Application of convex optimization techniques for feature extraction from EEG signals

The analysis of electroencephalogram signals is essential for extracting the relevant information from the human brain activities in order to diagnose brain diseases and abnormalities. The visual screening of long term electroencephalogram recordings is a time consuming and difficult task, and it is insufficient for capturing reliable information from brain activities. Challenging these issues, mathematical modelling and data analysis of electroencephalogram signals were presented, four automatic feature extraction methods were developed based on convex optimization problems, the classification of these signals was performed, and different statistical measures were investigated and employed. This research can facilitate better patient care services while save time and reduce costs. Of particular interest was the automated detection of transient electroencephalogram waveforms called K-complex and Epileptic seizure that contribute to sleep stage identification and seizures diagnosis respectively. The novel methods are also suitable for other types of signals.