Analysis of EEG data in Matlab

Do you want to learn how to analyze neurophysiological data, but aren’t sure how to translate concepts to code or how to implement your ideas in Matlab? During this course, you will learn the necessary mathematical foundations of signal processing and how to implement and use standard methods for the analysis of M/EEG data in Matlab.
Course levelAdvanced Bachelor/Master, open to PhD staff and professionals
Session 2
18 July to 1 August 2020
Recommended session 1 course combination
Neurodegenerative Diseases: Foundation, Tools and Perspectives, Single Cell Technologies in Life Sciences, The Beautiful Mind: Global Perceptions of Mental Health
Co-ordinating lecturersSimon Houtman
Other lecturersDr Klaus Linkenkaer-Hansen, Dr Sonja Simpraga, and others
Form(s) of tuitionLectures, excursions, discussions, group work
Form(s) of assessmentPoster (60% presentation, 40% content), exam (multiple choice)
ECTS3 credits
Contact hours45
Tuition fee€1150, read more about what's included.
This course is for junior scientists. Both Master’s students and PhD students who have experience with data analysis are welcome. You should be interested in developing a deeper understanding of the underlying mechanisms of data analysis of neural time series data. Bachelor students are welcome too, provided they have prior experience with Matlab programming. If you have doubts about your eligibility for the course, please let us know. Our courses are multi-disciplinary and therefore are open to students with a wide variety of backgrounds.
You will be using (and writing!) code in Matlab for the analysis of EEG data. Prior experience with Matlab programming is obligatory. There will be no time to learn Matlab from scratch during this course, so make sure you have followed at least one introductory course if you are not yet proficient with Matlab. The most important requirement is that you are enthusiastic!
Do you want to learn how to analyze neurophysiological data, but aren’t sure how to translate concepts to code or how to implement your ideas in Matlab? During this course, you will learn the necessary mathematical foundations of signal processing and how to implement and use standard methods for the analysis of M/EEG data in Matlab.

Neuronal oscillations are generated at many spatial and temporal scales of neuronal
organisation and are thought to provide a mechanism for the coordination of spatiotemporally distributed brain activity. Due to the multi-scale properties of neuronal oscillations (time, space, and frequency), it is important to analyse the neurophysiological data using multiple spatial and temporal measures. Doing so allows us to better understand changes in the brain caused by either disease, experimental manipulations, or therapeutic interventions.

Through the course, you will learn about several different measures related to EEG,  which are used commonly in the field of EEG research. These measures include power, and more novel methods, including the weighted phase-lag index, which is used to quantify phase synchronization of neuronal oscillations as a measure of information exchange between brain regions. You will also learn about detrended fluctuation analysis (DFA), which can be used to quantify long-range temporal correlations in neurophysiological data. There will be a guest lecture on critical brain dynamics given by a founder of the field of self-organized criticality within EEG.

This course will have a heavy focus on the fundamental skills required for the analysis of M/EEG data. Each day will consist of several lectures and Matlab practicals. During the practicals, you will learn how to use and implement code in Matlab, which will give you a foundation for understanding the concepts discussed during the lectures. There will be ample opportunity for questions and individual guidance (and coffee).

During the practicals, you will work in a group of around 5 students to create a scientific poster to address a specific research question. On the final day of the course, everyone will present their poster to the other groups.

The evaluation for the course will be based on the scientific poster (40% content and layout, 60% individual presentation) and an exam (multiple choice). We have computers available with Matlab pre-installed, but feel free to bring your own laptop with Matlab installed.

Upon completion of this course you should be able to:

  • Understand the biophysics of electroencephalography (EEG) and magnetoencephalography (MEG)
  • Understand the basics of sinusoidal signals and their mathematical representation
  • Understand the mechanisms of the Fourier transform to represent signals in the frequency domain
  • Understand and use complex wavelet convolution to extract time-frequency information from neural time series data
  • Use non-parametric statistics including the cluster-based permutation test
  • Simulate EEG data to test the effects of parameters within your analyses
  • Carry out preprocessing of the EEG to remove artifacts
  • Use independent component analysis (ICA) to remove eye artifacts
  • Perform source modeling to counteract the adverse effects of volume conduction and enhance the spatial resolution of EEG data
  • Analyze functional connectivity of source-reconstructed EEG data to study brain networks and compare across different groups or conditions
  • Write scripts and functions for analysis of EEG data
  • Perform clustering on EEG data and use machine-learning algorithms to classify groups or conditions

There will be a full-day excursion, including a lecture, to either BioSemi in Amsterdam or Philips EGI in Eindhoven. Both companies are important developers and producers of EEG hardware.
Facebook-icoon   linkedin-icoon   Instagram-icoon
;