Cognitive and Computational Neuroscience

Fundamentals of Neuroscience

The module provides an introduction to core aspects of contemporary neuroscience, and it will consider the current state of knowledge in the field, central theoretical issues and key practical approaches. Topics that are discussed include: neural signalling, sensation and sensory processing, movement and its central control, the 'changing brain' (development and plasticity in the nervous system) and complex brain functions.

15 credits

Computational Neuroscience 1: Biologically Grounded Models

This module starts with a primer on neuroscience and the role of computational neuroscience. The next part of the module covers abstract neuron models and introduce classic computational principles and learning rules related to neural networks. From there we move to more biologically grounded models and deal with single neuron models including leaky-integrate-and-fire and conductance-based neurons. Finally, we examine higher levels of description, in particular systems in context of reinforcement learning. While the emphasis throughout the module is on methodological issues, how models can be built, tested and validated at each level, we will also draw connections to specific brain regions to motivate and illustrate the models.

15 credits

Computational Neuroscience 2: Theoretical Models

The module builds on ideas developed in Computational Neuroscience 1 to explore networks of neurons, neural circuits and their dynamics, and models of complete brain systems. As in Computational Neuroscience 1, this is taught using both mechanistic (bottom-up) and theoretical (top-down) perspectives but, in this module, emphasis is placed on computational models of neuronal networks and systems.

15 credits

Mathematical Modelling and Research Skills

This module develops basic skills required to understand and participate in research in computational and cognitive neuroscience. The course begins with a refresher course in reading and writing skills, understanding of quantitative data and basic algebra and calculus. The course moves on to cover more advanced mathematical modelling techniques including matrix algebra, ordinary differential equations and optimisation methods. Programming skills are introduced via the MatLab modelling language. All topics are illustrated by application to concrete modelling examples relevant to contemporary neuroscience.

15 credits

Systems Neuroscience

The module provides an advanced understanding of the brain's major computational systems and the theoretical or model-driven approaches to research of these topics. Major processing units of the brain will be described and, where appropriate, emphasis will be placed on understanding each of these structures in terms of both their micro- and macro-circuitry. One focus of the module will be to impart an appreciation of how many fundamental questions relating to brain function requires study at a range scales, from single cell to whole brain and behaviour. The various strategies adopted for investigating and modelling brain-circuits, and the consideration of circuits as the defining feature of brain systems, will be presented.

15 credits

Research Project in Cognitive Neuroscience

The module allows students to work on an extended research project within computational neuroscience and/or cognitive neuroscience and/or systems neuroscience and/or analysis of brain imaging data. Students will learn and apply appropriate research techniques, analyse and interpret the results, and write up the research findings using recognised journal frameworks. Students will receive guidance and regular feedback from their supervisors. The project culminates in an oral presentation and a written dissertation.

75 credits

Pathway one:

You will train in the fundamentals of biological and artificial cognition, and the perspectives of computational biology, including a specialist module taught by the Department of Computer Science.

Fundamentals of Cognition

The module provides an overview of the fundamental issues in cognitive neuroscience and its contributory disciplines. The approach taken is in terms of its development over the past 50 years, providing an overview of the key concepts in the information processing approach and in cognitive science, followed by an analysis of the advances that have been made recently using cognitive neuroscience techniques. Topics include: fundamental issues in cognition (memory, attention, learning, language); theoretical approaches including cognitive neuropsychology, symbolic and sub-symbolic modelling; and methodological issues.

15 credits

Modelling and Simulation of Natural Systems

This module will provide a practical introduction to techniques used for modelling and simulating dynamic natural systems. Many natural systems can be modelled appropriately using differential equations, or individual based methods. In this module, you will explore and understand both modelling approaches. You will gain knowledge of the assumptions underlying these models, their limitations, and how they are derived. You will learn how to simulate and explore the dynamics of computational models, using a variety of examples mostly drawn from natural systems.  At the end of the module, we will introduce basic recurrent neural networks as examples of dynamical systems with multiple timescales. You should be aware that there are limited places available on this course.

15 credits

Pathway two:

You will build on your theoretical skills before exploring advanced topics in neuroimaging techniques, including multi-electrode electrophysiology, optical imaging, EEG and structural and functional MRI.

Neuroimaging 2

This module provides an overview of neuroimaging techniques and fundamental data analysis methodologies employed, specifically those based around functional magnetic resonance imaging (fMRI). The two aspects of neuroimaging (techniques and data analysis) will be taught over the semester. For neuroimaging techniques, after introducing the physical principles underlying fMRI, a description of fMRI-based methods for mapping brain structure and function will follow. For neuroimaging data analysis, the general linear model methodology will be introduced based on the software SPM (Statistical Parametric Mapping), which is one of the most widely used packages for fMRI data analysis. Issues concerning fMRI experimental de-sign and efficiency will also be discussed and taught in depth.

15 credits

Neuroimaging 1

This module provides an overview of neuroimaging techniques and fundamental data analysis methodologies. Specifically, it will focus on the functional imaging techniques of electrophysiology, EEG, and MEG, optical methods and calcium imaging, each of which will be introduced in the lecture component of the module. In the associated lab classes, students will gain first-hand experience of analysing and processing data sets arising from these techniques.

15 credits

Whichever route you choose, this training will develop your transferable skills in critical reading, writing, project management, computational modelling, data analysis and visualisation, and scientific programming using the languages Python and MATLAB.