Сенсорная нейропластичность,вызванная звуковыми стимулами,ассоциированными с монетарным подкреплением тема диссертации и автореферата по ВАК РФ 19.00.02, кандидат наук Горин Алексей Александрович

• Горин Алексей Александрович
• кандидат наук
• 2021

Оглавление диссертации кандидат наук Горин Алексей Александрович

Table of Contents

INTRODUCTION

Scientific novelty of the study

Theory and methodology

Approbation of the research

Research goals

Provisions for the defense

LITERATURE REVIEW

METHODOLOGY AND STUDY DESIGN

KEY RESULTS

CONCLUSIONS

ACKNOWLEDGMENTS

STRUCTURE OF THE WORK

REFERENCES

ATTACHMENTS

Attachment A. Paper "Correlation of cue-locked FRN and feedback-locked

FRN in the auditory monetary incentive delay task"

Attachment B. Paper "The monetary incentive delay (MID) task induces

changes in sensory processing: ERP evidence "

Attachment C. Paper "Cortical plasticity elicited by acoustically cued monetary losses: an ERP study "

Введение диссертации (часть автореферата) на тему «Сенсорная нейропластичность,вызванная звуковыми стимулами,ассоциированными с монетарным подкреплением»

Introduction

In the field of neuroeconomics, many decision-making models were developed to explain and predict choices between alternatives. Usually, the model includes parameters of the expected value (EV), such as valence, probability, and magnitude of the feedback, also entering a variable to control for the attention toward the alternatives (Collins and Shenhav, 2021). Additionally, there is a wide and consistent evidence of the sensory cortex plasticity inhuman brain. The studies in this field were focused mainly on the language and music learning and showed that repeated exposure to the previously neutral or indistinguishable stimuli may lead to the changes in pre-attentive processing of such stimuli, particularly in the auditory domain. These neuroplastic modulation of the sound processing is shown to be reflected in the appearance or growing of the so-called mismatch negativity, a component of the auditory event-related potentials (ERP) that could be registered during a passive listening to the trail of sounds.

Taking together the aforementioned facts, we hypothesized that repeated exposure to the initially neutral auditory stimuli used as monetary cues may affect the sensory processing of such sounds in the future. Since the MMN reflects pre-attentive processing of the sounds, we assumed that the sensory input to the associative cortices could be modulated by the previous experience, and the current decision-making models should be extended to cover the neuroplastic events in the sensory cortices.

Scientific novelty of the study

The neuroplastic changes of the MMN component were mainly studied during language learning or as a result of music lessons. In our set of studies, we focused on the MMN changes as a result of association between sound stimuli and certain monetary outcomes in both gain and loss domains. The presented evidence of the auditory sensory plasticity as a result of the learning of monetary outcome cued by a sound let us propose that the sensory input to the associative cortices are not stationary. Therefore, the same stimulus may lead to different decision before and after learning even if the EV variables and attention factor were unchanged, meaning that the existing models may be extended by a

stimulus salience variable that could reflect neuroplastic changes in the sensory cortex. Our results support the hypothesis of neuroplasticity of the sensory input to the neural ensembles that process the expected utility of stimuli and extend knowledge about sensitivity of the FRN component to the rewards, losses, and their magnitudes.

Methodological novelty

We developed a set of experimental paradigms that used auditory modality to associate stimuli and outcomes in the MID task that allowed us to manipulate magnitudes and probabilities of losses or rewards and combined this modified MID task with the oddball task to study induced plastic changes in auditory ERPs.

Empirical novelty

For the first time, we demonstrated significant changes in the auditory sensory ERP component as a result of the cue-outcome association during a monetary game and connected them to the parameters of the feedback-related ERP component.

Theory and methodology

To track the neuroplastic changes in the sensory cortex, we compared magnitudes of the MMN component before and after two sessions of a monetary game. Previously, MMN was described as a correlate of such plasticity in non-monetary contexts (Shtyrov et al., 2010; Pantev and Herholz, 2011). To elicit the MMN, we used oddball paradigm, a classic auditory experiment where a trail of standard sounds is interspersed with rare deviant sounds such as tones, vowels or pitches (Naatanen et al., 2007). We measured MMN changes after two sessions of the monetary incentive delay task (MID task, ), that we switched to the auditory domain. In the initial MID experiment, Brian Knutson and colleagues showed that, after a short training, visual cues encoding different EVs of the future outcomes elicited proportional activity in the nucleus accumbens, reinforcing the hypothesis of the effective cue-outcome association. In our paradigm, we combined a passive oddball session with auditory MID task to build a robust stimulus-outcome association, and study the subsequent neuroplastic changes of the MMN component.

