This thesis investigates how the neural system instantiates selective attention to speech in challenging acoustic conditions, such as spectral degradation and the presence of background noise. Four studies using behavioural measures, magneto- and electroencephalography (M/EEG) recordings were conducted in younger (20–30 years) and older participants (60–80 years). The overall results can be summarized as follows. An EEG experiment demonstrated that slow negative potentials reflect participants’ enhanced allocation of attention when they are faced with more degraded acoustics. This basic mechanism of attention allocation was preserved at an older age. A follow-up experiment in younger listeners indicated that attention allocation can be further enhanced in a context of increased task-relevance through monetary incentives. A subsequent study focused on brain oscillatory dynamics in a demanding speech comprehension task. The power of neural alpha oscillations (~10 Hz) reflected a decrease in demands on attention with increasing acoustic detail and critically also with increasing predictiveness of the upcoming speech content. Older listeners’ behavioural responses and alpha power dynamics were stronger affected by acoustic detail compared with younger listeners, indicating that selective attention at an older age is particularly dependent on the sensory input signal. An additional analysis of listeners’ neural phase-locking to the temporal envelopes of attended speech and unattended background speech revealed that younger and older listeners show a similar segregation of attended and unattended speech on a neural level. A dichotic listening experiment in the MEG aimed at investigating how neural alpha oscillations support selective attention to speech. Lateralized alpha power modulations in parietal and auditory cortex regions predicted listeners’ focus of attention (i.e., left vs right). This suggests that alpha oscillations implement an attentional filter mechanism to enhance the signal and to suppress noise. A final behavioural study asked whether acoustic and semantic aspects of task-irrelevant speech determine how much it interferes with attention to task-relevant speech. Results demonstrated that younger and older adults were more distracted when acoustic detail of irrelevant speech was enhanced, whereas predictiveness of irrelevant speech had no effect. All findings of this thesis are integrated in an initial framework for the role of attention for speech comprehension under demanding acoustic conditions.

Propofol is one of the most commonly used general anesthetics in the world, and yet precisely how it enables loss of consciousness still eludes us. It exhibits rich spectral characteristics on electroencephalogram (EEG) recordings from human patients, including alpha oscillations (8-14 Hz) and Slow Wave Oscillations (SWO, 0.5-2.0 Hz). Additionally, these two oscillations are phase-amplitude coupled (PAC) in a dose-dependent manner: low doses cause “trough-max” coupling where alpha power is maximal during the trough of the SWO cycle, while high doses cause “peak-max” coupling where alpha power is maximal during the peak of the SWO cycle. These propofol rhythms occur at the same frequencies as sleep spindles and sleep SWO, and likely use the same well-studied thalamocortical circuitry. The study of anesthesia therefore represents a safe method for investigating both how our brains sleep and the much-debated components of consciousness. In this dissertation, I use Hodgkin-Huxley-style computational models of both the thalamus and cortex to explain how the direct and indirect effects of propofol can generate such spectral phenomena. In the first part of this dissertation, I discuss results from a thalamic model. I illustrate how GABAA potentiation by propofol can create sustained alpha oscillations in the hyperpolarized thalamus by utilizing the same mechanisms used by sleep spindles. I then show how the thalamus, under artificial SWO conditions, can output trough-max or peak-max PAC depending on background excitation, GABAA potentiation, and H-current conductance. In the second part of this dissertation, I discuss results from a thalamocortical model. My analysis reveals how, in a simulated EEG signal, trough-max PAC can arise from competition between thalamocortical and intracortical synaptic currents, while peak-max PAC can arise from their cooperation. Furthermore, the coherence of cortical SWO rhythms can directly control whether the system expresses trough-max or peak-max PAC, while the indirect effects of propofol on acetylcholine are required for both PAC states. This culmination of years of work reveals just how complex the inner workings of anesthesia can be in enabling its profound effects.

