June , 28th – Thomas Wills
Geog Sciences G.11N SR1
June, 14th – Shana Silverstein
Geog Sciences G.11N SR1
June, 7th – Denize Atan
Geog Sciences G.11N SR1
Mai, 24th – Tim Howe
Geog Sciences G.11N SR1
Title: Extending evidence for REM-associated Replay in Hippocampal CA1 Place Cells.
Abstract: During periods of inactivity, hippocampal CA1 neurons with spatial receptive fields (“place cells”), reactivate in patterns that recapitulate previously experienced spatial sequences, a phenomenon known as replay. CA1 replay is most prominently associated with sharp-wave ripple (SWR) events during non-REM sleep or quiet wake, and has been implicated in the consolidation of episodic memory. Replay has also been reported during REM sleep [1], however evidence for this phenomenon is substantially less extensive than for replay during non-REM. Non-REM replay occurs on a temporally compressed timescale (approximately 8 times faster than during active behaviour) around brief, discrete SWR events. REM-replay appeared less temporally compressed and occurred during extended periods of elevated theta power, necessitating alternative detection methods to those established for SWR-associated replay. Using tetrode recordings from adult rat dorsal CA1, we present data that corroborate the existence of REM-replay. The activity of multiple place cells was recorded simultaneously while rats performed simple goal-directed maze tasks, and during subsequent extended rest periods in a sleep box. Replay was detected using a moving-window correlation algorithm (from [1]), and confirmed with complementary approaches including hidden Markov model (HMM) and Bayesian trajectory decoding. Extending evidence for REM replay paves the way for analyses exploring its experience-dependence, extra-hippocampal correlates and function contributions.
[1] Louie K & Wilson MA (2001) Neuron 21: 145-156
Mai, 17th- Anna Schapiro (University of Pennsylvania)
Geog Sciences G.11N SR1
http://sleepandcognition.org/anna-schapiro.html
Title: Learning and consolidating patterns in experience
Abstract: There is a fundamental tension between storing discrete traces of individual experiences, which allows recall of particular moments in our past without interference, and extracting regularities across these experiences, which supports generalization and prediction in similar situations in the future. This tension is resolved in classic memory systems theories by separating these processes anatomically: the hippocampus rapidly encodes individual episodes, while the cortex slowly extracts regularities over days, months, and years. This framework fails, however, to account for the full range of human learning and memory behavior, including: (1) how we often learn regularities quite quickly—within a few minutes or hours, and (2) how these memories transform over time and as a result of sleep. I will present evidence from fMRI and patient studies suggesting that the hippocampus, in addition to its well-established role in episodic memory, is in fact also responsible for our ability to rapidly extract regularities. I will then use computational modeling of the hippocampus to demonstrate how these two competing learning processes can coexist in one brain structure. Finally, I will present empirical and simulation work showing how these initial hippocampal memories are replayed during offline periods to help stabilize and integrate them into cortical networks. Together, the work provides insight into how structured information in our environment is initially encoded and how it then transforms over time.
Mai, 10th- Andrea Martin (MPI for Psycholinguistic)
Geog Sciences G.11N SR
Title: On neural systems, oscillations, and compositionality
Abstract: There continues to be vibrant controversy about the fundamental relationship between the information in biological signals and the neural systems that represent and process them. Compositionality is a property of a system such that the meanings of complex entities are derived from the meanings of constituent entities and their structural relations. It is a crucial part of what enables human thought and language to “make infinite use of finite means,” but also part of what makes human thought and language difficult to account for within extant theories of cognition, artificial intelligence, and human neurobiology. I focus on this foundational puzzle and discuss the computational requirements, including the role of neural oscillations, for what I believe is necessary in order to compose structures and meanings within the constraints of a neurophysiological system.
Mai, 3th – Camin Dean
Geog Sciences G.11N SR1
Cancelled
April, 18th – Bridget Lumb
Geog Sciences G.11N SR1
Postponed
April, 11th – Chris Bailey
43 Woodland Rd G.10 LR
Cancelled
April, 4th – Naoki Masuda
43 Woodland Rd G.10 LR
March, 29th- Sam Berens (University of York)
43 Woodland Road, LR G.10
Title: Learning and memory in an uncertain world
Abstract: We often need to pick up on subtle patterns and learn complex associations in our environment; even when its unclear which pieces of information are important. How is this achieved? I will discuss some of my recent behavioural and fMRI work exploring how we are able to acquire knowledge under uncertain conditions and in the absence of feedback (so-called ‘unsupervised learning’). These studies test various computational models of learning, investigate whether some types of information are preferentially retained or consolidated, and examines the role of metacognitive learning intentions.
March, 22th- Gareth Barker (University of Bristol)
43 Woodland Rd G.10 L
Title: There and back again: Investigations into associative recognition memory network function.
