Independent component analysis for identification of artifacts in magnetoencephalographic recordings
NIPS '97 Proceedings of the 1997 conference on Advances in neural information processing systems 10
| Hi-index | 0.00 |
To establish improved therapies for patients suffering from severe neurological and psychiatric diseases, a demand controlled and desynchronizing brain-pacemaker has been developed with techniques from statistical physics and nonlinear dynamics. To optimize the novel therapeutic approach, brain activity is investigated with a Magnetoencephalography (MEG) system prior to surgery. For this, a realtime data acquisition system for a 148 channel MEG and online signal processing for artifact rejection, filtering, cross trial phase resetting analysis and 3D reconstruction of the cerebral current sources was developed. The developed PCI bus hardware is based on a FPGA and DSP design, using the benefits from both architectures. The reconstruction and visualization of the 3 dimensional volume data is done by the PC which hosts the realtime DAQ and pre-processing board. The framework of the MEG-online system is introduced and the architecture of the realtime DAQ board and online reconstruction is described. In addition we show first results with the MEG-Online system for the investigation of dynamic brain activities in relation to external visual stimulation, based on test data sets.