Poster Sessions

Gilles-de-la-Tourette syndrome (GTS) is a congenital neuropsychiatric disorder characterised by rapid, uncontrolled repetitive movements (motor tics) and/or vocalizations (vocal tics), occurring suddenly without any recognizable purpose with commonly reported premonitory urges. Previous studies indicate the hypothesis of dysregulation of the dopaminergic system, in particular a dysfunction of the tonic / phasic dopamine release or a dopaminergic hyperinnervation. In view of the complex interaction of different neurotransmitters, especially in thebasal ganglia, it can also be assumed that an abnormal dopaminergic transmission also affects other transmitter systems, e.g. glutamate (Glu), glutamine (Gln) or gamma-Aminobutyric acid (GABA). Furthermore, recent studies indicate abnormalities in cerebral iron metabolism in GTS. Since iron is accumulated in dopamine vesicles and plays a central role in dopamine synthesis, this observation may also be related to the disorder of the dopaminergic system. In this study, we combine methods of positron emission tomography (PET), magnetic resonance tomography (MRI) and magnetic resonance spectroscopy (MRS) in order to compare the following parameters in patients with GTS and a control cohort: (i)the binding potential of D1-dopamine receptors -for first time in-vivo for GTS, (ii) the concentrations of Glu, Glutamine and GABA in the corpus striatum andin the cortex cingularis anterior (iii) the subcortical iron concentration.

Abstract MRI holds great promise for diagnosing Parkinson's, based on the disappearance of the swallowtail sign in T2*-weighted images, which past research assumed to be nigrosome 1. We studied the sign's anatomical underpinning, combining ultra-high field MRI in vivo and postmortem, 3D-reconstructed microscopy, and immunohistochemistry. Based on block-face images and calbindin-D28K immunohistochemistry, we constructed a 3D nigrosome atlas. We show that nigrosome 1 extends beyond the swallow tail sign by co-registering this atlas to in vivo MRI. As the swallow tail sign only partially overlaps with but is not identical to nigrosome 1, its interpretation needs to be revised.

The goal of language comprehension is to interpret the communicator’s intended message. To support this process, comprehenders can make use of the communicative context in order to generate predictions about upcoming input. If bottom-up input matches these top-down predictions, comprehenders are given a ‘head-start’ and processing is facilitated. However, if comprehenders encounter input that violates their predictions, additional processing demands are needed to fluidly adapt to the unexpected input. The planned project will investigate whether the brain engages distinct neural mechanisms in response to output that fulfills versus violates strong predictions based on three-sentence discourse scenarios. German native speakers will read and/or listen to such scenarios, in which the third sentence constraints for a broad event structure, e.g. {Agent cautioned animate-Patient}. In high constraint contexts, a two-sentence context about for instance a beach, lifeguards and sharks, additionally constrains for a specific target word {Agent cautioned Swimmers}. Low constraint contexts are vague and do not constrain for any specific word. We will measure ERPs on target words in the third sentence of the scenarios that fulfill and/or violate each of these constraints. We expect to find clear, dissociable effects to fulfilled semantic predictions (a reduced N400), to lexical prediction violations (an increased late frontal positivity), and to event structure animacy violations (an increased late posterior positivity/P600). We argue that these consequences of strong predictions reflect earlier-made expectations at the verb preceding the target word. At this verb position, we expect a reversed N400. Furthermore, we expect that these ERP responses are similar for visual and auditory presentation. The planned study will provide evidence whether confirmed and violated prediction manifest as distinct spatiotemporal neural signatures. Further, it will offer functional explanations of the observed pattern of ERP findings that will be mapped onto specific interactions within the cortical language circuit. The resulting theoretical framework will form the basis for our future transcranial magnetic stimulation/electroencephalography studies that probe the functional relevance of brain regions, which consequently will contribute to a deeper understanding of the neurobiology of language comprehension.

Sign language offers a unique perspective on the human faculty of language by illustrating that linguistic abilities are not bound to speech and writing. In studies of spoken and written language processing, lexical variables such as, for example, age of acquisition have been found to play an important role, but such information is not as yet available for German Sign Language (Deutsche Gebärdensprache, DGS). Here, we present a set of norms for frequency, age of acquisition, and iconicity for more than 300 lexical DGS signs, derived from subjective ratings by 32 deaf signers. We also provide additional norms for iconicity and transparency for the same set of signs derived from ratings by 30 hearing non-signers. In addition to empirical norming data, the dataset includes machine-readable information about a sign’s correspondence in German and English, as well as annotations of lexico-semantic and phonological properties: one-handed vs. two-handed, place of articulation, most likely lexical class, animacy, verb type, (potential) homonymy, and potential dialectal variation. Finally, we include information about sign onset and offset for all stimulus clips from automated motion-tracking data. All norms, stimulus clips, data, as well as code used for analysis are made available through the Open Science Framework in the hope that they may prove to be useful to other researchers: https://doi.org/10.17605/OSF.IO/MZ8J4

We are born with the natural ability to learn language. Infants first tune into acoustic properties (Nazzi & Ramus, 2003) and then employ acoustic information to infer abstract linguistic knowledge, such as word meaning and combinatorics (Gleitman & Wanner, 1982). It is notoriously hard to measure the emergence of linguistic knowledge, because of children’s limited attention and response abilities. We here combine measures of cortical tracking of speech acoustics and computational modeling to assess the development of speech tracking abilities and linguistic knowledge. To assess linguistic knowledge before infants can actually speak, we measure electrophysiological responses during the presentation of naturalistic children’s stories. We are collecting a quasi-longitudinal data-set covering the period between 2 months and 5 years of age. The developmental trajectories of acoustic tracking and linguistic knowledge will be analyzed using growth-curve models. To assess speech tracking, we measure the synchronization between the speech envelope and the electroencephalogram. We predict synchronization will increase during the first year of life, when speech processing becomes more native-language specific (Kuhl, 2004). To assess the inference of abstract linguistic knowledge, we will employ surprisal, a metric taken from information theory (Hale, 2001). Surprisal describes the amount of lexical–semantic or syntactic information conveyed by a word. We predict that neural sensitivity to surprisal will emerge gradually between the second and third year of life, depending on the degree of linguistic abstraction (see Morgan et al., 2020)—that is, electrophysiology should first mirror lexical–semantic, then syntactic knowledge. Importantly, given that the acquisition of abstract information presupposes acoustic tracking abilities, we hypothesize an interaction between acoustic synchronization and sensitivity to abstract linguistic information. In sum, this study has the potential to assess the emergence of linguistic representations before children can solve tasks and opens new ways to look at the interaction between acoustic speech processing and comprehension of meaning in early childhood.

Despite accumulating evidence on the clinical usefulness of neurofeedback (NF) in the treatment of various disorders, including overweight and obesity, the neurophysiological and clinical mechanisms underlying NF efficacy are unclear. Specifically, previous studies demonstrated that the real-time ('online') signal differs from the more precisely processed offline brain activity, that alterations in brain activity generalized from targeted regions to a ‘NF control network’ in the prefrontal cortex (PFC), and that brain-based, self-reported, and clinical NF success varied considerably between participants and sessions and depended on participants’ mental strategy use. In order to study the potential of NF in the treatment of eating disorders and to shed light on underlying NF mechanisms, our laboratory carried out the feasibility study “Near Infrared Spectroscopy Neurofeedback for Binge-Eating Disorder” (NIRSBED; DRKS00014752), the most prevalent eating disorder, characterized by recurrent binge-eating episodes in the absence of regular weight control behaviours. Adults with binge-eating disorder (n = 40) underwent 12 sessions of functional near-infrared spectroscopy NF, with feedback from individualized regions of interest in the PFC displayed via individually appetitive food pictures. Participants were instructed to use any mental strategy they could imagine, but to only continue the use of successful strategies to up-regulate the feedback signal. Participants performed up-regulation trials with and without immediate feedback and a passive viewing condition. Brain-based regulation success was defined by the difference in PFC response between up-regulation versus passive viewing during immediate and delayed feedback. First, the online and the offline regulation success will be related to participants’ self-reported regulation success after each session, assuming a stronger correlation between the online versus offline regulation and self-reported success. Second, based on self-reported food craving, urge to eat, and loss of control over eating before and after each session, we will investigate the clinical mechanisms underlying NF efficacy, hypothesizing significant decreases after versus before sessions and with increasing number of sessions. Third, the study aims to exploratively evaluate the effects of participants’ sessions-wise mental strategies on brain-based and self-reported regulation success.