In the series of experiments, we developed two versions of the auditory monetary incentive delay task, using them along with the auditory oddball tasks. To collect the data, in all experiments we used BrainVision actiCHamp amplifier (Brain Products GmbH) with a sample rate set at 500 Hz. All subjects were right-handed, with normal or corrected-to-normal vision. They did not report any history of psychiatric or neurological problems, and they all reported to be right-handed. The study was approved by the local ethics committee. All participants gave their written informed consent prior to their participation and received rewards. Approbation of the research

1. Annual Meeting of the Society for Neuroeconomics, October 6 - 8, 2017 (Toronto, Canada). Report: Short-term plastic changes in the primary sensory cortex elicited by monetary outcomes

2. Annual Meeting of the Society for Neuroeconomics, September 26-28, 2014 (Miami, USA). Report: Short-term plasticity in auditory cortical circuit evoked by monetary incentive delay task

3. Annual Meeting of the Society for Neuroeconomics, October 4 - 6, 2019 (Dublin, Ireland). Report: tDCS-induced modulation of the feedback-related negativity in the MID task

This work has been carried out in the Centre for Cognition and Decision Making, Institute for Cognitive Neuroscience, National Research University Higher School of Economics, Moscow, Russian Federation.

Research goals

1. Develop a research paradigm to study brain plasticity associated with financial outcomes in the economic game in the auditory modality using electroencephalography.

2. Study plastic changes in auditory ERP responses to the oddball stimuli associated with the monetary rewards varying in expected value using the newly developed combined MID-oddball paradigm approach.

3. Study plastic changes in auditory ERP responses to the oddball stimuli associated with the monetary losses varying in monetary value using the newly developed combined MID-oddball paradigm approach.

Provisions for the defense

1. We developed an original oddball-MID approach to test the auditory plasticity driven by the financial outcomes associated with the auditory stimuli. The RPE signal, a feedback-locked FRN was modulated by both the magnitude and the probability of outcomes during an auditory version of the MID task. Furthermore, the cue-locked dN200, which is associated with the update of information about the magnitude of prospective outcomes, correlated with the standard feedback-locked dFRN.

2. In the gain domain, we observed learning-related changes in the oddball task (P3a component) as a result of the MID task. MMN changes were not significant but correlated significantly with the FRN amplitude.

3. In the loss domain, we observed learning-related changes in the MMN component as a result of the cue-loss association during the MID task. The amplitude of the changes correlated significantly with the FRN signal.

Literature review

The traditional decision-making theory assumes that individuals' choices are driven by values that are associated with prospective outcomes. Numerous neurobiological studies have implicated the involvement of dopaminergic neurons in the valuation stage of the decision-making process (Schultz, 2006) and in behavioral adaptations (BrombergMartin et al., 2010). Strikingly, while popular neurobiological models of decision making (Rangel et al., 2008; Wang, 2012) acknowledge the key role of learning in reward-based decisions, they indirectly assume that the primary sensory inputs to dopaminergic (decision making) networks are stationary and independent from previous decisions. At the same time, many cognitive studies have demonstrated experience-induced plasticity in the primary sensory cortices (Atienza et al., 2005; Kujala and Naatanen, 2010; Shtyrov et al., 2010; Pantev and Herholz, 2011), indicating that repeated decisions could modulate sensory processing, which in its turn could modulate follow-up decisions. In this light, a separate study that would reconcile evidence of both sensory neuroplasticity and decision-making findings into a uniformed theory seems to be timely.

