DEVELOPMENT OF LOW FREQUENCY MEG HARDWARE AND SOFTWARE

点击翻译按钮获取中文摘要

The overall objective of the proposed research is to develop hardware, software, and techniques to substantially expand the utility of Magnetoencephalography (MEG), both as a clinical diagnostic tool and as a modality for basic studies in the neurosciences. MEG has been utilized extensively to date for the localization of neural generators in pathological conditions (e.g. focal epilepsy), as well as normal sources (e.g. auditory, visual, or somatosensory evoked responses). Such localizations always involve significant simplifications such as approximating sources as single current dipoles and heads as homogeneous spheres. One aim of this proposal is to develop new data analysis techniques and software to exploit the unique ability of MEG to extract more significant kinds of information from magnetic fields of physiologic origin. Among these will be methods to distinguish and localize weak generators which fire almost simultaneously with stronger ones, and methods to monitor subtle spectral changes in MEG signals accompanying changes in alertness or arousal or other changes in spontaneous cortical activity. A second capability of MEG is the completely non-invasive measurement of very slow (DC) shifts which are known to arise in a variety of pathological conditions. Such shifts cannot readily be studied using surface electroencephalography (EEG) because of impedance changes which inevitably occur at the electrode-skin interface. Unfortunately, MEG measurements are also contaminated by environmental magnetic noise (even in elaborately shielded rooms) and magnetic fields arising from regions of the patient's body not under study ("patient noise"). Such noise has been of little consequence in localization studies in which low frequency noise is filtered out and in which additional noise reduction is achieved by signal averaging over a large number of repetitive signals. This is not the case however, for the observation of spontaneous low frequency activity. To date, noise cancellation techniques provided by neuromagnetometer manufacturers have proven inadequate to deal with such problems, in some cases themselves introducing artifacts into the data. Preliminary measurements in our laboratory point the way towards a variety of solutions to these noise problems. The second aim of this proposal is to develop technology for reliable DC-MEG measurements, and to extend the capability of MEG to the measurement of absolute values of the biomagnetic field. For both aims, our techniques will be tested in phantoms, animal models and human patients and controls, as appropriate.