INTRODUCTION. Neurophysiological studies investigated the automaticity of syntactic processing under conscious conditions. However, it remains largely unknown whether syntactic analysis proceeds outside of conscious awareness. In this study, we employed a subliminal syntactic priming paradigm to address the automaticity of syntax in minimal phrases at the behavioural level. METHODS. We run four behavioural experiments to test whether recognition of a target’s category (noun or verb) could be influenced by its syntactic relationship with a prime (“a” or “he”) in the German language. Prime-target relationship was either congruent (“a” + noun, “he” + verb) or incongruent (“a” + verb, “he” + noun). In the pilot and Experiments 1 and 2, verbs included an overt morphological cue (the suffix “-t”), which was absent in Experiment 3. In Experiment 2 we included nonword primes as a baseline condition, to remove processing differences between nouns and verbs independent of syntactic context. Unconscious processing of the primes was achieved via a short prime duration (33ms) and a special masking technique. Prime awareness was assessed in a so-called ‘prime visibility task’. RESULTS. Experiments 1 and 2 revealed a subliminal syntactic priming effect: words preceded by syntactically congruent masked primes were processed faster than words preceded by incongruent primes. Furthermore, Experiment 2 revealed an inhibitory nature of syntactic priming: incongruent primes slowed down the processing of nouns and verbs. When no morphological cue was present (Experiment 3), we found no subliminal syntactic priming. DISCUSSION. Our study shows that abstract syntactic representations might be accessed unconsciously in an automatic fashion. Furthermore, the results indicate a role of morphological cues in the early and automatic steps of syntactic analysis. Though, further research is required to confirm our speculations, more specifically, to provide more evidence for the reliance of subliminal syntactic priming effects on morpho-syntactic markers.

The question to what extant preverbal infants are agents has been one of the central debates in developmental research and theory. Once infants are able to act, their interaction with the environment becomes more purposeful and rewarding. We review existing studies that have investigated agency in the first years of life. In the classical Mobile Paradigm, for example, infants increase the movement of a limb, if this limb is attached to a mobile. These paradigms indicate that infants detect multi-sensory contingency and have been interpreted as evidence for the development of a sense of agency. Here, we argue that neither of these measures allows conclusions about the presence of agency because they fail to indicate the infant’s control over the movement. Multi-sensory contingency learning, we argue, does not go beyond reinforcement learning. Consequently, it does not tell us whether infants perceive themselves as agents or have any control over their movements. We suggest that, instead, what will inform us about agency, are infants’ reactions to a break-down of contingency, and the extent to which these reactions show infants’ control over and flexible adaption of their actions. In sum, we take an interdisciplinary perspective to show how control, as an important criterion of agency in philosophy, has been overlooked in the empirical research on agency. We combine previous empirical work from different fields to derive an experimental suggestion for a test of agency in a non-verbal population. The experimental design of this experiment will be described on the poster.

Human listeners are capable of implicitly learning and memorising sounds just through repeated exposure. Remarkably, this capability for perceptual learning is not restricted to sounds that carry a semantic meaning. It is also observed for previously unknown random and abstract acoustic patterns such as tone-pip sequences or white noise. However, several determinants of perceptual learning remain unexplored. Most previous studies relied only on indirect measures of learning. Patterns were usually repeated seamlessly within a stimulus and not presented in isolation as it happens under more naturalistic conditions such as language learning. Moreover, little is known about whether perceptual learning of previously unknown acoustic patterns is shaped by contextual factors such as a given task or the presentation format. The current behavioural study (n = 50 participants to be tested in an online experiment by May 2021) sought to explore direct and indirect correlates of perceptual learning of random acoustic patterns in different contexts. During an implicit learning phase, two groups of listeners performed different perceptual tasks. Participants were presented with acoustic patterns that were randomly computer-generated, but shared statistical properties of natural sounds. One group performed a well-established repetition detection task that asked them to detect seamless pattern repetitions in a continuous sound. The other group performed a discrimination task that asked them to indicate whether two patterns presented as separate sounds with a silent interval in between were same or different. Unbeknownst to the listeners, in each experimental block one specific pattern was repeated across trials. In line with earlier findings, as an indirect measure of learning we expect an increase in task performance for the repeatedly presented pattern in comparison to patterns that were presented only once. Additionally, we directly tested recognition of the learnt patterns in an unexpected testing phase. Thereby, we aim to advance the understanding of auditory perceptual learning with regard to three aspects: We aim to (i) assess behavioural recognition as a more direct correlate of pattern learning, (ii) show that pattern learning also occurs in a more naturalistic presentation format in the absence of within-stimulus repetitions, and (iii) test whether learning and recognition are shaped by contextual factors (i.e. the participants’ task and the presentation format).

From early on, infants seem to track what others see (visual perspective taking), and recent research indicates that this may also affect their own representation of the environment. This influence of others on one’s own representation has been observed in adults and has been referred to as altercentric modulation. Here, we use the steady-state visual evoked potential (ssVEP) as a neural marker of object representation and probe to what extent the presence and visual perspective of an agent modulates infants’ and adults’ object processing. To this end, we presented 10-months-old infants and adults with videos in which flickering objects move across a table (visual on-off stimulation at 4 and 6Hz). We manipulated whether 1) the scene was experienced together with or without an agent and 2) the agent’s visual access to the object. More specifically, at the end of each trial, the object either disappeared into a tunnel (no visual access) or behind an occluder (continued visual access for the agent). We hypothesize a sustained ssVEP response in the occluder compared to the tunnel condition as the agent still has visual access to the object. Preliminary data (of N=15 infants, N=30 adults) confirms that infants and adults show enhanced neural activity at the entrained frequency while the object is presented. This activity declines continuously with the disappearance of the flickering object. We will contrast the ssVEP time course between frequencies, and report power differences between conditions.

The understanding of post-stroke language recovery is crucial for clinical interventions. A previous neuroimaging study demonstrated distinct mechanisms during different phases of language reorganisation after stroke: global network disturbance (i.e. diaschisis) during the acute phase, resolution of diaschisis, upregulation of bilateral domain-general networks and restitution of perilesional cortex during the subacute phase, and reintegration of left temporal language areas during the chronic phase. They also identified how these phase-specific mechanisms are dependent on lesion sites. However, it remains unclear how brain regions within these mechanisms interact during different phases after stroke. This doctoral project aims to investigate the effective connectivity in the language networks after temporo-parietal and frontal strokes. We will firstly explore the changes in time-specific and lesion-specific effective connectivity from the acute and subacute to the chronic stage post stroke. Further, we will compare the effective connectivity results between stroke patients and healthy controls. Finally, we would like to establish the relationship between neural coupling and language behaviour. This project will provide a better understanding of the neural mechanisms underpinning post-stroke language reorganisation. Moreover, these findings will bring insights to the clinical interventions of post-stroke language rehabilitation, such as new targets for non-invasive brain stimulation based on individual lesion locations. This project therefore clearly targets current needs in the field of neuroplasticity research as well as in the clinical treatment for neurological diseases.