Sensory cortices retain the capacity for experience-dependent changes, or plasticity, throughout life. These changes constitute the mechanism of perceptual learning (Gilbert et al., 2001). Numerous event-related potential (ERP) studies have shown training-induced neuroplastic changes in auditory information processing that could be explained by the reorganization of neuronal networks and changes in the sensitivity to and processing of relevant information (Atienza et al., 2005; Kujala and Naatanen, 2010; Shtyrov et al., 2010; Pantev and Herholz, 2011). In conditioning paradigms, where auditory tones are used as conditioning stimuli, the training results in associative representational plasticity, which selectively facilitates responses to the conditioned stimuli (Weinberger, 2007). According to the representational plasticity theory, the tuning of the neurons in the primary auditory cortex is selectively shifted towards the characteristics of the conditioned stimulus, thus biasing the whole sensory system to emphasize the behaviorally important stimulus (Diamond and Weinberger, 1986; Bakin and Weinberger, 1990; Edeline and Weinberger, 1993; for a review, see Weinberger,

2015). On the other hand, relatively little is known about the induced plastic changes in the auditory cortex caused by stimuli associated with monetary outcomes.

Main hypothesis

We hypothesized that the learning to discriminate stimuli with different economic values may lead to neuronal reorganization of the auditory cortical regions analogous to plasticity during the learning of foreign language or musical lessons.

Mismatch negativity

The plasticity of auditory processing is often reflected in the mismatch negativity (MMN) component of auditory ERPs. The MMN is a negative ERP component that peaks around 170 ms post-stimulus and reflects the electrophysiological signature of a pre-attentive process that detects alterations in a regular sound sequence (Naatanen, 1990; Winkler et al., 1996). The MMN is evoked by a deviant or rare (i.e., oddball) event embedded in a stream of repeated or familiar events (i.e., standards; Naatanen et al., 2007). The MMN is frequently explained in terms of predictive coding, which is a general theory of perceptual inference (Garrido et al., 2009; Carbajal and Malmierca, 2018). According to this theory, the brain actively learns the regularities of the sensory input and models an internal representation of this information. When the model's prediction of the forthcoming stimulus is violated, the mismatch signal is generated (Paavilainen et al., 1999; Naatanen et al., 2005; Winkler, 2007). Importantly, the amplitude of the MMN is modulated by previous experiences and correlates with behavioral discrimination performance. An initial poor differentiation of the deviant and standard stimuli, as well as inaccurate performance, are correlated with a low-amplitude MMN, while active learning to discriminate deviant stimuli results in larger MMN activity (Sams et al., 1985; Novak et al., 1990; Naatanen et al., 1993; Tiitinen et al., 1994; Cheour et al., 2002). Furthermore, learning-dependent changes of the MMN's amplitude have been demonstrated not only right after discrimination training but also several days later (Kraus et al., 1995; Tremblay et al., 1998; Menning et al., 2000; Atienza et al., 2002, 2005), a training-dependent long-term effect on pre-sensory processing in the auditory cortex.

Since previous studies have robustly demonstrated that training-induced changes of the MMN amplitude are reliable markers of experience-induced neuroplasticity, in our series of studies MMN has been used as a marker of plastic changes in the auditory cortex. To record MMN, we used two versions of the oddball task: a classic one where the standard sound was fixed and differed with deviants equally in intensity and frequency, and a roving standard oddball task (Cowan et al., 1993). During the roving oddball task, eight tones (sounds) formed a roving oddball sequence with 4-6 tone repetitions at each possible carrier frequency, where the first occurrences of a given frequency were considered deviants, while the rest (except for the second and third) were considered standards.

Feedback-related negativity and reward prediction error

The decision theory (Steele and Stefansson 2015) assumes that the values associated with prospective outcomes drive the individual's choices. To build an integrated parameter of expected value (EV), one must know both the magnitude and probability of the reward or punishment (Bandura 1977; Von Neumann and Morgenstern 1944). For example, the temporal difference model of RL (Rescorla and Wagner 1972) indicates that an individual assigns high values to states that predict future rewards when encountered unexpectedly. Reward prediction error (RPE) reflects the mismatch between the expected and obtained outcomes. For example, if a received reward was less than expected (or turns into a punishment, such as a monetary loss), it generates a negative RPE and vice versa. Subsequently, with the seminal work of Wolfram Schultz (1997), RL theory has come to play an important bridging role between economics (e.g., Camerer and Ho 1999; Erev and Roth 1998), psychology (Rescorla and Wagner 1972), and neuroscience (Schultz 1997). The dopaminergic system has been proposed to mediate a signal of prediction error in the form needed in reinforcement algorithms to drive convergence toward a standard dynamic programming value function (Barto and Sutton 1982). The dopaminergic RPE hypothesis has been tested with the use of a variety of neuroimaging techniques, including electroencephalography (EEG) (Duzel et al. 2009; Holroyd and Coles 2002; Knutson et al. 2005; O'Doherty et al. 2001; Pessiglione et al. 2006; Talmi et al. 2012). Using EEG-

evoked responses to the obtained and expected outcomes, Holroyd and Coles (2002) suggested that the feedback-related negativity (FRN) component of event-related potentials (ERPs) can code an RPE learning signal that modifies the performance of the task.

