A survey of signal processing Flashcards
Key questions addressed in this paper
Key Signal processing components of BCI, Signal Processing Algorithms used in BCI, promising Signal Processing Techniques
General Structure of BCI System was explored in this paper. What are the three components of BCI Transducer?
Artifact Processor, Feature Generator, Feature Translator
Artifact Processor
Removes the artifacts from the electrical signal after it has been amplified.
Feature Generator
Transforms resultant signals into feature values that correspond to the underlying neurological mechanism employed by the user for control
Feature Generator can be concatnetation of these three components
signal enhancement, feature extraction, feature selection / dimensionality reduction
In some BCI designs, why is pre-processing performed prior to the extraction of features?
To increase the signal-to-noise ratio of the signal.
What is the goal of feature selection / dimensionality reduction component?
Reduce the number of features and channels used so that very high dimensional and noisy data are excluded.
What does Feature Translator do?
Translates the features into logical(device-independent) control signals, such as a two-state discrete output.
What kinds of algorithm does translation algorithm use?
Linear Classification Methods[Classical Statistical Analyses] / Nonlinear Classification Methods [Neural Networks]
Feature Translator may consist of these two components
Feature Classification, Post-Processing
Main goal of feature classification component
classifying the features into logical control signals
The role of the BCI Transducer
Translating the user’s brain activity into logical (or device-independent) control signals.
What does Electrophysiological sources refer to in BCI systems
Neurological Mechanisms or processes employed by a BCI user to generate control signals.