Neural oscillations, including rhythms in the beta1 band (12-20 Hz), are important in various cognitive functions. Often brain networks receive rhythmic input at frequencies different than their natural frequency, so understanding how neural networks process rhythmic input is important for understanding their function in the brain. In the current thesis we study a beta1 rhythm that appears in the parietal cortex, focusing on the way it interacts with other incoming rhythms, and the implications of this interaction for cognition. The main part of the thesis consists of two stand-alone chapters, both using as a basis a biophysical neural network model that has been previously proposed to model the parietal beta1 rhythm and validated with in vitro experiments. In the first chapter we use a reduced version of this model, in order to study its dynamics, applying both analytic and numerical methods from dynamical systems. We show that a cell can respond at the same time to two periodic stimuli of unrelated frequencies, firing in phase with one, but with a mean firing rate equal to the other, a consequence of general properties of the dynamics of the network. We next show numerically that the behavior of a different cell, which is modeled as a high-dimensional dynamical system, can be described in a surprisingly simple way, owing to a reset that occurs in the state space when the cell fires. The interaction of the two cells leads to novel combinations of properties for neural dynamics, such as mode-locking to an input without phase-locking to it. In the second chapter, we study the ability of the beta1 model to support memory functions, in particular working memory. Working memory is a highly distributed component of the brain's memory systems, partially based in the parietal cortex. We show numerically that the parietal beta1 rhythm can provide an anatomical substrate for an episodic buffer of working memory. Specifically, it can support flexible and updatable representations of sensory input which are sensitive to distractors, allow for a read-out mechanism, and can be modulated or terminated by executive input.

The main aim of the experiments reported in this thesis was to explore the neural mechanisms underlying the temporal orienting of attention. In Chapter 3, I explored the possible dissociation between exogenous and endogenous temporal orienting by comparing reaction times to targets appearing after rhythmic or symbolic cues. Behavioural results provided evidence for the existence of dissociable exogenous and endogenous types of temporal orienting of attention. The experiment in Chapter 4 combined spatiotemporal expectations using rhythmic moving cues to test the modulatory effect of exogenous temporal orienting in the brain. Specifically, I used EEG to test the effect of temporal orienting on perceptual and motor stages of target analysis, as well as on anticipatory oscillatory brain activity. The time-frequency analysis revealed that rhythmic cues can entrain slow brains oscillations, providing a putative mechanism for enhancing the perceptual processing of expected events. Spatiotemporal expectations also modulated the amplitude of visual responses and the timing and amount of preparatory motor activity. In Chapter 5, I used a novel task to explore the neural modulatory effects of spatial and temporal expectations acting in isolation or in coordination. For the first time, the analysis of early visual responses demonstrated that temporal expectations alone, independently of spatial orienting, can enhance early visual perceptual processes. The time-frequency analysis in this experiment showed a desynchronisation of alpha oscillations focused over central-parietal electrodes induced by rhythmic cues that were independent of spatial expectations. When rhythmic cues carried spatiotemporal information, the alpha desynchronisation also spread over contralateral occipital electrodes. In Chapter 6, fMRI was used to test the possible neural dissociation between motor and temporal orienting. The results confirmed the large overlap between these two processes, but also indicated independent behavioural and neural effects of temporal orienting. Temporal orienting activated the left IPS across motor conditions, further implicating the left IPS in temporal orienting. Based on the results of these experiments, directions for future studies are discussed.

In our daily lives, we often make decisions that require the use of self-control, weighing trade-offs between various attributes: for example, selecting a food based on its health rather than its taste. Previous research suggests that re-weighting attributes may rely on selective attention, associated with decreased neural oscillations over posterior brain regions in the alpha (8-12 Hz) frequency range. Here, we utilized the high temporal resolution and whole-brain coverage of electroencephalography (EEG) to test this hypothesis in data collected from hungry human subjects exercising dietary self-control. Prior analysis of this data has found time-locked neural activity associated with each food’s perceived taste and health properties from approximately 400 to 650 ms after stimulus onset (Harris et al., 2013). We conducted time-frequency analyses to examine the role of alpha-band oscillations in this attribute weighting. Specifically, we predicted that there would be decreased alpha power in posterior electrodes beginning approximately 400 ms after stimulus onset for the presentation of healthy food relative to unhealthy food, reflecting shifts in selective attention. Consistent with this hypothesis, we found a significant decrease in alpha power for presentations of healthy relative to unhealthy foods. As predicted, this effect was most pronounced at posterior occipital and parietal electrodes and was significant from approximately 450 to 700 ms post-stimulus onset. Additionally, we found significant alpha-band decreases in right temporal electrodes during these times. These results extend previous attention research to multi-attribute choice, suggesting that the re-weighting of attributes can be measured neuro-computationally.