Numerous electroencephalography (EEG) studies have shown that the feedback-related negativity (FRN) component reflects a neural activity that underlies learning and performance monitoring (Holroyd and Coles, 2002; Montague and Berns, 2002; Montague et al., 2004; van Meel et al., 2005; Sambrook and Goslin, 2016). The FRN is a negative deflection with a frontocentral maximum occurring 240-340 ms after receiving negative feedback. According to Holroyd and Coles (2002), the FRN reflects a phasic decrease in dopaminergic activity that disinhibits the anterior cingulate cortex, which signals an RPE (Hajihosseini and Holroyd, 2013). A number of studies have provided evidence for the links between the FRN and mid-frontal theta oscillations with individual behavioral changes (for a review, see Luft, 2014). We recorded the FRN during a monetary game and then studied the sensitivity of FRN to RPE parameters and correlation of the FRN's amplitude with changes in the MMN and P3a, which was recorded using the oddball paradigm before and after the monetary game. Monetary incentive delay task

One of the paradigms that can be utilized to study the association between cue-locked (FRN-like) dN200 and standard FRN is a monetary incentive delay task (MID). The MID task is an elegant tool to study the different stages of RL from reward anticipation to its delivery (Knutson et al. 2000, 2005). It can be used to delineate the neural mechanisms of performance monitoring during behavioral acts with different expected values and RPEs. Initially, the MID task was used in fMRI studies on the neural processing of gains and losses (Knutson et al. 2000). Subsequent EEG and MEG studies utilized the MID task to examine the neural dynamics of reward processing with a temporal resolution in the millisecond range (Broyd et al. 2012; Donamayor et al. 2012; Thomas et al. 2013). The MID task introduces incentive cues that signal both the magnitude and the probability of prospective outcomes. It enables the investigation of the effects of these two

components of expected value on neural activity associated with the processing of incentive cues and feedback (Knutson et al. 2005). In the classic MID task, visual stimuli, such as circles, squares, and triangles, are utilized as incentive cues that code the probabilities and magnitudes of outcomes. To study auditory neuroplasticity caused by learning to differentiate auditory stimuli with different monetary values, we developed a novel auditory modification of the MID task that relied on the sounds of different physical characteristics as incentive cues to construct robust associations between auditory stimuli and monetary outcomes.

To sum up, this research consists of three experiments where we introduced a new approach to studying sensory cortex plasticity as a result of assigning to previously neutral stimuli certain economic values. We studied how different EV encoded by the auditory cues affected FRN responses during the MID task and how the features of FRN correlate with the changes in the MMN responses. To investigate how the MMN magnitude changes after the association of the stimuli, we created an auditory version of the MID task that allowed us to associate stimuli and outcomes. In the auditory MID, we recorded a robust FRN response that is a correlate of RPE processing. Throughout the series of our experiment, we used two versions of our auditory MID paradigm - one with gains and another with losses, where we manipulated the context of the outcome. In addition, we complemented the MID with the oddball paradigm, including its roving-standard modification in the loss-experiment, to record MMN as a robust correlate of plastic changes in the auditory cortex.

Methodology and study design

To collect the EEG data, in all experiments we used BrainVision actiCHamp amplifier (Brain Products GmbH) with a sample rate set at 500 Hz. All subjects were right-handed, with normal or corrected-to-normal vision. They did not report any history of psychiatric or neurological problems, and they all reported to be right-handed. The study was approved by the local ethics committee. All participants gave their written informed consent prior to their participation and received rewards for the participation.

