Back to the main page.
Bug 2882 - Write a tutorial for analysing NIRS data (or NIRS/EEG data)
| Status | CLOSED FIXED |
| Reported | 2015-04-24 13:44:00 +0200 |
| Modified | 2019-08-10 12:37:29 +0200 |
| Product: | FieldTrip |
| Component: | external |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P5 normal |
| Assigned to: | Jörn M. Horschig |
| URL: | |
| Tags: | |
| Depends on: | 2881 |
| Blocks: | |
| See also: |
Jörn M. Horschig - 2015-04-24 13:44:01 +0200
probably requires input/data from Biophysics (Marc van Wanrooij et al.)
Robert Oostenveld - 2017-09-21 09:22:37 +0200
we might want to start with an example script on the wiki, that has fewer requirements in terms of formatting. Are there publicly shared NIRS datasets?
Jörn M. Horschig - 2017-10-11 13:18:44 +0200
We can use our data - we have several example datasets. I would propose to use two of them. Both are measured on the motor cortex. One is a two-channel recording, the other is a 24-channel recording. Basic analysis would be pretty simple (bandpass-filtering, average, voila). The two-channel recordings has a shallow-channel, allowing for correction of superficial skin activation, i.e. a somewhat more advanced option. The 24-channel measurement can then be taken as an example for a multichannel recording and plotting, maybe even for stats. I propose the following: 1. intro to NIRS and paradigm 2. single channel analysis on the deep channel, i.e. basic processing pipeline 3. short-channel subtraction 4. 24-channel and topographic plotting in a separate tutorial we could cover advanced short-channel correction i.e. with several channels and different techniques and stats and those things. What do you guys think?
Jörn M. Horschig - 2017-10-11 13:20:47 +0200
Created attachment 849 Our internal, incomplete FieldTrip tutorial I started with a tutorial for courses we gave already last year. This is the last version (Aug 2016) - not complete, but can serve as a start/
Jörn M. Horschig - 2017-10-11 13:28:12 +0200
btw, we can also keep in the same order as in our example tutorial: beginning with the 24 channel to do basic analyses, then do some fancy stuff with the 2-channel measurement.
Robert Oostenveld - 2017-10-13 12:02:14 +0200
Hi Jorn, Great to see that there is already something that close to a NIRS tutorial. For didactical reasons I think it is better to start with the 2-channel dataset. Plotting/exploring the data is easy using standard MATLAB, which keeps the complexity low. The 24-channel requires another way of exploring/visualizing (i.e. layout and multiplot). It might also require a more elaborate explanation on how opcodes (transmitters/receivers) relate to channels. Both are relevant and valuable, but I think that the gradual build up of complexity is better served by starting with the 2-channel one. And I am not sure whether there should be "fancy stuff" for the 2-channel dataset, or whether that should be left for the user (i.e. as an in-line exercise).
Robert Oostenveld - 2018-10-31 20:37:24 +0100
there are now two NIRS tutorials.