The problem with topographic brain maps in microstate analysis

💬 Support & Discussions
1

Dear colleagues
could you please tell me why we can’t get classic topographic maps in microstate analysis (using the Pycrostates)? I read the forum and saw that data quality is important, but we checked it very carefully. I use 19 channels, I read that this is enough.I did a similar analysis with 63 channels and everything turned out fine. Please tell me what could be the potential problem.

Could this be due to the peculiarities of the package and the small number of channels, or is it more likely to be a peculiarity of the sample? We have a clinical sample.

PIC
PIC1286×705 98.4 KB

2

Hey @Kate,

From what I can see, the topographies look ok.

Microstates analysis being a data-driven method, having topographies different from the literature is not a bad result in itself.

Here are a few suggestions for reassuring yourself about the results:

  • check the topographies for k=5 states to see if you get something closer to the maps reported in the literature.
  • redo the analysis using spatial filter to see if results are consistent.

Hope it helps

3

Victor, thanks so much for your help! I’ll try that.