Abstract
Independent component analysis (ICA) is a powerful tool for separating
signals from their observed mixtures. This area of research has
produced many varied algorithms and approaches to the solution of this
problem. The majority of these methods adopt a truely blind
approach and disregard available a priori information in order
to extract the original sources or a specific desired signal. In this
contribution we propose a fixed point algorithm which utilises a
priori information in finding a specified signal of interest from the
sensor measurements. This technique is applied to the extraction and
channel isolation of sleep spindles from a multi-channel
electroencephalograph (EEG).
Go back