Echtzeitsystem für Phasenrücksetzanalysen und Neuro-Rückkopplungen am MEG

Modern medical imaging systems like Magnetic Resonance Tomography (MRT), Positron Emission Tomography (PET), Single Photon Emission Computed Tomography, Electroencephalography (EEG), or Magnetoencephalography (MEG) are used for the investigation of complex brain functions and their localization. In MEG magnetic fields outside the brain, caused by electrical currents due to the neural activity within the brain, can be measured with high temporal resolution [Cohen03], [Mamivuo95], [Hamlainen93]. In the field of neuroscience the MEG is e.g. used to analyze rhythmic brain activity under stimulation. Based on the high temporal resolution of the MEG simultaneously active neural populations may be identified. These measurements are commonly used in both, research and clinical applications. Existing MEG systems acquire the signals from a huge number of sensors, pre-process the signals measured by the magnetic field sensors and store them for later offline analysis. Neither instantaneous analysis of the measured data, visualization and localization of the electrical sources nor feedback stimulation is facilitated. But, these features would open up new desirable MEG applications. Real time feedback stimulation of the measured and reconstructed brain activity would be very helpful in order to scientifically explore methods for the manipulation of synchronization processes in brain. For this, a data acquisition system for a 148 channel MEG, capable of online reconstruction of the cerebral current density distribution, phase analysis and realtime feedback stimulation has been developed. The developed system is designed as an add-on for an existing BTi MAGNES-2500WH MEG. The usage of a optical splitter allows the utilization of the given fibre optic interface for parallel data acquisition of the new system. Therefore the new MEG-Online system does not affect the original MEG, but adds realtime functionality for feedback experiments and online data analysis, reconstruction and visualization. The realized hardware concept is based on three different signal processing units. The combination of the FPGA, DSP and PC architecture (hybrid system) provides sufficient computational power to fulfil the requirements. In particular, the reconstruction of the current densities of up to 10 voxels can be done within 1 ms. The developed PCI bus add-on hardware is based on a FPGA and DSP design, using the benefits from both hardware architectures. This hybrid technology board enables a standard PC to handle all time critical calculations for the realtime data acquisition, reconstruction of a cerebral current course and feedback signal generation. This enables realtime stimulations based on acquired signals or a reconstructed cerebral current time course. The 3D reconstruction and visualization of the 3-dimensional volume data is done by the PC, which hosts the powerful DAQ and pre-processing board. The control of the whole online system, the online analysis, i.e. the reconstruction of the current densities of all voxels, the phase resetting analysis and finally the visualisation is done on the PC. A user friendly application software for experiment control, data analysis and visualization has been developed. Filters, phase analysis, spectrum analysis and reconstruction algorithms are implemented. The acquired or processed data and volume sets can be saved to the hard disc for further analysis with standard packages