Part I. Twenty-seven subjects (17 women, 23 ± 3 years old) participated in an EEG experiment, in which both behavioral and electrophysiological data were collected. Data from 18 additional subjects were excluded because of the insufficient number of trials in one of the conditions for the averaging procedure or because of excessive EEG artifacts. Acoustic cues signaled a high or low prospective reward probability (0.80 and 0.20, respectively) and a high or low prospective reward magnitude (4 and 20 rubles, the equivalent of 0.07 USD or 0.4 USD, respectively). The participants were given the cumulative reward they had earned, and this was, on average, equal to the cost of a dinner. A set of four sounds (cues) consisted of two levels of frequency (fundamental frequencies of 562 and 487 Hz) and two levels of intensity (55 and 80 dB) to encode the prospective reward probability and magnitude. The probability and magnitude of reward were encoded differently in the two experimental groups. In group 1 (n = 14), the outcome magnitude was encoded by the intensity of the acoustic cue, whereas the gain probability was encoded by the frequency of the acoustic cue. In group 2 (n = 13), the encoding of the gain magnitude and gain probability was reversed. To decrease the effects of the physical parameters of stimuli on ERP processing, we polled the data of two experimental groups.

During the auditory MID task (Figure 1b), the participants were exposed to acoustic cues encoding the prospective gain magnitude (4 or 20 rubles) and probability of a win (0.80 or 0.20). After a variable anticipatory delay period (2000-2500 ms), the participants responded with a single button press immediately after the presentation of a visual target (white square). 800 ms after the button press, the subsequent (2000 ms long) feedback

notified the participants about both current and cumulative outcomes. The 800-ms delay of the feedback aimed to eliminate the possible confound of the visual target duration on feedback-locked ERPs. The overall duration of a single trial was 8 s. The probability of a win was manipulated by changing the average target duration through an adaptive timing algorithm that followed the subjects' performance, such that they would succeed in ~ 80% of the high-probability trials and in ~ 20% of the low-probability trials (Knutson et al. 2005). Thus, in high-probability trials, the participants had more time to give a response than in low-probability trials. The outcomes were positive (a gain of 4 or 20 rubles) or negative (omission of gain—participants did not gain 4 or 20 rubles). At the beginning of the task, the initial duration of the target was based on the reaction times (RTs) collected during the training session. Prior to the MID task, the participants were instructed on which acoustic cues corresponded to which probabilities and magnitudes of outcomes.

In this study, we focused on the auditory cue associated N200 component and feedback-locked FRN component in the MID task.

Part II. In the next stage of our research, we analyzed two sessions of the oddball tasks to study changes in the auditory ERP components.

Forty-two subjects (17 females) participated in an EEG experiment in which both behavioral and electrophysiological data were collected. Five subjects were excluded from the analysis because of excessive EEG artifacts or too few artifact-free. Data of 37 subjects (15 females, 23 ± 3 years old) were included in the final statistical analysis.

The monetary game was analogous to the aforementioned version of MID task.

To probe the learning-related neuroplasticity of the auditory processing, the subjects participated in two identical passive oddball tasks, with the first session of the oddball task performed on Day 1 before the first MID session, while the second session of the oddball task was performed after the second MID session on Day 2 (Figure 1a). The standard stimuli in the oddball paradigm were composed of three sinusoidal partials. Four distinct deviant tones differed from the standard tone in both frequency and intensity

14

such that the probability of an increment or decrement was even. All stimuli lasted 200 ms (including 5 ms rising and falling times). The same four deviant oddball stimuli were also used as acoustic reward-predictive cues for the auditory MID task. The acoustic cues signaled high or low prospective reward probabilities (0.80 and 0.20, correspondingly) and high or low prospective reward magnitudes (4 or 20 RUB, correspondingly. During the oddball tasks, infrequent deviant stimuli were pseudo-randomly interspersed with a standard stimulus presented with a probability of 0.80 and with an 800 ± 100 ms onset asynchrony. Each session started with a training session of four standard stimuli. During passive oddball sessions, the subjects read a book of their own choice.

Part III. The results of the 2nd experiment led us to the new experimental paradigm, where we modified the MID task to switch to the loss domain, fixed the probability of the outcome, and added context of the game. To be able to work with plenty of stimuli, we also used the roving oddball task.