Recent fMRI studies have shown that the human spinal cord exhibits spatially distinct patterns of spontaneous activity at rest, such as enhanced functional connectivity (FC) between the dorsal horns and between the ventral horns [1, 2], mirroring the anatomical divide into sensory and motor function. The link between such spinal activity patterns and behavior is still unclear, but there are initial reports of altered resting-state spinal cord FC due tospinal cord pathology in humans [3] and non-human primates [4]. While this has led to suggestions that resting-state spinal cord FC might be used as a diagnostic and prognostic biomarker for sensory and motor disorders (such as chronic pain and multiple sclerosis), a first essential step is to quantify its reliability. Test-retest reliability of spinal cord resting-state FC has been investigated at 7T [5], an approach that we here aim to extend to the clinically prevalent field strength of 3T. We acquired fMRI data from 24 healthy participants using a gradient-echo EPI sequence with slice specific z-shimming [6] on a 3T MRI system (Siemens Prisma) equipped with a 64-channel RF head-neck and a spine-array coil. 24 5-mm thick transversal slices with an in-plane resolution of 1x1mm were acquired, covering a field of view from the 2nd cervical vertebra to the 1st thoracic vertebra (TR: 2312ms, TE: 40ms, FA: 84°, FOV: 128×128mm, GRAPPA-factor: 2). Two functional runs with 250 volumes each were acquired with a maximal temporal gap between runs of ~10 minutes. Preprocessing and data analysis were performed using MATLAB, FSL and the Spinal Cord Toolbox [7]. First of all, similar to previous studies, we observed significant positive connectivity between the dorsal horns(r=0.10, p<.001 and between the ventral horns p dorsal-ventral connectivity within hemicords was not significant. in terms of reliability these patterns icc dorsal-dorsal whereas for ventral-ventral thus exhibiting fair reliability. within-hemicord between-hemicord those that essentially did show any effects were poor range with iccs respectively.>

Processing time windows of 2–3 seconds constrain the sampling of language and consequently influence comprehension. While prior work converged on a window size of about 2.7 seconds on the group level, this window size likely varies between individuals. We here aimed at uncovering inter-individual differences in the duration of these linguistic processing windows. We tested 40 participants on ambiguous sentences (e.g., Mrs Groß called the nurse of the pensioner who always joked.), where the relative clause "who always joked" can attach either to the nurse (= NP1; the nurse who always joked) or to the pensioner (= NP2; the pensioner who always joked). Attachment was monitored through two-alternative forced-choice comprehension questions (e.g., Who joked always?). We aimed at revealing individual temporal constraints by manipulating the duration of the sentence up until the end of "pensioner". We created seven levels of auditory stimuli, where duration of this lead-in part ranged from ~2–4 seconds. We hypothesized that attachment would depend on duration: While short lead-in parts should lead to termination of a processing window after NP2 and trigger attachment to NP1, long lead-in parts should elicit segmentation after NP1 and thus NP2 attachment. The duration at which attachment switches from NP1 to NP2 would indicate individuals’ time constraint. Our results confirm a change in attachment as a function of duration. Yet, strikingly, attachment did not change linearly. Instead, fluctuations of attachment could be best fitted by a sinusoidal function. This suggests that attachment might reflect underlying oscillatory activity; specifically, participants may always attach to the NP that arrives at a specific phase angle of the underlying oscillation. This is in line with recent evidence that found processing to be facilitated when information arrives at an excitable phase of an ongoing neural oscillation. Importantly, we suggest that the frequency of the individual sinusoidal fits may directly reflect the alternation of excitable states of the underlying oscillator and is thereby the source of inter-individual variations in comprehension. Further research using MEG aims at confirming this hypothesis.

The auditory system is organized tonotopically. As a consequence, the position of a sound source has to be inferred based on a number of implicit sound localization cues. The most important ones are interaural time and intensity differences, which correspond to a sounds azimuth, and direction dependent spectral filtering trough the head and ears, from which the sounds elevation is inferred. Because of it's fragmented nature we can manipulate single aspects while leaving the rest of auditory spatial perception intact. This makes sound localization an interesting model to study the relationship of perception and neural correlates. With over a hundred years worth of psychoacoustical research, it is well understood how sound localization cues shape our spatial perception sound. In addition, electrophysiological studies found location sensitive neurons at various stages in the auditory pathway in different mammals and birds. Less known, however, is how sound source position is encoded on a population level. We investigate the encoding of sound source elevation by playing noise bursts from loudspeakers at different positions while recording the listeners EEG. We use stimulus-specific adaptation to isolate the elevation dependent component of the brain response. Since only a small minority of auditory neurons responds selectively to sound source elevation we use a signal separation technique called joint decorrelation which allows us to filter the EEG signal for changes across elevations. By comparing the changes in the EEG signal with the listeners localization performance we can demonstrate that the signal does encode sound source elevation. Using permutation cluster analysis, we can find the time intervals in which the activity differs significantly across elevation and model these differences to reach a better understanding of how sound source elevation is encoded in the brain.

Stress and stress-associated diseases are the health epidemic of the 21st century, currently exacerbated by the Covid-19 pandemic. Interestingly, despite experiencing roughly similar amounts of stress than those falling ill, some individuals seem to be protected against the “wear and tear of daily life”. In the current study, we aim to identify abilities that protect individuals from mounting a strong emotional and physiological response to acute psychosocial stress exposure. In detail, we focus on facets of mindfulness and compassion. Both abilities are frequently trained in increasingly popular meditation-based interventions such as, for example, the Mindfulness-Based Stress Reduction Program (Kabat-Zinn, 1990) and Compassion Focused Therapy (Gilbert, 2009). To assess different aspects of mindfulness and compassion, 130 men and women completed the Five Facet Mindfulness Questionnaire (FFMQ), the Self-Compassion Scale (SCS) and the Compassion Scale (COSN). To induce acute psychosocial stress, the Trier Social Stress Test, a standardized laboratory paradigm (Kirschbaum et al., 1993) was conducted, and levels of the hypothalamic-pituitary-adrenal axis end hormone cortisol were assessed over the course of two hours. We found that the FFMQ scale Describe and the COSN scale Common Humanity were inversely associated with stress-induced cortisol release. Higher scores on the scales SCS Self-Kindness and COSN Kindness, on the other hand, were associated with higher stress reactivity. These results give valuable insights into the components of mindfulness and compassion that may be central to stress reduction in long-term training intervention.

Asymmetry is a key organisational feature of the human brain and related to the hemispheric preference of functional processes, for example language. However, to what extent the functional organisation of the cerebral cortex is asymmetric and whether this asymmetry is heritable are not known to date. The present study aims to investigate asymmetry of functional organization in healthy humans, study its heritability using a twin-sample, and analyze the asymmetry in macaque monkeys to map the phylogenetic conservation. To capture the functional organization, we employed the gradient approach with diffusion map embedding, which describes how the neuronal populations are integrated and segregated in the functional connectome. Our findings revealed that large-scale functional gradients were asymmetric and the asymmetry was heritable (particularly the language network), indicating that individual variation in asymmetry of functional organization is under genetic control. Notably, macaques show no significant asymmetry of functional gradients, potentially indicating that asymmetry of functional organization is, phylogenetically, a relatively novel adaptation in functional brain organization. Functional decoding underscored the functional relevance of the observed patterns, with regions showing left-ward asymmetry associated with language processing, whereas right-ward asymmetries in organization related to cognitive control. Overall, our findings suggest a genetic and evolutionary basis for a hemispheric difference in intrinsic functional organization.

