Magnetoencephalography: from sensor data to source distribution using traditional and advanced methods for solving the inverse problem

Magnetoencephalography (MEG) is a method of functional brain imaging that allows us to noninvasively and contactlessly visualize fast-flowing electrical processes in the brain with a spatial resolution of about one centimeter. This fantastic property of MEG is caused both by the nature of the signals recorded with the help of MEG, and by the result of hardware and algorithmic advances implemented in modern systems and specialized software for processing multichannel MEG measurements. To convert the measurements of MEG sensors into activation maps of the cerebral cortex, it is necessary to solve an underdetermined inverse problem. Ways to combat such underdeterminacy are actively being investigated in the modern scientific community and make a critical contribution to the resulting spatial resolution of the entire MEG technology.
The presentation will first present the engineering mathematical apparatus for solving the inverse MEG (and EEG) problem. Then a family of methods for solving inverse problems developed in our laboratory is described, which provide increased spatial resolution and allow visualizing functional connections between cortical sources. In conclusion, a brief description of the software developed by us for functional visualization of cortical activity based on MEG and EEG in real time on a conventional PC will be given!