Calibration of a multichannel MEG system based on the signal space separation method

For an efficient use of multichannel MEG systems, an accurate sensor calibration is extremely important. This includes the knowledge of both channel sensitivities and channel arrangement, which can deviate from original system plans, e.g., because of thermal stresses. In this paper, we propose a new solution to the calibration of a multichannel MEG sensor array based on the signal space separation (SSS) method. It has been shown that an inaccurate knowledge of sensor calibration limits the performances of the SSS method, resulting in a mismatch between the measured neuromagnetic field and its SSS reconstruction. Given a set of known magnetic sources, we show that an objective function, which strongly depends on sensor geometry, can be derived from the principal angle between the measured vector signal and the SSS basis. Hence, the MEG sensor array calibration is carried out by minimizing the objective function through a standard large-scale optimization technique. Details on the magnetic sources and calibration process are presented here. Finally, an application to the calibration of the 153-channel whole-head MEG system installed at the University of Chieti is discussed.