Structural predictions—e.g. expecting a noun or an adjective after processing a determiner like “a”—have been proposed as the mechanism supporting the fast route of syntactic analysis (Lau et al., 2006). Correlative studies implicate Broca’s area’s involvement in categorical prediction (Bonhage et al., 2015), but causal evidence for this claim is still missing. In this study, we tested this claim by simultaneously combining online Transcranial Magnetic Stimulation (TMS) and Electroencephalography (EEG) in a two-word auditory paradigm. We employed German grammatical (i.e., a determiner followed by a noun, or a pronoun followed by a verb) and ungrammatical (i.e., a determiner followed by a verb, or a pronoun followed by a noun) structures to elicit the Early Syntactic Negativity (ESN, Hasting & Kotz, 2008). We reasoned that, if Broca’s area is causally involved in categorical prediction, impeding its functioning with short bursts of TMS at the predictive stage (i.e., the first word) would have decreased the ESN effect (ungrammatical – grammatical) compared to two control conditions (sham and superior parietal lobe stimulation). Cluster-based permutation tests were conducted (N=29). Furthermore, by employing state-of-the-art modelling (Weise et al.,2020) we quantified and correlated the TMS-induced electrical field in our target region with changes in the ESN relative to the sham condition. A main effect of grammaticality was observed (ESN and a late positivity), replicating previous findings (Hasting & Kotz, 2008; Jakuszeit et al., 2013). Contrary to our hypothesis, TMS over Broca’s area did not affect the ESN amplitude. Similarly, the correlation between changes in the ESN effect and the TMS-induced electrical field in Broca's area was non-significant. Our findings suggest that Broca's area may not be causally involved in categorical prediction. They are compatible with a different role of this region in syntactic composition, namely the bottom-up integration of words into syntactic structures (Bhattasali et al., 2019). Accordingly, in our study TMS did not affect the ESN component because stimulation of Broca’s area occurred before this region could build a constituent in a bottom-up fashion, as no syntactic rule could be evaluated on a single function word. This study has important implications on the understanding of language processing and provides a testable computational hypothesis for future studies.

Despite being more complex compared to other types of languages (Lupyan & Dale, 2010, Trudgill, 2011), languages with noun classification and agreement systems are widespread across the world: they are present in languages genetically and geographically non-related, like Cantonese, German and Sesotho. These systems seem to provide a communicative advantage in terms of language processing, as shown by studies on Cantonese and German (Tsang, Chambers & Mozuraitis, 2017; Dye et al., 2016) which suggest that agreement facilitates word recognition. On the other hand, noun classification and agreement systems appear to be more costly in terms of language acquisition: they require learning several morphological encodings both for the noun class and the agreement. However, recent research has shown that this morphological redundancy might facilitate noun class acquisition. For example, Demuth and Weschler (2012) investigate noun class acquisition in Sesotho. Sesotho has several noun classes with their corresponding agreement encodings. However, in certain grammatical conditions, Sesotho allows null prefixes. The results of this study are that children make mistakes in noun classification and agreement only with nouns which can take the null prefix. This suggests that agreement strengthens the robustness of children’s lexical representations. However, this study does not provide a direct test of the role of morphological agreement in noun class learning. My dissertation investigates how agreement affects noun class acquisition through an artificial language learning paradigm. There are three experimental conditions: reduplicative, arbitrary and no agreement, which vary in terms of the type and number of morphological encodings. In the reduplicative agreement condition, the two suffixes are identical, whereas in the arbitrary agreement condition, the two suffixes are different. Therefore, the second condition has morphological redundancy which facilitates learning, but poses an extra challenge, as it requires to learn two encodings. Participants will be divided into three groups and assigned to one condition. The main prediction is that the no agreement condition will yield worse learning results compared to the other two, as it lacks multiple cues to noun class. A second prediction the reduplicative agreement condition will yield better results than the arbitrary one.

The ReSource project is a longitudinal mental training study. Over nine months, participants underwent training of distinct meditation modules, using a wide range of techniques designed to enhance either attention and interoceptive awareness (Presence Module), socio-affective capacities such as compassion, loving-kindness and gratitude, (Affect Module), or socio-cognitive capacities such as perspective taking on self and other. In the current study, we explored how individual improvement in the specific abilities trained per module (measured using paper-pencil questionnaires) moderated training-induced stress reduction (measured in terms of hormonal, sypmathetic and emotional activation in response to a psychosocial laboratory stress task).

Image reconstruction is the first post-processing step in magnetic resonance (MR) imaging and determines the base quality for further processing. Modern techniques driven by sparsity constraints lead to better image quality and more options to accelerate the acquisition compared to methods strictly based on analytical models of the imaging process. Deep Decoder (DD) [1], a promising approach of this kind, tunes a convolutional neural network (CNN) as an adaptive function transforming a fixed noise vector to the image of interest. The CNN represents an implicit prior on the structure of clean images leading the network to create essentially a denoised or artifact reduced version of the target image. DD has been successfully applied to MR data with Poisson-sampling [2]. Looking for improved image reconstruction of our own multi-parametric mapping (MPM) MR data [3] acquired in sub-sampled cartesian grid pattern, we want to evaluate the applicability of DD. Following [2] the original DD was modified to process axial slices of MPM data of a human head. The subsampled k-space was zero filled, Fourier transformed, sliced in readout direction and zero padded to 256x256 voxels before processing. Final magnitude images were generated by root-sum-of-squares combination across channels. For comparison different number of filters per layer (128 to 512) and iterations (10k to 100k) were tested. Qualitative inspection showed that models using less than 384 filters reconstructed only very blurry images. Output quality generally increased with number of iterations. All reconstructed images retained a significant amount of SENSE ghosting, specifically for high contrast regions like the skull, rendering the results unusable. Since the DD relies on the affinity of CNNs for structures over random noise, this is not surprising and difficult to counteract. With the cartesian sampling pattern given, DD can in consequence only be used to target data without structured artifacts. In conclusion the DD as used in [2] is not directly applicable to data sampled in cartesian fashion. Future research could instead try to include DD for optimized sensitivity estimation in mSENSE or recovering k-space directly in GRAPPA like methods. [1] R. Heckel and P. Hand, arXiv:1810.03982, Feb. 2019. [2] S. Arora, V. Roeloffs, and M. Lustig, in ISMRM 2020. [3] N. Weiskopf et al., Front Neurosci, vol. 7, p. 95, Jun. 2013.

Introduction: Most movements in daily life are embedded in motion sequences forming linked movements, which are associated with specific motor memories. Interference tasks, where reaches to a target have to be adapted to one of two opposing force fields, have been used to investigate the ability to form two separate motor memories. If the direction of a linked prior movement (pre-movement) is predictive of the force field’s direction, adaptation to the forces is possible, while other cues (e.g., static visual) do not allow this. It is yet unclear whether the facilitative effect of linked movements could also be achieved by motor imagery of pre-movements. The aim of the project is therefore to investigate how the sensorimotor system can be modulated by linked active and imagined pre-movements to enhance motor learning. Methods: We compared motor learning performance, measured by direction specific adaptation in an interference force field paradigm, of three experimental groups: 1) active pre-movement group, 2) motor imagery pre-movement group, 3) visual cue (no pre-movement) group. Movement kinematics (Kinarm-Exoskeleton-Lab, Kinarm, Canada) and EEG (64 electrodes EEG cap, Brain Products, Germany) were recorded simultaneously to investigate the underlying neural mechanisms involved in the motor learning process. Results: In line with previous research, we show that active pre-movements facilitate adaptation to opposing force fields. Additionally, preliminary behavioral results replicate previous findings that stationary visual cues, indicative of force field direction, do not allow the formation of separate motor memories. Moreover, we have first tentative evidence that motor imagery of pre-movements might be sufficient for motor adaptation to occur at least to a certain extent. Outlook: More data needs to be collected to confirm our tentative behavioral results. In addition, we will analyze EEG data to shed light on neural mechanisms during learning of linked movements, including how neural states and movement kinematics relate to each other.

