiety, and measure the severity of the overall anxiety level.The STAI, which is appropriate for those who have at least a sixth grade reading level, contains four-point Likert items. The instrument is divided into two sections, each having twenty questions. Approximately 15 minutes are required for adults to complete the both STAI. The number on the scale is positively correlated to the anxiety related to in the question.

24,997 citations

Proceedings ArticleDOI
01 Jul 1992
TL;DR: A training algorithm that maximizes the margin between the training patterns and the decision boundary is presented, applicable to a wide variety of the classification functions, including Perceptrons, polynomials, and Radial Basis Functions.
Abstract: A training algorithm that maximizes the margin between the training patterns and the decision boundary is presented. The technique is applicable to a wide variety of the classification functions, including Perceptrons, polynomials, and Radial Basis Functions. The effective number of parameters is adjusted automatically to match the complexity of the problem. The solution is expressed as a linear combination of supporting patterns. These are the subset of training patterns that are closest to the decision boundary. Bounds on the generalization performance based on the leave-one-out method and the VC-dimension are given. Experimental results on optical character recognition problems demonstrate the good generalization obtained when compared with other learning algorithms.

11,211 citations

Journal ArticleDOI
TL;DR: Although mental disorders are widespread, serious cases are concentrated among a relatively small proportion of cases with high comorbidity, as shown in the recently completed US National Comorbidities Survey Replication.
Abstract: Background Little is known about the general population prevalence or severity of DSM-IV mental disorders. Objective To estimate 12-month prevalence, severity, and comorbidity of DSM-IV anxiety, mood, impulse control, and substance disorders in the recently completed US National Comorbidity Survey Replication. Design and Setting Nationally representative face-to-face household survey conducted between February 2001 and April 2003 using a fully structured diagnostic interview, the World Health Organization World Mental Health Survey Initiative version of the Composite International Diagnostic Interview. Participants Nine thousand two hundred eighty-two English-speaking respondents 18 years and older. Main Outcome Measures Twelve-month DSM-IV disorders. Results Twelve-month prevalence estimates were anxiety, 18.1%; mood, 9.5%; impulse control, 8.9%; substance, 3.8%; and any disorder, 26.2%. Of 12-month cases, 22.3% were classified as serious; 37.3%, moderate; and 40.4%, mild. Fifty-five percent carried only a single diagnosis; 22%, 2 diagnoses; and 23%, 3 or more diagnoses. Latent class analysis detected 7 multivariate disorder classes, including 3 highly comorbid classes representing 7% of the population. Conclusion Although mental disorders are widespread, serious cases are concentrated among a relatively small proportion of cases with high comorbidity.

10,951 citations

Journal ArticleDOI
TL;DR: A package of computer programs for analysis and visualization of three-dimensional human brain functional magnetic resonance imaging (FMRI) results is described and techniques for automatically generating transformed functional data sets from manually labeled anatomical data sets are described.

10,002 citations

Frequently Asked Questions (1)
Q1. What contributions have the authors mentioned in the paper "A causal test of affect processing bias in response to affect regulation" ?

In this study the authors merged methods from machine learning and human neuroimaging to causally test the role of self-induced affect processing states in biasing the affect processing of subsequent image stimuli. Beyond its contribution to their understanding of the neural mechanisms that bias affect processing this work demonstrated the viability of novel experimental paradigms triggered by pre-defined cognitive states. This line of individual differences research potentially provides neuroimaging scientists with a valuable tool for causal exploration of the roles and identities of intrinsic cognitive processing mechanisms that shape their perceptual processing of sensory stimuli. CC-BY-NC-ND 4. 0 International license available under a was not certified by peer review ) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. The authors found that self-induction of a positive valence brain state significantly positively biased valence processing of subsequent stimuli.