Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Last revision Both sides next revision
getting_started:neuromag [2013/10/17 18:13]
137.110.172.50 [Set Path]
getting_started:neuromag [2017/05/10 10:20]
jan-mathijs [Frequently Asked Questions]
Line 1: Line 1:
-{{tag> elekta neuromag dataformat ​MNI headmodel sourceanalysis ​meg}}+{{tag>​elekta neuromag dataformat meg}}
  
-====== Getting started with Elekta-Neuromag data ======+====== Getting started with Elekta/Neuromag data ======
  
 ===== Introduction ===== ===== Introduction =====
Line 57: Line 57:
  
 ===== Special issues ===== ===== Special issues =====
 +
 +  * if you have STI001 up to STI008, the TTL values (single bits) in those channels will be combined into an event of type '​Trigger'​ with an integer value between 0 and 255. 
  
   * Reading .fif mri-data with fieldtrip and making a single shell headmodel (example script can be found [[example/​read_neuromag_mri_and_create_single-subject_grids_in_individual_head_space_that_are_all_aligned_in_mni_space|here]]).   * Reading .fif mri-data with fieldtrip and making a single shell headmodel (example script can be found [[example/​read_neuromag_mri_and_create_single-subject_grids_in_individual_head_space_that_are_all_aligned_in_mni_space|here]]).
  
-Not yet documented or fixed but encountered:​ +  ​The default behavior of ft_read_event is that it assumes that event values ​below 5 are noise. ​However, in the new systems ​(Elekta Neuromag VectorView ​or Triux (306 channels both) this is seldom ​the case. 
-  ​read_trigger,​ this function ​assumes that event_values ​below 5 are noise. ​In the new system ​(Elekta Neuromag VectorView 306) this is not always ​the case. +
  
 +===== Frequently Asked Questions ===== 
 +\\
 +    - Can I do source reconstruction with combined planar and magnetometer channels? ​
 +        *Yes, by specifying cfg.coilaccuracy=1 or 2 during ft_preprocessing
 +    - Can I do source reconstruction with combined MEG and EEG channels? ​
 +        * Almost!
 +    - How can I visualize planar gradient data? 
 +        *After combining, but also by pulling them apart in side-by-side layouts
 +    - How can I do stats with clustering on data from the planar gradiometers? ​
 +        * After combining, or using some smart (still to be defined) neighbourhood definition
 +    - Can I do stats with clustering on combined planar and data? 
 +        *Yes, using two (non-neighbouring) concatenated neighbourhood definitions
 +    - Can I combine multiple runs of an experiment (over multiple files)? ​
 +       *Yes, by using ft_appenddata...but see next question
 +    - How can I deal with rank-deficient maxfiltered data from multiple datasets/​runs? ​
 +        *After combining data from separate runs using ft_appenddata,​ you can run PCA using ft_componentanalysis followed by ft_rejectcomponent such that the rank of your covariance matrix is a number less than 64.
 +    - Should I use or avoid using MaxFilter? ​
 +        *It depends on several factors, including the level of noise in your recording, the presence of artefacts from outside the helmet and large amounts of head movement. Optimising source-localisation for Maxfiltered data is still under development. N.B. If you used Internal Active Shielding (IAS) running Maxfilter is obligatory