Could MEG aid in mirror neuron research better in terms of validity, reproducibility and yielding stronger evidence for the existance and localization of human mirror neurons/networks? There is a substantial amount of fMRI studies but only a few using MEG (pubmed results ratio 220 to 15). Is there any reason for preferring fMRI (or TMS) to MEG?
The combination of MEG and MRI shows some promise, but since both machines are prohibitively expensive, has not yet caught on in general use.
EEG is not prohibitively expensive and is often used in combination with fMRI. However both scalp-EEG and MEG currently have poor spatial resolution, so they are not replacements for single-unit (i.e. looking at individual neurons) work in electrophysiology if the question of interest involves particular neurons.
For human mirror "networks", both EEG and MEG can be useful when asking questions related to the time-scale or frequency of the network, while fMRI could be useful when asking questions related to the location or connectivity of the network.
I would say definitely not! Have a look at Coll et al freely available here https://sites.google.com/site/geoffbirdlab/papers
MEG permits a way better localisation of cortical sources then EEG without losing the temporal resolution. But the results are doubtful when it comes to subcortical sources. Most likely it is rarely used as (1) it is very expensive (2) the point of departure of MEG was localisation in epilepsy. Nevertheless, it remains the non-invasive tool which comes closest to direct electric corticography. There are a lot of problems when dealing with distributed sources or overlapping sources which remain to be resolved.
The theoretical and simulation work that I have seen do not support the idea that MEGs provide better localization. This is partly because the MEG sensors are far away from the scalp. If magnetometers could be placed closer to the scalp then the 2D resolution would be better than EEG because MEG would ignore volume conduction effects and would not suffer as much from field spread effects (however localization to a 3D surface in both modalities would still require extra assumptions).
MEG and EEG do differ however in what they measure. Due to the typical tangential direction of current flow in the cortex, MEG sensors are better at finding activity that lives in sulcal walls (e.g. some auditory responses) while EEG is better at finding activity that lives in gyral crowns.
I refer to Electric Fields of the Brain (Nunez and Srinivasan, 2006) and companion papers:
http://www.sciencedirect.com/science/article/pii/S016502700700307X
- Badges
- Science topic