Machine Learning on Neuroimaging Data
This course is an introduction to the elds of neuroimaging (acquiring the data from the brain) and machine learning (techniques to extract useful knowledge from the data). We will have a look at such neuroimaging techniques as EEG (electroencephalography), fMRI (functional magnetic resonance imaging), fNIRS (function near-infrared spectroscopy), intracranial recordings and study the data these methods produce: continuous signal, spiking data, voxel activation data. After that we will explore the notion of machine learning, see how our data can be described in machine learning context and what useful information machine learning can provide us with. On the last lecture we will introduce one particular application of machine learning on the brain data: brain-computer interface and play with Emotiv EPOC - portable EEG device.
Lecture 1: Neuroimaging Techniques
- The technology behind EEG, fMRI, fNIRS and microelectrode arrays
- EEG data and Fourier transform
- Spiking data from intracranial recodings
- Voxel activation data from fMRI
Practice Session 1: Spiking data & Fourier Transform on EEG Data
- Extract the orientation the neuron is tuned to from the spiking data
- Perform Fourier transform on EEG data
Lecture 2: Introduction to Machine Learning
- Machine learning terminology and concepts
- Performance measures to estimate accuracy of the model
- Demonstration on a simple example
- Representing EEG data for machine learning
Practice Session 2: Machine Learning on EEG data
- Perform the signal processing on real EEG data
- Run the machine learning algoritm
- Analyze the results
Lecture 3: Brain-Computer Interfaces
- The concept and the purpose
- Real-time machine learning on EEG data
- Best existing solutions
- Criticism on the mass media coverage of the BCI research
Practice Session 3: Hands on Emotiv EPOC
- Record some data using Emotiv EPOC device
- Apply the techniques developed during second practice session
- Analyze the results
Course prerequisites: student should be familiar with basic algebra and have basic programming skills (Matlab/Octave). These requirements are not strict, the course is mostly self-contained and it is possible to learn all the required concepts on the go.
Tutor
Mr. Ilya Kuzovkin
Country: Estonia
Place of employment: PhD student, University of Tartu, Estonia
Spheres of researches: Machine Learning in general and it's applications to Brain-Computer Interface systems. Data mining. Text algorithms. Cryptology. Bioinformatics.
E-mail:This email address is being protected from spambots. You need JavaScript enabled to view it.