Abstract
Brain source imaging and time frequency mapping (TFM) are commonly used in magneto/electro encephalography (M/EEG) imaging. However, these methods suffer from important limitations. Source imaging is based on an ill-posed inverse problem leading to instability of source localization solutions, has a limited capacity to localize high frequency oscillations and loses its robustness for induced responses (ill-defined trigger). The drawback of TFM is that it involves independent analysis of signals from a number of frequency bands, and from co-localized sensors. In the present article, a regression-based multi-sensor space–time–frequency analysis (MSA) approach, which integrates co-localized sensors and/or multi-frequency information, is proposed. To estimate task-specific brain activations, MSA uses cross-validated, shifted, multiple Pearson correlation, calculated from the time–frequency transformed brain signal and the binary signal of stimuli. The results are projected from the sensor space onto the cortical surface. To assess MSA performance, the proposed method was compared to the weighted minimum norm estimate (wMNE) source imaging method, in terms of spatial selectivity and robustness against an ill-defined trigger. Magnetoencephalography (MEG) recordings were performed in fourteen subjects during two motor tasks: finger tapping and elbow flexion/extension. In particular, our results show that the MSA approach provides good localization performance when compared to wMNE and statistically significant improvement of robustness against ill-defined trigger.
Highlights
Magnetoencephalography (MEG) is routinely used in non-invasive dynamic functional brain imaging
We propose a new methodological approach to overcome the following drawbacks of widely applied source imaging and time frequency mapping (TFM): (a) sensitivity to ill-defined triggers; (b) limited frequency resolution, and limited capability to localize high frequency oscillations; (c) potential instability caused by ill-posed inverse problem of source imaging, and (d) difficulties in processing multi-sensor/multi-frequency data by TFM
The cortical maps of the absolute activity values were obtained for each participant and on average across participants using the multi-sensor space–time–frequency analysis (MSA) and weighted minimum norm estimate (wMNE) methods for the “FingerTap” experiment (Figure 3, left) and the “FlexElbow” experiment (Figure 4, left)
Summary
Magnetoencephalography (MEG) is routinely used in non-invasive dynamic functional brain imaging. Functional brain imaging is commonly performed using source imaging methods. The basic principle of source imaging involves the use of the dipolar biophysical model to reconstruct brain sources from scalp measurements, assuming that the MEG signal can be expressed as a linear mixture of brain sources in the presence of noise [1]. The time course of each dipole current is estimated from the MEG signal. Brain source imaging uses a high number of dipoles The number of dipoles generally exceeds the number of sensors (up to 306). This results in an ill-posed inverse problem that requires a Sensors 2020, 20, 2706; doi:10.3390/s20092706 www.mdpi.com/journal/sensors
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
