Winter 2018

Posted on by conor.houghton
 
September, 21st – Mark Humphries (University of Nottingham)
 
3.34 Physics Building
 
The plasticity of population activity in prefrontal cortex is independent of learning
 
The prefrontal cortex is thought to represent our knowledge about what action is worth doing in which context. But we do not know how the activity of neurons in prefrontal cortex collectively changes when learning which actions are relevant. Here we show in a trial-and-error task that population activity in rat prefrontal cortex is persistently changing, irrespective of whether the animal shows evidence of learning. Only during overt learning of the correct action are the accompanying changes to population activity carried forward into sleep, suggesting a long-lasting form of neural plasticity. Our results suggest that representations of relevant actions in prefrontal cortex are acquired by reward imposing a direction onto ongoing population plasticity.
  
 September, 28th – Tom Baden (University of Sussex)
 
3.34 Physics Building
 
The Evolution of Function in the Brain: What can we learn from the vertebrate retina? 
 
Sighted animals use their eyes in vastly different ways, and therefore evolved a staggering array of visual specialisations to navigate their individual visuo-ecological niches. These specialisations are deeply rooted at every level of visual systems, from the optical properties of eye to functional and structural retuning of neuronal microcircuits in the retina and brain. Taking advantage of the exquisite experimental accessibility of the larval zebrafish visual system and drawing on available knowledge gathered in the retina of other species with different visuo-ecological demands, I will present our lab’s recent efforts to better understand how animals can retune their retinal circuits for efficient sensory processing. Focussing on colour vision, I will highlight how zebrafish use different circuit motifs in different parts of their eyes to simultaneously support differential visual requirements imposed by the need for feeding, predator avoidance and object recognition.
 
Key References
Zimmermann MJY*, Nevala NE*, Yoshimatsu T*, Osorio D, Nilsson DE, Berens P, Baden T § . 2018. Zebrafish differentially process colour across visual space to match natural scenes. Current Biology 28(1-15).
Franke K*, Berens P*, Schubert T, Bethge M, Euler T § , Baden T § . 2017. Inhibition decorrelates visual feature representations in the inner retina. Nature; doi: 10.1038/nature21394. link. 
 Baden T*, Berens *P, Franke K*, Roman Roson M, Bethge M, Euler T § . The functional diversity of mouse retinal ganglion cells. Nature. doi:10.1038/nature16468.
 
October, 5th – Demian Battaglia (University Aix-Marseille)
 
Physics 3.34
Perception, cognition and behavior rely on flexible communication between microcircuits in distinct cortical regions.
It has been proposed based on growing experimental evidence that changing patterns of oscillatory coherence support flexible information routing. The stochastic and transient nature of oscillations in vivo, however, is hard to reconcile with such a function.
Here, through a computational modelling approach, we add a new chapter to this debate between “oscillo-partisans” and “oscillo-skeptics”, by showing  that models of cortical circuits near the onset of oscillatory synchrony are well able to selectively route input signals despite the short duration of oscillatory bursts and their stochastic-like irregularity. In canonical multi-areal circuits, we find that gamma bursts spontaneously arise with matched timing and frequency and that they organise information flow by large-scale routing states. We thus hypothesise that a self-organized network-wide re-organization of routing could be induced by suitable weak control perturbations or minor modulations of background activity.
Information routing constitutes nevertheless just a component of neural information processing by neural circuits. Moving to the analysis of actual electrophysiological recordings in hippocampus, enthorinal cortex and prefrontal cortex of anaesthetised and sleeping rats, we investigate whether dynamic changes between oscillatory modes also affect ongoing computational manipulations of information within local circuits, beyond inter-circuit routing. Through an unsupervised algorithmic approach, we are able to identify a multiplicity of internal “computing micro-states”, characterized by the flexible recruitment of alternative hub neurons, transiently specialising in different primitive operations of information processing (buffering and funneling). We find that global oscillatory states have an impact on both the “dictionary” of available computing micro-states and on the “syntax” of their sequences, whose complexity is systematically boosted by e.g. the presence of theta oscillations vs slow-oscillation dominated states.

