Functional Tomography of Complex Systems Using Spectral Analysis of Multichannel Measurement Data

A new method has been proposed for determining the structure of complex biological and physical systems from their electromagnetic fields. The method is based on spectral analysis of multichannel time series. Optimization of the Fourier transform is achieved by integrating long-term time series. Fine tuning to a given frequency is also possible to increase the signal-to-noise ratio. When analyzing a detailed multichannel spectrum, the signal is reconstructed at each frequency and the inverse problem is solved for the resulting field map. Using the model of one elementary source allows one to correctly solve the inverse problem by exhaustive search. The set of found elementary sources for all frequencies represents the functional structure of the complex system being studied. The method was verified on computer and physical models, after which it was successfully applied in various biological problems. The separation of the encephalogram into a signal from the brain and physiological noise was obtained.