Back to the main page.
Bug 619 - easycapXX-avg layouts have to be improved
| Status | CLOSED FIXED |
| Reported | 2011-04-27 21:40:00 +0200 |
| Modified | 2011-11-09 14:37:15 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P1 normal |
| Assigned to: | Jörn M. Horschig |
| URL: | http://fieldtrip.fcdonders.nl/tutorial/layout#easycap_layouts |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2011-04-27 21:40:52 +0200
The current ones are based on polhemus recordings which were made for a limited number of subjects, averaged and then projected on a 2D plane. Much better layouts can be made by taking the bitmaps from the easycap site, see http://www.easycap.de/easycap/e/contents.htm and click on electrode arrangements. The easycap naming scheme should be used (e.g. "M10" for the cap that we are mostly using at the DCCN), see below: Recordable Area Demonstrated by 10%-System (M1) American versus German Electrode Nomenclature 10%-Arrangements M25: International 10/20-System (used in EC20) M3: Extended 10/20-System with 30 Channels M22: Small Equidistant 29-Channel-Arrangment (used in Braincap32) M23: Large Equidistant 32-Channel-Arrangement M24: Large Equidistant 34-Channel-Arrangement (used in EC40) M11: 61-Channel-Arrangement ("10%-System") (used in BrainCap64) M1: 74-Channel-Arrangement (used in EC80) M15: 128-Channel-Arrangement Triangulated Equidistant Arrangements M7: Spherical 32-Channel-Arrangement M10: Equidistant 61-Channel-Arrangement M16: Equidistant 88-Channel-Arrangement M14: Spherical 124-Channel-Arrangement Miscellaneous Arrangements M20: BESA 32-Channel-Arrangement for Epilepsy Diagnostics M17: 29-Channel-Arrangement for Language Research M19: 2x64-Channel-Kombi-Montage
Jörn M. Horschig - 2011-06-01 10:55:13 +0200
How to label electrodes that have no label on the easycap webpage? example: http://www.easycap.de/easycap/_img/content/ea_08_M24.gif As far as I can see the problem is that some electrodes are freely positional.
Jörn M. Horschig - 2011-07-27 13:03:23 +0200
Created attachment 108 EASYCAP m1
Jörn M. Horschig - 2011-08-31 10:20:45 +0200
Created attachment 128 Template GIF vs. FT layout While working on this, I realized that there is a mismatch between the FT topo and the one on the homepage. E.g. channel 6 is not on the midline, channel 73 is not on the outer ring of the head, etc... I suppose this is FT related, because there is just no way I can adjust the exact position of the channels. The layout is constructed by taking the relative distance of the sensors to each other.
Jörn M. Horschig - 2011-08-31 11:21:16 +0200
I finished layouts for M1, M10, M11, M14, M15 and M16 M7 is a variation of M14, meaning that I just have to rename or delete some channels M1 can be the model template for M3, M17, M19, M20, M22, M23, M24 and M25, see above what this means. However, the only question left is, how should I label channels that have an empty label on the webpage?
Robert Oostenveld - 2011-08-31 17:33:57 +0200
the empty circles are placeholders, they are not included in the actual cap, right? Could you check the electrode count? If needs be, I can provide you with the names (there is a consistent naming scheme that also includes those).
Robert Oostenveld - 2011-08-31 17:39:53 +0200
Please note that ft_layoutplot has the option cfg.image where you can specify the bitmap. Please use that to ensure the correct placement. Should we update/improve the tutorial at http://fieldtrip.fcdonders.nl/tutorial/layout?
Robert Oostenveld - 2011-08-31 17:42:39 +0200
(In reply to comment #3) > While working on this, I realized that there is a mismatch between the FT topo > and the one on the homepage. E.g. channel 6 is not on the midline, channel 73 > is not on the outer ring of the head, etc... I suppose this is FT related, This is because of the scaling of the channel positions to fit exactly in the circle, i.e. the initial problem reported by Johanna that caused us to trigger these efforts.
Jörn M. Horschig - 2011-09-01 12:12:34 +0200
Placeholders: I am not sure why they are there... but e.g M19 has unlabeled electrodes, and M20 has the same amount of electrodes and the labeled ones from M19 coincide with the ones in M20. It's just that the formerly unlabeled got a place and name in M20, now. Should I just exclude the placeholder electrodes, or exclude the whole layout? ft_layoutplot: Yesterday in the meeting, I heard about about that functionality and the tutorial the first time... I wrote the function for creating the layouts myself some weeks ago. The problem is not that the tutorial is bad (it is quite good actually), but when you search for 'layout', it does not show up in the search results. I don't quite get why, because the tutorial is tagged with 'layout' correctly. Scaling: Does this mean I should also switch to creating .mat files instead of .lay (see bug 890)?
Robert Oostenveld - 2011-09-01 13:52:40 +0200
(In reply to comment #8) > Placeholders: > I am not sure why they are there... but e.g M19 has unlabeled electrodes, and > M20 has the same amount of electrodes and the labeled ones from M19 coincide > with the ones in M20. It's just that the formerly unlabeled got a place and > name in M20, now. Please look up the Oostenveld &Praamstra 2001 paper on electrode nomenclature for extended 10-20 systems. Many of the unlabeled positions in the bitmaps can be labeled based on either the 10-20, the 10-10 or the 10-5 layout. The M19 cap at http://www.easycap.de/easycap/e/electrodes/18_M19.htm suggests that it is a 2x64 channel system. So the white circles are no placeholders here, but actual channels. However, the particular layout is "The graphic shows as an example an arrangement by ... Jena" not standard. So M19 should not be done. M25 for example mentions at http://www.easycap.de/easycap/e/electrodes/04_M25.htm that there are 23 electrode positions. The white circles therefore are placeholders but not actually in the cap. > Should I just exclude the placeholder electrodes, or exclude the whole layout? just exclude all the fuzzy layouts >... > some weeks ago. The problem is not that the tutorial is bad (it is quite good > actually), but when you search for 'layout', it does not show up in the search > results. I don't quite get why, because the tutorial is tagged with 'layout' > correctly. should be discussed with Eelke. > Scaling: > Does this mean I should also switch to creating .mat files instead of .lay (see > bug 890)? yes, please synchronize your actions with Arjen, who is now coordinating this effort. I am afraid that you missed out on a particular FT meeting in which bug #890 was cerated/discussed.