 


October, 19th – Ruth Betterton (University of Bordeaux)
 
Physics 3.34
 
A biophysical network model of CA3, hippocampus: functional architecture and learning induced changes
A key function of the brain is the storage and recall of information as memories. The hippocampus and specifically area CA3 are involved in the rapid encoding of short-term spatial, episodic, and contextual memory. A unique feature of the CA3 network is the presence of recurrent excitatory cell connectivity which has led to the theory that CA3 acts as an attractor or auto-associative network. It is thought that, during encoding of new memories connection weights between activated excitatory cells within CA3 are rapidly enhanced through Hebbian plasticity creating a micro-network of cells known as an assembly. The potentiated assembly provides an attractive candidate for the location of the neuronal ‘engram’ or cellular correlate of memory. Using a combination of mathematical modelling and in vitro slice recordings we are gathering evidence of this theory of the importance of recurrent connectivity in hippocampal function. Presented here is the development of a network model of biophysical, multi-compartmental neurons carried out in the simulation environment NEURON. The network includes many features of the in vivo CA3 region including specific inputs, recurrent connectivity and region specific plasticity rules.


November, 2nd – Gareth Barnes (UCL)
 
Physics 3.34
 
A new generation of MEG scanners
 
I will talk about collaborative work between University College London and the University of Nottingham to use optically pumped magnetometers (OPMs) for human brain imaging. These sensors have comparable sensitivity to current cryogenic devices but do not require cooling. This means that the sensor array can be worn (rather than climbed into) and the smaller separation between sensor and brain means optimal (and improved) signal to noise ratio in all subject cohorts. I will talk about our initial modelling and experimental work with these new sensors. One of the exciting advances has been to keep these arrays operational during head-movement through a static magnetic field. This has opened up many new clinical and neuroscientific possibilities and I will talk about some of our experiences with these new paradigms.
 

November, 9th – George Stothart (University of Bath)

Physics 3.34

Using Fast-Periodic-Visual-Stimulation to assess cognition in neurological disorders 

 Fast Periodic Visual Stimulation (FPVS) provides a new objective method for assessing an individual’s ability to discriminate between different categories of visual stimuli. Using a combination of steady state visual evoked potentials and oddball paradigms it has been demonstrated to be a powerful measure of visual discrimination in single subjects. Importantly what defines the visual categories can range from low-level perceptual properties to abstract cognitive properties. We have adapted this approach to examine a range of cognitive processes and will demonstrate that the technique can be used to assess the integrity of semantic categorisation, short term memory and visuo-spatial processing in single subjects in as little as 3 minutes EEG recording time. The implications for the objective assessment of cognition in dementia and the potential as an early diagnosis tool will be discussed.

researchportal.bath.ac.uk/en/persons/george-stothart

 

November, 23th – Jiaxiang Zhang (Cardiff University)
 
Physics 3.34
 

The neurocognitive mechanisms of voluntary decision  

We can voluntarily make decisions to fulfil our goals and desires, even when all the options are similar to each other. This talk will discuss our work on using brain imaging, electrophysiology and computational modelling to understand the cognitive processes underlying such voluntary decisions. First, I will present fMRI evidence that during voluntary decision, a decision network in the medial frontal cortex accumulates action intention until a threshold is reached. Second, the behavioural randomness in a sequence of decisions fluctuates over time and correlated with both fMRI and MEG activity in the frontopolar cortex. This suggests a cortical network sensitive to information regularities, which concurrently monitor the choice in voluntary decisions. Last, I will present recent results on how perceptual salience and action outcomes affect the behavioural, EEG, and metacognitive measures in voluntary decisions. Our results highlight the potential and challenges of establishing a neurobiological theory of voluntary behaviour in humans.

 
 
November, 30th –  Dimitrije Marković (TU Dresden)
Physics 3.34

Anticipating changes: decision-making with temporal expectations 

Being able to experience time and build temporal expectations about future events is essential for our everyday activities and survival. Despite the central role that time plays in our lives, the neuronal and computational mechanisms that link our experience of time with decision-making remain poorly understood. In this talk, I will focus on the computational underpinning of decision-making with temporal expectations and present a probabilistic behavioural model that enables a systematic investigation of the interplay between temporal expectations and behaviour. The central assumption here is that humans form prior beliefs about the temporal regularities of a dynamic environment; these beliefs shape both the inference and the planning process. Using a sequential reversal learning task, I will illustrate the key properties of the model and demonstrate how it can be applied to behavioural data to infer prior beliefs of participants and to investigate interindividual behavioural differences.

Leave a Reply

Your email address will not be published. Required fields are marked *