Language arose around 100 thousand years ago and is assumed to be a human trait. In the human brain, language is supported by a neural network connecting Broca’s area with Wernicke’s area through a dorsal and a ventral pathway. Despite extensive research, how the language network evolved is still debated. Given that soft tissue does not fossilise, it is impossible to compare human brains to those of other hominids as one can only extrapolate brain features based on the size and shape of the skull. To gain insights on the evolution of brain connectivity it is then necessary to look at our closest ancestors, chimpanzees and macaques. Furthermore, chimpanzees show milestones in the emergence of fine motor skills and socio-cognition comparable to humans, placing them as a model of great interest to study language evolution. Most of the current knowledge on primate brain connectivity originates from tracer studies in monkeys, such as macaques. Researchers now use comparative neuroscience to retrace the evolution of our brains based on the shared features between species. By using knowledge of the tracts anatomy of both humans and NHPs they challenge their results and revise the anatomy of those tracts using a common technic in all species. The last decade of research has provided insights in the anatomy of language relevant tracts like the arcuate fascicle and its homologue in the NHPs. A striking feature seems to be the endings of the tract. Indeed, the arcuate ends at the supra temporal gyrus in chimpanzees but reaches further in the human temporal lobe. In macaques the arcuate homologue has weaker frontal projections and does not reach further than the supra temporal gyrus. Such knowledge can be combined with data on vocal production in NHPs to propose an evolutionary model of brain connectivity.

Neurofeedback is a type of biofeedback that uses the real-time display of neural signals to facilitate self-regulation of brain activity. During neurofeedback training, participants try to up- or down-regulate a certain dynamical parameter calculated from the recorded neuronal activity (e.g., average amplitude of oscillations). It has been shown that these training sessions can lead to clinical improvements in a number of brain disorders: ADHD, stroke, Parkinson’s disease, PTSD, etc. However, previous studies report that up to 50% of the participants are unable to properly modulate their brain activity via neurofeedback, and the reasons why this occurs are still unclear. While neurofeedback learning success is typically evaluated by examining changes in the target neuronal activity, the co-occurring modulation of distributed neuronal networks is usually not investigated. We hypothesize that the observed success in self-regulation can be attributed to subject-specific spatio-temporal patterns of distributed activity and connectivity preceding up- or down-regulation of the target neuronal parameter during neurofeedback. In the present study, we plan to develop a computational approach for the extraction of such patterns and apply it to publicly available datasets of neurofeedback training protocols. Since patterns of self-regulation can greatly depend on the mental strategy that participants use, we will first validate our method using the data from a brain-computer interface (BCI) training with feedback, where subjects had a fixed and well-investigated mental strategy – motor imagery of left and right hand movements. After validation, the approach will be applied to neurofeedback recordings where subjects have no explicit strategy, in order to analyze individual aspects of their neurofeedback learning.

Introduction: Languages of the world strongly differ from one another in all dimensions (Evans and Levinson, 2009), and each language relies on a particular neural network adapted to its processing demands (Ge et al., 2015; Goucha, 2019). Similar to the specialization of the brain to the characteristics of the mother tongue, efficient processing of novel structures in adult foreign language learning was previously related to brain plasticity (Li et al., 2014; Qi and Legault, 2020). Here we analyzed the longitudinal structural changes of the white matter language connectome during adult second language learning in a large and well controlled cohort. Method: We recruited 60 healthy right-handed Arabic native speakers (mean age, 25.9 years; range, 19-34) for an intensive German course over a 6-months period. We acquired high spatial resolution diffusion MRI data from each participant at the beginning (time point 0: TP0), after 3 months (TP1) and after 6 months (TP2) of language learning. Computed the structural network between all language-related areas in both hemispheres. We first analyzed global intra-hemispheric connectivity change by testing the brain lateralization at each time point in a paired t-test. We then tested the intra- and inter-hemispheric connectivity change longitudinally using a Linear Mixed Effects (LME) model with each time point as fixed effects. To localize the learning-induced connectivity change across each time point to specific subnetworks, we used the Network-Based R-statistic (NBR) mixed-effects models (p-threshold = 0.01, K = 3000 permutations)(Zeus Gracia-Tabuenca et al., 2020). Results: The initial lateralization test showed a leftward lateralization of the network for the initial and the middle timepoint (TP0: Left >> Right, t= 3.11, p=0.003; TP1: Left > Right, t = 2.04, p= 0.046; TP2: Left ≯ Right, t = 1.79, p= 0.08). And we found a significant dynamic decreased inter-hemispheric connectivity during leaning with the strongest effect in the second half of the learning period (TP0-TP1: t = -1.1, p = 0.27；TP1-TP2: t = -6.2, p = 1.4e-08，TP0-TP2: t= -8.1, p =1.2e-12). The NBR showed an increased intra-hemispheric connectivity in sub-networks, including the bilateral parietal-temporal system and the right IFG mainly in the second half of the learning period. Additionally, the connectivity of sub-networks including connections of orbital IFG – aSTG, parahippocampal- lateral temporal lobe and inter-hemispheres were decreased.

The ability to understand speech is crucial for daily communication and social interactions. The healthy brain is capable of compensating for and adapting to degraded speech signals, e.g. the ability to focus on a single speaker and understand the speech within a noisy environment. Deficits in speech perception, including perception of degraded speech, has been documented in patients with Alzheimer’s disease (AD) and primary progressive aphasia (PPA). However, there is a lack of dynamic tests that can probe these processes experimentally in the context of neurodegenerative diseases. Here, we present data on the development of new dynamic probes of speech perception in two linked experiments. First, we use noise-vocoding as a model to parametrically reduce the amount of information present within speech signals and ascertain the minimal amount of ‘channels’ needed for speech to remain intelligible. Second, we explore associative processing of vocoded emotional speech stimuli. Results show profile differences in perception of degraded speech, verbal and nonverbal, amongst AD and PPA patients. These results are a step towards furthering our understanding of auditory processes and networks, the differences in auditory processing between two major groupings of neurodegenerative diseases and allowing for the possibility of utilising dynamic probes of speech perception as a promising biomarker in AD and PPA.

Vascular risk factors (VRFs) in midlife have been hypothesised to play a prominent role in the aetiology of depression in later life: VRFs are thought to slowly induce damage in white matter tracts, and older adults are at a high risk of developing depressive symptoms if this damage predominantly affects white matter tracts underlying mood regulation. However, direct evidence from longitudinal studies on the mediating role of white matter is still lacking. In the current study, we analysed longitudinal data from 7246 middle-aged and older adults from the UK Biobank. Vascular risk factors were assessed at baseline, white matter microstructure at the first follow-up, and depressive symptoms at the second follow-up. Longitudinal mediation models were fitted to the data, while including demographic and socioeconomic variables as covariates. Our results revealed that VRFs show a small association with depressive symptoms at follow-up. Exploratory analyses suggested that VRFs are driven by somatic symptoms of depression. However, these associations were not mediated by white matter microstructure; neither by overall nor specific tracts. Our results do not support the hypothesis that integrity of white matter tracts links vascular risk and depressive symptoms. Future studies should explore promising alternative mechanisms, such as inflammation, while focusing on associations with specific types of depressive symptoms.