Abstract: Associative recognition memory, our ability to form an association between an object and its spatio-temporal context, is critical for everyday memory function. A network of brain regions critical for associative recognition memory has been identified, however how these brain regions function as a network during associative recognition memory formation is poorly understood at present. We investigated the role of connections between three key nodes in the network, the hippocampus, medial prefrontal cortex and nucleus reuniens of thalamus, by using a combination of optogenetic and chemogenetic approaches.
By manipulating specific connections within this thalamo-cortico-hippocampal memory network, we have revealed that distinct types of associations rely on anatomically distinct projections and have identified distinct, but interleaving circuits for associative recognition memory encoding and retrieval.
March, 15th- Rui Ponte Costa (University of Bristol)
43 Woodland Rd G.10 LR
Title: Powerful learning via cortical microcircuits
Abstract: Cortical circuits exhibit intricate excitatory and inhibitory motifs, whose computational functions remain poorly understood. I will start out by introducing our work on how state-of-the-art recurrent neural networks used in machine learning may be implemented by cortical microcircuits. In addition, our new results suggest that such biologically plausible recurrent networks exhibit better learning of long-term dependencies. However, learning in such networks relies on solving the credit assignment problem using the classical backpropagation algorithm that appears to be biologically implausible. I will finish my talk discussing our recent work on a biologically plausible solution to the credit assignment problem using well-known properties of cortical microcircuits, which approximates the backpropagation algorithm. Overall, our work demonstrates how cortical microcircuits may enable powerful learning in the brain.
March, 8th- Helen Barron (University of Oxford)
43 Woodland Rd G.10 LR
https://www.mrcbndu.ox.ac.uk/people/dr-helen-barron
Title: Inhibitory engrams in memory storage and recall
Abstract: Memories are thought to be represented in the brain by activity in groups of neurons described as memory engrams. Although memory engrams are typically thought to be made up of excitatory neurons, several recent studies suggest that inhibitory neurons also contribute. Indeed, by matching their excitatory counterparts, selective inhibitory interneurons may facilitate a stable storage system that allows memories to lie quiescent unless the balance between excitation and inhibition is perturbed. Here I will present a set of studies that show evidence for selective neocortical inhibition in the human brain using ultra-high field 7T MRI and brain stimulation. I will show that matched excitatory-inhibitory engrams provide a stable storage mechanism for neocortical associations, and protect memories from interference. Finally, I will explore how neocortical memory engrams might interact with the hippocampus during recall, to selectively perturb excitatory-inhibitory balance.
March, 1st- Natalie Doig (University of Oxford)
43 Woodland Rd G.10 LR
https://www.mrcbndu.ox.ac.uk/people/dr-natalie-doig
Title: Structure is Function: Cellular and Network Substrates of Basal Ganglia Dynamics
Abstract: In order to fully understand how the dynamic functions of the nervous system are realised we must evaluate its structure through static measures. In this talk I will discuss two studies which employed a range of neuroanatomical methods to reveal specific cellular and network principles of the organisation of the basal ganglia. In the first study I will discuss the use of modern trans-synaptic tracing techniques to examine the cell type selective connections between nuclei of the basal ganglia. Second, I will highlight the features of a novel connection between the dorsal hippocampus and the nucleus accumbens that shapes memory guided appetitive behaviour. Using these examples, I would like to promote a discussion on the advantages and disadvantages of specific neuroanatomical techniques and what they can tell us about the substrates underlying the neural dynamics of the basal ganglia.
February, 22th- Maria Wimber (University of Birmingham)
43 Woodland Rd G.10 LR
http://www.memorybham.com/maria-wimber/
Title: Tracking the temporal dynamics of memory reactivation in the human brain
Abstract: Our memories are not static. Each attempt to retrieve a past event can adaptively change the underlying memory space. Here I discuss my work on the neurocognitive mechanisms that enable the selective retrieval of episodic memories. I present behavioural and electrophysiological (M/EEG) work that provides insight into how a memory trace unfolds in time during retrieval, on a sub-trial scale. Further, I show evidence from a series of fMRI studies in which we track the representational changes that occur in a memory trace over time and across repeated retrievals. The latter findings demonstrate that retrieval adaptively modifies memories by strengthening behaviourally relevant and weakening behaviourally irrelevant, interfering components. Together, this work sheds light onto the neural dynamics of the retrieval process, and informs theories of adaptive memory.
February, 15th- Jim Dunham (University of Bristol)
43 Woodland Road, G.10 LR
Computing pain – Real time signal processing in human pain nerves.
January, 18th- Quentin Huys (UCL)
Geog Sciences G.11N SR1
Perceptual conditioning
https://www.quentinhuys.com/research.html
January, 11th – Vitor Lopes dos Santos (Oxford)
Life sciences G14
Neural oscillations