AAL3 scouts with MEG

Discussions
1

Dear Francois,

I would like to ask for your advice.

Me and my team are currently analysing MEG signals. We want to untangle the connectivity patterns between the cerebellum and the cerebrum. For parcellation we use the AAL3. For our source analysis we use dSPM.

We know that MEG is not so accurate for deep brain structures. So, we decided to omit some of the AAL3 scouts like the thalamus and other deep structures. For instance we skipped:
121, 122 Thalamus, Anteroventral Nucleus
123, 124 Lateral posterior
125, 126 Ventral anterior
127, 128 Ventral lateral
129, 130 Ventral posterolateral
131, 132 Intralaminar
133, 134 Reuniens
135, 136 Mediodorsal medial magnocellular
137, 138 Mediodorsal lateral parvocellular
139, 140 Lateral geniculate
141, 142 Medial Geniculate
143, 144 Pulvinar anterior
145, 146 Pulvinar medial
147, 148 Pulvinar lateral
149, 150 Pulvinar inferior
157, 158 Nucleus accumbens
159, 160 Ventral tegmental area
161, 162 Substantia nigra, pars compacta
163, 164 Substantia nigra, pars reticulata
165, 166 Red nucleus
167, 168 ocus coeruleus
169 Raphe nucleus, dorsal
170 Raphe nucleus, median

Related Publication (look at tables 2,3): https://www.sciencedirect.com/science/article/pii/S1053811919307803?fbclid=IwAR3BwE01sfsBXSZZSNgctWhmSymnYron6SwZK0NpmTBMpkQ-vGA8eA75RQc

However, we believe that the connectivity path could start from the cerebellum and pass through the thalamus and other deep structures before it reaches the cerebrum. For this reason we kept some of the deep structures like:
77, 78 Lenticular nucleus, Putamen
79, 80 Lenticular nucleus, Pallidum
81, 82 Thalamus
151, 152 Anterior cingulate cortex, subgenual
153, 154 Anterior cingulate cortex, pregenual
155, 156 Anterior cingulate cortex, supracallosal

What do you think about this approach?

We appreciate your help!

Kind regards,
Konstantinos Tsilimparis
Medical Physics and Digital Innovation Lab
Aristotle University of Thessaloniki

2

MEG is capable of recording some of the activity of these deeper regions, but buried in the signal coming from cortical areas and ambient noise. It is possible to reveal some of it by averaging hundreds or thousands of repetitions of the same stimulus - but studying connectivity is a different topic.
Minimum norm source imaging could theoretically help separating the signals coming from deeper vs. superficial sources. But in practice, I'm personally not confident that this is enough for studying functional connectivity from resting state recordings.

@ecoffey and @Sylvain have experience in recording deeper brain structures with MEG (e.g. https://www.nature.com/articles/ncomms11070). They are currently exploring various options for studying connectivity between cortical and subcortical regions.
They could give you more constructive and encouraging feedback than mine on your experimental protocol :slight_smile:

3

Thank you for your direct answer!

We have only a few repetitions per subject per condition(only 50). Also, we work on an ERP paradigm and not on resting-state recordings.

My question is: if we want to run phase transfer entropy through brainstorm is it okay to skip some of the AAL3 scouts? Will the connectivity analysis results still be valid?

4

Yes, I think this is OK to use any subset of ROIs from the AAL parcellation.
I would recommend however that you keep both the cortical and subcortical ROIs, and make sure that they are not identical. If you have some cortical ROIs that seem to show exactly the same behavior as the subcortical ROIs you are interested in (shape of the scouts time series for the average per condition and/or connectivity patterns), then you might be observing only activity from the cortex projected on the deeper structures by the minimum norm solution.
If you want to make sure that what you observe in deeper scouts is really coming from subcortical regions, then you should be able to observe differences with the cortical scouts, with features that are specific to the subcortical scouts time series.

1 Like
5

Hi @contsili,

Agree with what Francois said. It's okay to use a subset of ROIs (probably wise). Whether or not you can see anything with 50 trials depends on the SNR for your specific signal. It doesn't seem like very many trials to me, but I work mostly with really teeny signals.

Although connectivity measures are fancy and many people are using them with MEG, I am finding them a bit too variable between measures and parameter-dependent for my liking, and that is even before we go to unconstrained / subcortical / deep sources. We are exploring those issues in some known signals before we'll be relying on them too much. I doubt that your expected connectivity pattern is going to jump out of your data.
You may find that a combination of theory/literature driven highly restricted connectivity questions and/or looking at basic things like which ROI signal peaks first might be a stronger and more convincing way to go for now and might still allow you to speculate on connectivity in that system.

Since you're working on cerebellum, this paper might be useful for your general reference: https://www.sciencedirect.com/science/article/pii/S1053811920303049

Good luck,

Emily