Twenty-nine healthy, right-handed participants (12 males, mean age 23) participated in this study. In order to discover the changes in cortical responses associated with the expected value, we framed losses in three different contexts: low, high, and widely varying losses (LL, HL, WL), where one could lose 1 or 2, 50 or 51, and 1 or 50 MU, respectively. The task was split into six blocks: two blocks per context. In each block, only two stimuli and two corresponding monetary outcomes were used (see Supplementary Fig. S1). Prior to the MID session, participants received an endowment of 4000 MU (~ 70 USD). They were instructed that they might lose a part of the initial endowment during the game and that they could complement their compensation for participating in the experiment with any remaining balance. Importantly, six auditory stimuli (325, 381, 440, 502, 568, and 637 Hz) composed three pairs of incentive cues that predicted small and large losses: - 1 or - 2 MU ('low losses' context, LL-trials), - 50 or - 51 MU ('high losses' context, HL-trials), - 1 or - 50 MU ('widely varying losses' context, WL-trials). Therefore, in the LL- and HL-trials, the difference between the outcomes was equal to 1 MU, which was irrelevant in the context of the (4000 MU) initial endowment, whereas in the WL-trials, the difference between the outcomes was equal to

49 MU; therefore, the participants were more motivated to discriminate the cues, i.e., perceptual learning was only relevant for maximizing monetary outcomes in the WL-trials. This design allowed us to separate the effect of context (behavioral relevance) on participants' behavior from the effect of outcome size. Each pair of auditory monetary cues was randomly presented within mini-blocks of 50 trials: LL-trial blocks, HL-trial blocks or WL-trial blocks.

In this experiment, we could not use the classic oddball task since MMN amplitude is proportional to the difference of frequency of the standard and deviant stimuli. In our case, we had 6 experimental stimuli plus 2 control sounds, when in the classic oddball we could use only 2 stimuli that could equally differ from the standard. To solve this problem, we adapted a roving oddball ask that allows using more different pitches (Cowan et al., 1993). During the roving oddball task, eight tones (sounds) formed a roving oddball sequence with 4-6 tone repetitions at each possible carrier frequency, where the first occurrences of a given frequency were considered deviants, while the rest (except for the second and third) were considered standards (Figure 1c). Each sound type was presented randomly 50 times under the constraint that each consecutive sound was required to be from 1 to 3 steps higher or lower than the previous. It should be noted that the lowest and highest sounds were not used in the MID task and were analyzed separately as control condition stimuli that had no association with monetary outcome.

In this experiment, we focused on the changes in the MMN component induced by the cue-loss association in different game contexts during the MID task.

Taken together, we ran three experiments with 116 subjects in total to investigate auditory ERP components and changes in their amplitude during active and passive sound perception in the context of the monetary game in loss and gain domains together with feedback associated ERP component.

Key results

Part I (EEG Study I). We showed that feedback-locked FRN was modulated by both the magnitude and the probability of outcomes during an auditory version of the MID task. Furthermore, the cue-locked dN200, which is associated with the update of information about the magnitude of prospective outcomes, correlated with the standard feedback-locked dFRN, which is associated with the processing of favorable and unfavorable outcomes (RPE). The results further expand our knowledge of the interplay between the processing of the evaluation of ongoing predictive events and future outcomes and the following revision of these predictions during outcome delivery.

Part II (EEG Study II). We tested whether repeated exposure to the stimuli that signal different incentive values in the MID task changes their sensory processing when tackled in the oddball tasks. In the absence of the group MMN effect, we observed learning-related changes of the P3a, indicating a stronger reallocation of attention to the incentive cues. The correlational analysis of individual MMN amplitudes with the MID-session FRN responses revealed that a stronger RL signal was associated with more finegrained discrimination of the incentive cues. Our results showed that plastic changes associated with better discrimination could be sensitive to the continuing valuation of incentive cues that leads to enhanced involuntary attention switching.

Part III (EEG Study III). We studied auditory cortical neuroplasticity signaled by the MMN magnitude changes in the auditory oddball experiment. Prior to the oddball experiment, a cue-loss association was established during the monetary incentive delay task (MID) where participants were losing initial monetary endowment trial by trial. The MID task had three contexts: in two of them, different loss magnitudes were of the same order - either big or small, whereas, in the third, the losses significantly varied thus creating a widely varying losses context. As a result, we observed significant growth of the MMN for the acoustic cue that predicted larger loss in the widely-varying loss context, where sufficient learning was necessary for successful performance. The magnitude of differences, dMMN, significantly correlated with the individual sensitivity of the FRN to the loss amount (dFRN). We also analyzed the source distribution of the signal using a

standard head model and found that the major part of the activity was located in the temporal cortex, supporting our hypothesis of the plasticity in the auditory sensory cortex.