Glioblastoma multiforme (GBM) is the most malignant and heterogenous primary brain tumour with an overall survival of only 15 months after therapy. There are two main hypotheses that suggest the formation of GBM tumour in the brain- one suggests that it arises from dysregulated neural stem cells (NSCs) that transform to form the tumour, and the other theory suggests that cancer or glioma stem cells (GSCs) initiate and cause progression of GBM. It is crucial to understand the origin of the tumour to elucidate specific and effective targets for therapy. To understand the gene expression differences between these two types of stem cells, RNA Seq analysis was carried out on 20 RNA sequence samples of NSCs and GSCs each, selected from the SRA study SRP200400 on the NCBI database. Using the T-Bioinfo platform™, gene expression data in FPKM (Fragments Per Kilobase of transcript per Million mapped reads) units was obtained which was normalized and log scaled and further used to perform principal component analysis (PCA), differential expression analysis (DGE), hierarchical clustering and gene ontology (GO) studies on DAVID. DGE analysis showed that there were 192 significantly differentially expressed genes in GSCs (padj. value <0.05, log2fold change >3). and 156 significantly differentially expressed genes in NSCs (padj. value <0.05, log2fold change

Anticipatory looking (AL) paradigms that measure mentalizing have been increasingly challenged for their reliability and validity, because of the mixed results in replication studies. This study attempts to advance an AL paradigm by adding a prompted instruction to trigger action prediction, using a multi-trial paradigm and matched false-belief and true-belief conditions. Notably, we prompted participants to predict actions, but not mentalize beliefs, which remained the mentalizing implicit. Through comparing the prompted AL paradigm with a well-established explicit mentalizing task, we aim to investigate if it can reliably trigger action prediction. Since compensation is one of the factors that may influence mentalizing performance, our second aim is to explore its relationship with individual differences in compensatory and autistic traits. Mentalizing abilities of 62 adults were measured by the aforementioned tasks. Individual differences in compensatory behavioural and autistic traits were assessed with questionnaires. All the false-belief and true-belief conditions were able to trigger action prediction in the AL task. Although the false-belief reasoning in the AL task and the self-reported compensatory behavioural traits were not related, they were both significantly correlated with the explicit mentalizing performance. Both mentalizing tasks therefore seems to tap into the same underlying cognitive mechanism, supporting the prompted AL task as a mentalizing test. Only the explicit mentalizing task appears to involve compensation, possibly triggering alternative strategies to reason logically about mental states, whilst the prompted AL task was not susceptible to compensation, so it seems to be a purer measure of mentalizing.

Critical decisions, such as in domains ranging from medicine to finance, are often made under threatening circumstances that elicit stress and anxiety. The negative effects of such reactions on learning and decision-making have been repeatedly underscored. In contrast, here we show that perceived threat alters the process by which evidence is accumulated in a way that may be adaptive. Participants (n = 91) completed a sequential evidence sampling task in which they were incentivized to accurately judge whether they were in a desirable state, which was associated with greater rewards than losses, or an undesirable state, which was associated with greater losses than rewards. Prior to the task participants in the ‘threat group’ experienced a social-threat manipulation. Results show that perceived threat led to a reduction in the strength of evidence required to reach an undesirable judgement. Computational modelling revealed this was due to an increase in the relative rate by which negative information was accumulated. The effect of the threat manipulation was global, as the alteration to evidence accumulation was observed for information which was not directly related to the cause of the threat. Requiring weaker evidence to reach undesirable conclusions in threatening environments may be adaptive as it can lead to increased precautionary action.

In functional magnetic resonance (fMRI) studies of Parkinson’s disease (PD), a general linear model (GLM) associated with blocks of finger tapping and rest is a consolidated method to assess motor activity. However, this method has two limitations: First, due to the strong magnetic field and the confined environment of the cylindrical bore, it is troublesome to accurately monitor motor output and, therefore, variability in the performed movement is typically ignored. Second, given the loss of dopaminergic neurons and ongoing compensatory brain mechanisms, a PD patient's motor control is abnormal. These two limitations directly impact the standard GLM predictor that assumes a regular and constant amplitude movement output. Thus, consideration of results from online recordings of the real motor output may lead to significant sensitivity improvements. This was shown in previous work using a non-magnetic glove to capture details of the patients' finger movements. In the present study we aimed to test the robustness of this method in an expanded cohort (n=31), applying more conservative statistics. Our findings suggest that the integration of kinematic modeling offers a general improvement in detecting activations in subcortical structures, such as the basal ganglia. Kinematic modeling can specifically benefit investigations of neurodegenerative diseases with disruption of key elements of the motor loop.

The hippocampal-entorhinal system is remarkably efficient in inferring and organizing hidden relations of sensory input in the form of a cognitive map. It is likely that such a representation enables goal-directed behaviour and facilitates the maximization of rewards. Therefore, the question arises what form such a cognitive map should take. Is it most useful to represent structural relationships veridically, or could it be useful to reflect reward contingencies in a structural representation? Here we aim to address this question in a two-part fMRI study. Participants will initially acquire knowledge about the relationships between visual objects, which follow a hidden graph structure. Behavioural data suggest that participants can acquire such relational knowledge and utilize this to find shortcuts in the decision-making task. In the second part of the study, each of these objects is associated with fluctuating reward values. The current value of each object depends on an underlying latent reward structure. Critically, two parts of the graph structure share the same reward contingencies. We find that participants can realize this and use it to their advantage by generalizing over states sharing the same reward contingencies. Additionally, we find that participants can apply their acquired structural knowledge to correctly infer the values of objects whose values they never directly experienced. We hypothesize that these abilities rely on a map-like representation of the graph structure in the hippocampal-entorhinal system, which interacts with an overlapping latent reward structure. This reward structure could either directly wrap the initial representation in the hippocampal-entorhinal cortex or could be represented independently in mPFC. Mechanistically, we hypothesize that generalization and inference require reactivation of objects that are informative about the relevant reward value during feedback or choice. We will test this hypothesis by applying an fMRI decoding approach.

Everyday decisions require us to plan ahead and predict how valuable different choice options will be in the future. Efficient prediction of such future values is facilitated by an internal model representing state transitions and associated reward contingencies in a task. Recently, it has been suggested that the hippocampal-entorhinal system represents relationships between states in a map-like format akin to its representation of relationships between landmarks in physical space. Such a map-like representation allows for efficient computation of future states. It is unclear, however, whether and how these spatial coding principles and geometric computations might contribute to mapping abstract dimensions such as values. Here, we aim to test the idea that the hippocampal-entorhinal system supports prospective decision making by representing values of choice options in a two-dimensional space. To this end, we introduce a novel decision task where participants repeatedly decide between two options with changing values over a sequence of timepoints. Crucially, these sequences represent trajectories in an abstract two-dimensional value space akin to physical space. Recognizing the nature of these trajectories allows for prediction of future values of the two options. We hypothesize that fMRI responses in the entorhinal cortex (and possibly value-coding prefrontal regions) associated with these trajectories exhibit a hexadirectional modulation in line with grid cell coding found in spatial navigation research. Moreover, as such a map-like representation enables efficient computation of future states, we expect a correlation between the strength of this grid signal and participants’ model-based choice behavior. Furthermore, accurate prediction of future states should be reflected in prospective activation in line with the model-based option before choice. Ultimately, this study can help us elucidate whether the hippocampal-entorhinal system adopts spatial coding principles to support efficient prospective value-based decision making.

Brain responses vary considerably from moment to moment, even to identical sensory stimuli. This has been attributed to changes of instantaneous neuronal states reflecting the system’s excitability. Yet the spatiotemporal organization of these dynamics and their influence on the perception of external stimuli remain poorly understood. In a series of three somatosensory stimulation paradigms in humans, we examined the interplay of two markers of cortical excitability, pre-stimulus oscillatory activity in the alpha band (8-13 Hz) and thalamo-cortical excitatory post-synaptic potentials (EPSP) inferred from short-latency somatosensory evoked potentials (SEP) in the EEG, as well as their association with the perceived stimulus intensity. Furthermore, we characterized the temporal structure and the spatial specificity of these dynamics on a single-trial basis using scaling analysis and source reconstruction of hand- and foot-related SEPs, while controlling for variability in peripheral nerve activity. Pre-stimulus alpha oscillations influenced initial cortical SEPs, which were in turn associated with changes of the perceived stimulus intensity – yet unrelated to variability of subcortical or peripheral nerve activity. Furthermore, these fluctuations of cortical excitability, reflected in both oscillations and SEPs, consistently followed a temporal power law, thus indicating the existence of long-range temporal dependencies. Moreover, comparing these dynamics between spatially distinct somatosensory regions (i.e., foot and hand areas), we observed a somatotopic organization of the relation between ongoing neuronal states and stimulus-evoked responses. Our findings demonstrate that already at initial cortical processing instantaneous neural excitability modulates how intense a stimulus is perceived. Despite these cortical dynamics representing topologically confined effects, they also seem to be embedded in global system activity that is organized in a scale-free manner, as indicated by the observed long-range temporal dependencies. This may reflect a delicate balance between robustness and flexibility of neural responses to sensory stimuli, enabling the brain to adaptively change the neural encoding of even low-level features, such as the stimulus´ intensity.