The selective deployment of attention over time optimises our perception and action at the moments when relevant events are expected to happen. Such "temporal orienting" to moments when something is going to happen is especially useful when this information can be combined with predictions about where and what events are likely to occur. A large body of research has already established how temporal predictions dynamically influence our perception and action, but questions remain regarding the neural bases of these attentional mechanisms. In this thesis I present three magnetoencephalography (MEG) studies that I conducted to investigate anticipatory neural dynamics associated with spatial-temporal orienting of attention for perception and action. I also investigate and discuss how such anticipatory dynamics change with ageing and neurodegeneration in Parkinson's disease (PD), and how these anticipatory neural dynamics behave in situations where a complex, hidden spatial-temporal structure is present. In Chapter 1, I introduce the topic of this thesis by reviewing the literature on temporal orienting of attention and by introducing my specific research questions. In Chapter 2, I present an MEG study on anticipatory neural dynamics of joint spatial-temporal orienting of attention in the visual domain, in younger and older adults. This study shows that neural dynamics with spatial, temporal and spatial-temporal orienting are all differentially affected by ageing. In Chapter 3, I describe an MEG experiment that investigates anticipatory neural dynamics during spatial-temporal motor preparation and compares PD participants to healthy control participants. This study reveals that both behavioural and neural dynamics with temporal orienting are affected in PD. In Chapter 4, I describe an experiment that explores how an implicit spatial-temporal structure is utilised to predict and prepare for upcoming actions. This study shows that motor cortical excitability is dynamically modulated in anticipation of the location and timing of events, even when such expectations are hidden in complex visual-motor sequences that remain largely implicit. In Chapter 5, the General discussion, I place these results in their wider context and discuss limitations and future directions.

No description available.

Many living organisms have the ability to execute complex behaviors and cognitive processes that are reliable. In many cases, such tasks are generated in the absence of an ongoing external input that could drive the activity on their underlying neural populations. For instance, writing the word "time" requires a precise sequence of muscle contraction in the hand and wrist. There has to be some patterns of activity in the areas of the brain responsible for this behaviour that are endogenously generated every time an individual performs this action. Whereas the question of how such neural code is transformed in the target motor sequence is a question of its own, their origin is perhaps even more puzzling. Most models of cortical and sub-cortical circuits suggest that many of their neural populations are chaotic. This means that very small amounts of noise, such as an additional action potential in a neuron of a network, can lead to completely different patterns of activity. Reservoir computing is one of the first frameworks that provided an efficient solution for biologically relevant neural networks to learn complex temporal tasks in the presence of chaos. We showed that although reservoirs (i.e. recurrent neural networks) are robust to noise, they are extremely sensitive to some forms of structural perturbations, such as removing one neuron out of thousands. We proposed an alternative to these models, where the source of autonomous activity is no longer originating from the reservoir, but from a set of oscillating networks projecting to the reservoir. In our simulations, we show that this solution produce rich patterns of activity and lead to networks that are both resistant to noise and structural perturbations. The model can learn a wide variety of temporal tasks such as interval timing, motor control, speech production and spatial navigation.

