Applications of Source Detection in Biomedical Signals

The detection of the sources is a mathematical problem that can find application in all that fields in which a mixture of different signals is available: it is widely used in acoustics, where different sounds can be received by multiple microphones, in decoding communication signals taken by antennas, in separation of seismic data and in image processing, just to name a few. In biomedical signal processing, the source detection can help in decoding the complex brain activity from electroencephalographic (EEG) data, in studying muscle activation from electromyographic (EMG) recordings and also finds some interesting applications in electrocardiography (ECG), where maternal and foetal ECG were considered as two distinct sources and thus were separated. The term blind source separation (BSS) arises from the fact that the solution is recovered with a blindly approach: limited or no a priori information is assumed on the sources and on the propagating medium in which the signals travel to reach the receiver. With this approach, also useful information on the interposed medium can be obtained. In the following, an introduction on BSS and on the most used methods will be given. Subsequently, some biomedical applications in the electroencephalography and in the electromyography will be discussed with examples and simulation results.