In our everyday life, we need to execute numerous goal-directed actions. To successfully carry them out, it is essential to have experience-based knowledge about associated environmental or bodily effects. Hence, the anticipation of desired effects is critical for goal-directed behaviour. The anticipatory behaviour control theories, closely related to predictive coding, further suggest that the actions are cognitively represented in terms of their anticipated sensory consequences and these consequences mediate planning and initiating of voluntary actions. However, the question of how we are able to flexibly select and compare different abstractly represented action plans (especially, the plans with comparable goal outcomes) remains an open issue. With this research project, we will investigate the neuronal mechanisms underlying the abstract representation and comparison of goal-directed action sequences. We propose “cognitive maps” as a potential mechanism for such representation. Placing different action sequence options on the continuous two-dimensional cognitive map along the common dimensions (for example, according to the predicted sensory outcomes) could give us a very flexible criterion for comparing those options. Importantly, this mechanism could further explain our ability to form more effective action plans on the abstract level. In the current study, we will focus on approach and avoidance actions, planning to investigate whether these types of actions could be organized in abstract conceptual spaces using the information about predicted successful goal interaction or avoidance. We will first concentrate on approach actions and use a behavioural paradigm comprising several behavioural tasks, that are executed in immersive virtual reality. Analogously to an object location memory task often used to investigate conceptual spaces, participants will learn to execute certain action sequences and remember the associated environmental and bodily effects to get the goal items. We hypothesize, that gathering information about actions in the physical space should result in the extraction of goal-related key effects and, based on those effects, the construction of abstract action effect space. To this end, we will collect MRI data and test whether the action effect space was constructed in the hippocampal-entorhinal, prefrontal and parietal cortices.

We study patterns of partial synchronization in a network of FitzHugh–Nagumo oscillators with empirical structural connectivity measured in human subjects. We report the spontaneous occurrence of synchronization phenomena that closely resemble the ones seen during epileptic seizures in humans. In order to obtain deeper insights into the interplay between dynamics and network topology, we perform long-term simulations of oscillatory dynamics on different paradigmatic network structures: random networks, regular nonlocally coupled ring networks, ring networks with fractal connectivities, and small-world networks with various rewiring probability. Among these networks, a small-world network with intermediate rewiring probability best mimics the findings achieved with the simulations using the empirical structural connectivity. For the other network topologies, either no spontaneously occurring epileptic-seizure-related synchronization phenomena can be observed in the simulated dynamics, or the overall degree of synchronization remains high throughout the simulation. This indicates that a topology with some balance between regularity and randomness favors the self-initiation and self-termination of episodes of seizure-like strong synchronization.

Speech comprehension is subserved by cyclic electrophysiological activity, so-called neural oscillations. Oscillations within the delta band (i.e., < 4 Hz) are thought to synchronize or entrain to the pace of intonational phrase boundaries (IPBs). By means of entrainment, oscillations inherit acoustic rhythms that are present in speech to persist after stimulation offset (e.g., Kösem et al., 2018). Therefore, it could be possible that through entrainment, delta-band oscillations maintain the prior intonational rhythm (Breen, 2014; Frazier et al., 2006), thus triggering the perception of an IPB in an upcoming sentence. To test this, we combined an initial prosody entrainment with a subsequent visual target sentence. Target sentences were ambiguous, that is, they allowed for two different segmentation options (Hoeks et al., 2002; e.g., “Max sees Tom and Karl laughs”). We reasoned that the duration of the prior contour, corresponding to the duration of one of the segmentation options, would affect segmentation, and thus interpretation of the upcoming sentence. We conducted three online experiments. A trial started with one of two prosodic contours, repeated three times to induce entrainment. The contour was followed by word-by-word rapid serial visual presentation of a target sentence; the sentence-final words were self-paced. Segmentation was inferred from reading times, as well as from RTs to comprehension questions. The results demonstrated a significant speed-up after a longer prosodic contour at the first self-paced word—apparently, participants expected the sentence to continue after this word, thus accelerating. Therefore, long contours may enable the prediction of long upcoming segments, while short contours trigger the prediction of short segments. Prosody entrainment may thus serve a predictive function at the prosody–syntax interface, not only helping the reader to segment the current speech input, but also inducing predictions of the duration of upcoming constituents (Breen, 2014; Frazier et al., 2006; Grosjean, 1983). We are currently conducting an MEG experiment to assess the underlying electrophysiological dynamics: we expect to find delta-band phase differences, as well as power enhancement at stimulation frequency. References: Breen (2014) LANG LINGUIST COMPASS. Frazier et al. (2006) TRENDS COGN SCI. Grosjean (1983) LINGUISTICS. Hoeks et al. (2002) J MEM LANG. Kösem et al. (2018) CURR BIOL

The neocortex facilitates a wide variety of cognitive functions such as sensory processing, motor output and associative capabilities serving as higher-order perceptual frameworks. While functional properties of different neocortical areas are distinguishable and have been investigated extensively, the microstructural properties and the resulting implications remain unclear in a number of areas. In this study, I plan to investigate microanatomical properties of the primary auditory cortex in the mouse, primate and human brain tissue using a variant of the IDISCO+ tissue clearing technique coupled with Light Sheet Microscopy for large-scale three-dimensional image acquisition. By employing a pattern recognition algorithm I termed Gabor Spherical Shell filtering, I previously managed to reliably identify cell soma positions and cell radii in a cleared whole-brain sample of mouse brain. I am planning to map structural lateralization of social communication in mice, non-human primates and humans using the set of tools I previously established. This will give new insights into structures that might facilitate the specialization of the left auditory cortex hemisphere on social cues in different mammalian species.

Navigation and processing of complex environments is greatly facilitated by forming and applying a variety of different concepts. Concepts represent combinations of features shared by similar entities and allow generalisation from limited experience to novel situations. Recent research suggests that concept formation is supported by multidimensional cognitive maps within the hippocampal-entorhinal system. Cognitive maps are proposed to encode multiple relationships between experiences in a common representational format allowing for flexible decision making and novel inferences. How this map-like representation of concepts affects inferences in situations with incomplete information will be investigated in the planned study. In particular, it will be examined to what degree the hippocampus together with visual areas employ abstract representations (i.e. prototypes) for the completion of a missing feature. Participants will first be trained to categorise a set of exemplars based on the ratio of their features. Subsequently, they will encounter exemplars that exhibit only one of the two initial features and are instructed to complete the missing feature according to their category label. Our model-based analysis approach will test whether an unseen prototype is formed during categorisation that attracts feature inferences towards the category center and which can be decoded from hippocampal and visual cortex activity patterns.

