Block-Seminar "Algorithms for Brain Reading and Writing"

This seminar is an optional course in the module "Machine Learning 1 (ML1-X)" and is worth 3 LP (3 ECTS credits).

General Information

Description

Many results in modern neuroscience are derived from non-invasive neuroimaging data, most notably electrophysiology (electro- and magnetoencephalography, EEG/MEG) and functional magnetic resonance imaging (fMRI). In this seminar, we will review the steps that are needed in order to eventually draw neurophysiological conclusions from such data ("brain reading"). The focus will be on algorithms and mathematical methods used within each step along the processing pipeline. The seminar will cover EEG forward models and inverse solutions, MRI segmentation and registration, statistical source separation as well as en- and decoding approaches. We will also cover "brain writing" using transcranial current stimulation and algorithms for optimal targeting.

Each student is required to choose a particular topic from a predefined list of topics and to present it at the end of the semester. The presentation will be based on original research papers, which are provided for each topic.

Basic knowledge in mathematical topics such as linear algebra, numerical optimization, and statistics are helpful for this course.

If you are interested in taking the course, please send Stefan Haufe an email.