Eeg Signal Compression Based On Compressed Sensing Sampling

In this dissertation, a research has been made on the multichannel EEG signals compressed sampling based on compressed sensing, firstly compressed sample the single channel EEG signal, then joint compressed sample the multichannel EEG signals.In medical practice, EEGs are often collected over numerous channels and trials, providing large data sets. How to deal with these data effectively is a problem to be solved. In recent years, there has been a new approach to solve the problem. Compressed sensing is an emerging field which is based on the revelation that a small collection of linear projections of a sparse signal contains enough information for reconstruction. Firstly, this paper introduces a priori knowledge of EEG and compressed sensing theoretical framework. The next research is compressed sampling single channel EEG signals based on the theory of compressed sensing, including finding the redundant dictionary for the better sparse decomposition by comparing the different experiments, the result turns out that the redundant dictionary with the atoms of the Gaussian function and its first and second derivatives is the better one for the EEG signals. Then by using different measurement matrix, such as Gaussian random matrix and Toeplitz matrix and so on, measure the sparse decomposition coefficient vector, finally using orthogonal matching pursuit algorithm to recovery the original signals. Lastly, the paper makes the multichannel EEG signals joint sparse decomposition, joint compression and joint reconstruction by saving more atoms and reducing the times of measurement.