Representational Similarity Analysis (RSA) has emerged as a popular tool in cognitive neuroscience to compare representations of different modalities, species, and models (e.g., brains or deep neural networks, DNNs). RSA works by comparing representational dissimilarity matrices (RDM), which characterize the dissimilarity between all pairs of conditions (e.g., visual stimuli). However, classical RSA assumes that individual features of measured representations or models (e.g., brain voxels, DNN units) all contribute equally to the final similarity estimate. This is a strong assumption and contrasts with other common methods, such as multivariate decoding, that allow for a linear reweighting of individual features. Recent developments counteract this by reweighting features of DNN models for use in RSA, and indeed find improved performance (Jozwik et al., 2017; Peterson et al., 2018; Storrs et al., 2020). Here, we systematically evaluated the degree to which this feature-reweighted RSA (FR-RSA) can serve as a general purpose method for improving the sensitivity of measured representations. To this end, we used three publicly available datasets of real-world object images (84, 92, and 118 conditions, respectively), testing the correspondence between DNNs, behavior, fMRI, and MEG for MEG-fMRI fusion. For all datasets and models, we found that FR-RSA consistently increased the correspondence of pairs of RDMs, sometimes doubling the amount of shared variance. Sanity checks and the fitting algorithm’s design ensured that this was not due to overfitting. RSA is commonly deployed to estimate how well different models capture how the brain represents relations between stimuli or conditions. FR-RSA could become a general purpose method that further refines this process, ultimately facilitating the selection between competing models.

Endovascular interventions (EIs) have become the gold standard in cardiology as they are far less invasive than open surgery and are associated with shorter recovery times. However, due to their minimally invasive nature, the operator is faced with visuo-motor challenges while conducting such a procedure. Therefore, in order to perform an EI successfully a diverse set of skills and extensive training is necessary. Nevertheless, to date it is not clear which core skills drive successful performance and which factors cause differences in low -and high-level performers. In this study, we aimed to tease apart the driving force behind learning to perform an EI by examining training-related plasticity in grey matter volume using voxel-based morphometry. Further, we examined whether we can predict performance based on baseline grey matter volume. To examine grey matter plasticity as a result of learning to perform endovascular procedures, we trained medical students naïve to EIs over three days on an endovascular simulator. A sperate control group performed the basic first part of this EI to control for visuo-motor execution and the training environment, but were not otherwise trained on EI procedures. At baseline, one day prior to training and directly after the last training session, T1-weighted scans were acquired from participants in both groups. The results from this study will be presented on this poster.

A wealth of previous research has investigated the spatial profile of receptive fields in the human visual cortex, with receptive fields generally considered to increase in size along the visual hierarchy. In contrast, little is known about the temporal dynamics of information transmission through the visual cortex. In the present study, we investigated how receptive field size might vary over time as information flows through the visual hierarchy. Participants (n=9) viewed a circular stimulus sequentially flashed across the visual field while electroencephalogram (EEG) data were recorded. We used forward encoding modelling to map the visual field and investigate the temporal dynamics of population receptive fields. First, we showed that the precision of position representations (in polar angle) was roughly equal across eccentricities. This indicates that neurons responding to regions of visual space closer to the fovea have receptive fields covering a smaller visual angle than peripheral visual neurons, consistent with previous studies that used invasive recordings or functional magnetic resonance imaging. In addition, we observed that population receptive fields transiently narrowed 100ms and 200ms after stimulus onset. This timing is consistent with activation of early visual cortex during an initial feedforward sweep of activity, and a subsequent reactivation following a feedback sweep. Our study demonstrates how EEG can provide unique insight into the temporal dynamics of spatial information in the visual system.

This study investigates the neural and behavioural dynamic of two people engaging in naturalistic social interaction. In preparation of the experiment, participants are asked to learn facts about a set of obscure items (musical instruments, ancient obejcts, exotic food and animals). During the lab visit, they teach (and learn) these facts to (and from) their research partner. During the 16 trials of the social learning task, participants can interact without particular constraints - a part from the designated and alternating role of teacher and student. For half of the trials, a separator is placed in between participants, obscuring each others' view. Brain activity is measured via fNIRS from a total of 38 channels per participants across bilateral parietal, temporal and frontal areas. Measurements of breathing, speech, head movement and video recordings are also collected. After the lab visit, learning performance is measured via free recall. A multiple choice quiz is then completed to take measures of their immediate performance (same day) and long-term recall (one week later). This study is being currently at the phase of data collection. We aim to develop an analysis pipeline that can help asnwer questions about the inter-brain dynamic of social learning and how this relates to behavioural measures (breathing, movement, speech) as well as performance.

It is widely believed that when someone copies you, you will like that person more. Is this also true for copying choices? Here, we use an online interactive art preference task to examine how participants respond to confederates who do or do not copy their art choices. To implement this study, we used Zoom to have live conversations between participants and trained confederates, QuickTime Player to capture video footage of the conversations and Gorilla to obtain outcome measures of participant impressions. This involves the complex integration of multiple software systems to try to replicate live social interactions in an online setting and maximise ecological validity. The results will show how people respond to being copied by a confederate and showcase how interactive psychology research can be conducted in an online environment.

Our preferences influence our choices (we choose the things we like) but recent evidence suggests that our choices also influence our preferences (we like the things we have chosen). Here we present two studies investigating how participants learn about their own preferences through choice, and whether they are motivated to do so. In the first study, participants rated the value of, made choices about and rated again multiple activities. Participants showed a choice-induced preference change, whereby enjoyment ratings of activities increased after participants had chosen the activity, but decreased for those not chosen. We found that this effect was strongest when: i) alternatives were more similarly valued by participants; ii) participants were less certain about the values of the alternatives in the first place; iii) participants were eventually able to reach a more confident choice. This lends support to the hypothesis that individuals use the process of choosing to reassess and increase certainty about their own preferences. In a second study, participants were able to freely switch between a sustained attention task for monetary reward and a preference choice task without monetary reward. Participants were willing to forgo monetary reward in order to spend time making preference choices, either for themselves or another person, but were less willing to forgo money to make perceptual choices about the item images. These studies suggest that the process of choosing enables individuals to learn about and gain certainty over their own preferences, and potentially because of this, they find making preference choices rewarding. Next, we will assess age-related changes in these processes over adolescence, a period where gaining certainty over our preferences may be particularly important, and how they relate to symptoms of depression, in which difficulty in decision making is a core symptom.

Performance of actions is suggested to rely on predictive processing mechanisms mapping together sensory inputs and statistical regularities of simple action-goal relationships. These mappings take the form of different action representations that need to be constantly accessed and retrieved in the execution phase, in order to guide action in complex behavioral contexts. However, little is known concerning what brain areas are supporting the retrieval of action representations. To address this point, we developed an event-related functional magnetic resonance imaging (fMRI) experiment in which participants are required to perform finger-tapping sequences with their right hand. The experiment is structured so that four types of stimuli are going to be presented, each one coding a specific sequence or one of the following classes of sequences: sequences sharing (i) similar motor features; (ii) an abstract rule; (iii) no specific motor or abstract information. During every trial, participants will be provided first with a class or sequence cue (contextual signal) and then exclusively with a sequence cue (execution signal), corresponding to the sequence to be performed. We hypothesize that the posterior-most subregion within left Brodmann Area (BA) 44 is responsible for accessing mental representations of actions and that its activity reflects the specificity of information carried by a certain stimulus. We believe that the more information is accessed when seeing a specific contextual signal, the more the activity will be observed within BA44. Such a finding would enrich our knowledge concerning the various processes the region can be assumed to correlate with: i.e., retrieving and maintaining mental representations of actions until execution. Furthermore, this would bring evidence to the assumption that action hierarchies—contrary to language-derived ones—are to be found in representational levels and not in structural and syntactic-like ones. At this stage, the experiment has been piloted at a behavioral level (n = 20). Results show that the lower the contextual signal is in the representational hierarchy (i.e., the more specific it is), the shorter the onset time (OT) of the finger-tapping sequence. However, the transition across finger movements is not affected by the position of the signal within the hierarchy. This suggests that context guides behavior providing help in selecting premotor representations, but it does not affect motor execution itself.