This thesis investigates how the neural system instantiates selective attention to speech in challenging acoustic conditions, such as spectral degradation and the presence of background noise. Four studies using behavioural measures, magneto- and electroencephalography (M/EEG) recordings were conducted in younger (20–30 years) and older participants (60–80 years). The overall results can be summarized as follows. An EEG experiment demonstrated that slow negative potentials reflect participants’ enhanced allocation of attention when they are faced with more degraded acoustics. This basic mechanism of attention allocation was preserved at an older age. A follow-up experiment in younger listeners indicated that attention allocation can be further enhanced in a context of increased task-relevance through monetary incentives. A subsequent study focused on brain oscillatory dynamics in a demanding speech comprehension task. The power of neural alpha oscillations (~10 Hz) reflected a decrease in demands on attention with increasing acoustic detail and critically also with increasing predictiveness of the upcoming speech content. Older listeners’ behavioural responses and alpha power dynamics were stronger affected by acoustic detail compared with younger listeners, indicating that selective attention at an older age is particularly dependent on the sensory input signal. An additional analysis of listeners’ neural phase-locking to the temporal envelopes of attended speech and unattended background speech revealed that younger and older listeners show a similar segregation of attended and unattended speech on a neural level. A dichotic listening experiment in the MEG aimed at investigating how neural alpha oscillations support selective attention to speech. Lateralized alpha power modulations in parietal and auditory cortex regions predicted listeners’ focus of attention (i.e., left vs right). This suggests that alpha oscillations implement an attentional filter mechanism to enhance the signal and to suppress noise. A final behavioural study asked whether acoustic and semantic aspects of task-irrelevant speech determine how much it interferes with attention to task-relevant speech. Results demonstrated that younger and older adults were more distracted when acoustic detail of irrelevant speech was enhanced, whereas predictiveness of irrelevant speech had no effect. All findings of this thesis are integrated in an initial framework for the role of attention for speech comprehension under demanding acoustic conditions.

Fear memories are necessary to initiate anticipatory fear responses when we are confronted with cues that predict an impending threat. However, when a cue no longer predicts threat, an extinction memory is formed that actively inhibits the expression of the fear memory. Failure to acquire, consolidate, or recall extinction memories causes fear memory expression (i.e., fear responding) in the absence of threat, which is a hallmark characteristic of most anxiety-related disorders and post-traumatic stress disorder (PTSD). Of further importance, these disorders occur approximately twice as often in women than men, which is thought to partially rely on sex hormone mediated differences in fear extinction. Moreover, deficits in extinction memory processing can also hinder the success of extinction-based exposure therapy, which is commonly used to treat these disorders. Thus, a better understanding of the factors determining the quality of extinction memories is of utmost importance. The present thesis focuses on three of these factors including the female sex hormone 17β-estradiol (E2), fear extinction timing, and the noradrenergic arousal system. To examine the role of E2 (Manuscript 1; low E2 levels or high E2 levels) and fear extinction timing (Manuscript 2; either immediately or delayed after the initial fear memory formation), we used a special differential fear conditioning procedure that allowed us to separately assess fear memories and extinction memories via peripheral arousal responses (measured via skin conductance responses [SCR]) and, most importantly, via central neurophysiological indicators (measured via electroencephalography [EEG]). Concerning EEG parameters, we were especially interested in neural oscillations (especially in the theta and gamma range). To further advance the understanding of the neurophysiological foundations of both memory systems, we also aimed at disentangling oscillatory and non-oscillatory brain activity (Manuscript 2). Moreover, the crucial role of the noradrenergic arousal system for the quality of extinction memories is highlighted in a review of relevant rodent and human studies (Manuscript 3). By using the described multi-methodological approach, we were able to demonstrate for the first time that peripheral arousal as well as fear-related theta oscillations are sensitive to E2. This was indicated by less fear responding (attenuated peripheral arousal and attenuated theta oscillations) during the recall of fear and extinction memories under high peripheral E2 levels (Manuscript 1). Concerning the role of fear extinction timing, we demonstrate that delayed extinction is advantageous over immediate extinction in reducing peripheral arousal during the recall of the extinction memory (Manuscript 2). Additionally, by disentangling oscillatory and non-oscillatory brain activity, we demonstrate for the first time that oscillatory and non-oscillatory brain activity is sensitive to fear expression. Moreover, by reviewing different rodent and human studies, we highlight the important role of noradrenergic arousal for the recall of extinction memories and, importantly, provide a detailed mechanistic framework of how extinction deficits might be caused after immediate extinction (Manuscript 3). In sum, the present thesis underscores the important role of E2, fear extinction timing, and the noradrenergic system for the recall quality of fear memories and extinction memories in humans.