I am looking to calculate and compare the localization error of different source estimation methods. However, I am facing a challenge in selecting meaningful ground truth data to compare with the estimated source data obtained through each method.
It is crucial to choose appropriate ground truth data to obtain the most plausible localization error for each method.
I would appreciate a detailed response on this matter.
I would suggest looking up previous recent papers on source localization accuracy and see what they did. I am not sure the state of the art offhand! You could start with phantom data for example, but it’s useful to work with simulated data. You could also look into Adding new dataset (ground-truth for source localization) · Issue #7694 · mne-tools/mne-python · GitHub to see if the dataset was ever published (along with a paper).
Thank you very much for your valuable notes.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.