Talk:Feature Selection and Extraction for a BCI based on motor imagery

From AIRWiki
Jump to: navigation, search

Part 1: Project profile

Project name

Feature Selection and Extraction for a BCI based on motor imagery

Project short description

The aim of this project is to explore and analyze methods, pointed out from literature, for feature selection and extraction in a BCI system based on motor imagery.

Dates

Start date: 0?/0?/2008

End date:

People involved

Project head(s)
Other Politecnico di Milano people
Students currently working on the project

Laboratory work and risk analysis

Laboratory work for this project will be mainly performed at AIRLab-IIT/Lambrate. The main activity consists in the acquisition of brain signals through an EEG amplifier for on-line or off-line processing. This is a potentially risky activity since there is an electrical instrumentation that is in direct contact with the human body. It is thus important to keep the system isolated from the power line. The EEG amplifier (as all biomedical instrumentations) is certified by the vendor to be isolated and the acquired data are transferred to the PC using an optic fiber connection . Anyhow for increased safety the PC and any other electronic device connected to the system must be disconnected from the power line. Standard safety measures described in Safety norms will be followed.

Part 2: Project description

L'obbiettivo principale che questo progetto si prefigge è l'analisi di algoritmi di filtraggio spaziale applicati in un sistema BCI basato su Motor Imagery. L'obbiettivo specifico dell'analisi è la valutazione dell'impatto dei suddetti algoritmi sulle capità di classificazione del sistema attualmente in fase di sviluppo presso IIT-Lab. Selection of algorithm to analize was made looking for the most performant once inside literature and also BCI Competition' results. After this preliminary selection three filter's algorithm was retained:

  • ICA
  • CSP
  • CSSD

Brief description of Motor Imagery BCI architecture

Even if all BCI system share the same goal, create a new communication pathway from mind to an external device, this isn't always true for architecture, due to different type of brian signal observed. However, we could trace a common design shared from most BCI application. Project work was focused on Signal Processing block, more specifically on Featur Filtering subblock.

Analysis Tool

An offline analysis tool was developed to generate spazial filter starting from data recorded by BCI2000 enviroments. Tool also provide functionality to test and visualize filters' application result. Tool was develop inside Matlasb eviroments due to compatibility with BCI2000 analysis tool and also for

Data Import Module

Prefiltering Module

Prefiltering module preprocess data to remove unuseful information for filter generation process. Application provides two kinds of proprecess step:

  • Whitening White imported data removing mean and normalizing variance to 1
  • Frequency filtering Bandpass filter imported data in selected frequency range

Filter Selection Module

ICA

Independent Component Analysis is a blind source separation tecnique applied to recover sorce from a mixture signal observation. ICA algorithm emploies statistical mutal indipendence of component to extract soruce signal from mixture one. We found different algorithm depending on contrast funcion used to maximize independence. We analize two ICA algorithm, Infomax and FastICA, which seems to have good reults when applied to BCI data.

Infomation maximization algorithm, also kwon as Infomax, was based on approach developed from Bell & Sejnowski, that find a solution maximization the joint entropy of the output H(y). This is the same to minimize mutual information I(y) of the output of a neural network. An implementation fo this algorithm was found... Algorithm have many parameters:

  • learning rate Learning rate of neural network learning algorithm. Default value was set to
  • momentum Momentum of neural network learning algorithm. Default value was set to
  • bias
  • max step Maximum number of step that algorithm shuold be perform
  • initial weight A random matrix of inizial weight
  • epsilon

FastICA algorithm search for independent component maximizing the negentropy J(y) of each mixtured signal. Algorithm parameters are:

  • approach Value could be symmetric, in wich all signal was ,or deflaction
  • nonlinarity Type of nonlinearity used to compute contrast function
  • epsilon Lowest bound of prediction error
  • max iteration
  • sample size [0,1] Percentage value of imported dato to use as training set
CSP

Common Spatial Pattern algorithm simultaneus diagonalize covariance matrix of class labeled signals to find a linear subspace, i.e., a linear trasformation matrix, in such a way that combined features maximize variance for one class and minize its for other class. Standard CSP is designed for binary classification's problem, then extension to multi-class problem was done, as suggest in literature, by splitting original problem into several binary decision, computing each filter like one-vs-rest. After computing n matrix, with n number of class to analize, then algorithm output a spatila matrix Algorithm params that could be tuned are:

  • sample size [0,1] Percentage value of imported dato to use as training set
  • retained eigenvalues Integer. Number(s) of eigenvalue(s) from each class matrix to retain in output spatial matrix.
CSSD

Common Spatial Subsample Decomposition is a statistical method derived from common spatial pattern analysis. Also this method, like CSP, look for linear trasformation of orginal signal to obtain combined features that hold maximum variance, i.e. most usefull information for discrimint process, for selected class. Algorithm params that could be tuned are:

  • sample size [0,1] Percentage value of imported dato to use as training set
  • retained eigenvalues Integer. Number(s) of eigenvalue(s) from each class matrix to retain in output spatial matrix.

Part 3: References

Part 4: Project Tracking