Taken together, this series of experiments focused on the plastic changes and dynamics of auditory ERP components and alterations in the FRN component that is connected to the reward prediction error processing. We studied the modulation of the FRN by the magnitude of expected outcome in gain and losses domains, by the probability - in the gain domain. We also studied how auditory ERPs change as a result of cue-outcome association and found that in the loss domain change in the MMN component was most prominent only if the sound cue predicted big loss in the context of much smaller. We localized the source of the differential activity in the temporal cortices. Taken together, our results support our initial hypothesis of alterability of the sensory inputs to the reward processing associative cortices and extend our knowledge in the field of reward processing dynamics.

Conclusions

• In the auditory version of the MID task, we observed FRN component as a response to the monetary feedback;

• FRN component has been modulated by probability and magnitude of the outcome;

• The cue-locked dN200 signal showed significant correlation with the FRN signal;

• As expected, MMN and P3a components were observed during the oddball task;

• In the gain domain, we observed insignificant modulation of the MMN signal, that has been correlated with individual FRN magnitudes;

• In the gain domain, the P3a component changed its amplitude as a result of the stimulus-outcome association during the MID task;

• In the loss domain, amplitude of the MMN component changed after 2 sessions of the MID task. The changes were significant in the context, where learning was most significant for the performance;

• The changes of the MMN in the loss domain were correlated with the amplitudes of the individual FRN component.

Initially, we hypothesized that the sensory input to associative cortices could be modulated by the association of a stimulus with a particular outcome. To check the hypothesis, we modified the MID task and switched it to the auditory modality, we encoded the probability and magnitude of the outcome by frequency and intensity of the cues. EEG results demonstrated that the FRN signal has been modulated by features of the outcome. Therefore, we added two passive oddball sessions - prior to and after MID sessions and focused on the auditory ERPs observed in the oddball. The results indicated that the P3a component changed its amplitude as a result of the MID task, but MMN changed insignificantly. Nevertheless, individual changes in MMN correlated with the magnitude of FRN in the MID task. To simplify the MID task, we fixed the probability (and intensity of the sounds). We further reasoned that since people are more sensitive to

losses, it would be more efficient to study the modification of the sensory input in decision-making in the loss domain. As a result, we demonstrated that MMN amplitude increased for the cue that encoded bigger monetary loss, that this growth has been correlated with individual parameters of FRN, and that sources of this differential signal were spread over the temporal cortices. Our results supported our initial hypothesis and may inspire further studies of the plasticity of the sensory cortex as a result of economic games.

Список литературы диссертационного исследования кандидат наук Горин Алексей Александрович, 2021 год

References

1. Naatanen, IL. Paavilainen. P.. Rlnne. T. & Alho. k. 'Ihe mismatch negativity (MM N) in basic research at central auditory processing: A renew. Clin. Heuraphynol. 11«, 2344-2590 (2007).

2. Kupala. T. & Naatanen. R. The adaptwe brain: A neurophvsiological perspective, frog- NeurobioL 91. 55-67 (2010).

3. Ka as, III. & Collins. C E. The organization oi sensory cortex. Clirr. Opoi. Seuwbiat 11. 498-504 (2001).

4. Rossini. P. M. et dL Short-term brain 'plasticity' in humans: transient linger representation changes m sensory cortex socnafioecipv following ischemic anesthesia. Bruin Res. 642.169-177(1994).

5. W eiss. T., Miltner. W. 11. R. Lie-pert, L. Meis&ner, W. St Tauh E. Rapid functional plasticity in the primary somatomotor cortex and perceptual changes alter nerve block. Eur. I Seurosa. 20.3413*1423 (2004).

6. Bakin. 1 S. & Weinberger. N. M. Classical conditioning induces CS.speclhc receptive field plasticity in the auditory cortex of the guinea p«g. Brum Rex 51*. 271-186 (1990).

7. Reed, A. et aL Cortical map plastiaty improves learning but it not necessary for improved performance. Neuron 70. 121-131 (2011).

Scientific Reports | (:o:o)io::u6x | https://doi.orq/10 1038M1598-020-78211-7 nature research

H. Elaas. G. A., Bieszczad, K. M. 8r Weinberger. N. M. Learning strategy refmemenl reverses early sensory cortKal map expansion but not behavior: Support ior a theory of directed carlical substrates of learning and memory. Sevrobiol. Learn. Mem. 126. 39-55 (2015).