Back to the main page.
Bug 1580 - FAQ: What kind of volume conduction models are implemented?
| Status | CLOSED FIXED |
| Reported | 2012-07-02 15:04:00 +0200 |
| Modified | 2018-03-10 22:01:05 +0100 |
| Product: | FieldTrip |
| Component: | documentation |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P3 normal |
| Assigned to: | Lilla Magyari |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Lilla Magyari - 2012-07-02 15:04:09 +0200
This is not entirely informative: http://fieldtrip.fcdonders.nl/faq/what_kind_of_volume_conduction_models_are_implemented because it says: "all forward models supported by the Neuromag meg-calc toolbox" but it doesn't mention what kind of forward models are implemented in the Neuromag meg-calc toolbox. I would rather say here this: The implemented methods are listed in the reference of ft_prepare_headmodel (with link to ref page). If you want to know more about each model, consult the Reference to implemented methods page (with link to the page). Lilla
Robert Oostenveld - 2012-07-02 15:37:34 +0200
This should refer to the http://fieldtrip.fcdonders.nl/reference/ft_prepare_headmodel page
Lilla Magyari - 2012-07-02 19:05:57 +0200
(In reply to comment #1) I have looked at the reference of ft_prepare_headmodel, and I have found the following inconsistencies: (1) Methods listed in the reference of ft_prepare_headmodel are not consistent with method names in the script of ft_prepare_headmodel (2) it is not easy to match the method names in the reference of ft_prepare_headmodel to the articles in References to implemented methods. (3) Not all methods have articles in references to implemented methods. (4) But there is also no explanations in the reference of ft_prepare_headmodel what the methods are. Some references and short explanations of the methods are in the reference of the lower-level functions (e.g. help ft_prepare_singleshell). But those can not be seen by the users because the individual functions (e.g. ft_headmodel_bemcp) are called from ft_prepare_headmodel and the help of ft_prepare_headmodel does not refer to the help of these lower level functions. (5) But it would be not easy to tell in the help that one can check out also the lower-level functions because the naming of the lower-level functions are not consistent (e.g. ft_headmodel_bemcp vs. ft_headmodel_bem_dipoli). (6) Some of the “methods” of ft_prepare_headmodel are not computations but reading functions of already made head models. (7) Some of the "methods" that I see in the script of the ft_prepare_headmodel are not documented in its help (e.g. ft_headmodel_fem_simbio). ALL of these problems are worked out in detail for each "method" here: http://fieldtrip.fcdonders.nl/development/documentation_source?&#what_kind_of_volume_conduction_models_of_the_head_are_implemented_in_fieldtrip My suggestions are: - help should be corrected with the right method names (that are in the script) - reference to implemented methods should be extended, and each article should be introduced with the method name of the reference documentation of ft_prepare_headmodel - undocumented functions should be documented - in the reference documentation of the ft_prepare_headmodel, each method should refer to its lower-level function if someone wants more information on the method (or?) Lilla
Robert Oostenveld - 2012-07-02 22:02:15 +0200
I have renamed the forward/ft_headmodel_xxx.m functions so that they don't have BEM_XXX but only XXX as method mbp> svn commit Deleting forward/ft_headmodel_bem_asa.m Adding forward/ft_headmodel_asa.m Deleting forward/ft_headmodel_bem_dipoli.m Adding forward/ft_headmodel_dipoli.m Deleting forward/ft_headmodel_bem_openmeeg.m Adding forward/ft_headmodel_openmeeg.m Deleting forward/ft_headmodel_fdm_fns.m Adding forward/ft_headmodel_fns.m Deleting forward/ft_headmodel_fem_simbio.m Adding forward/ft_headmodel_simbio.m I have also cleaned up the documentation in fieldtrip/utilities/ft_datatype_headmodel.m mbp> svn commit ft_datatype_headmodel.m ../forward/private/ft_datatype_headmodel.m Adding forward/private/ft_datatype_headmodel.m Sending utilities/ft_datatype_headmodel.m Transmitting file data .. Committed revision 6194. PS please remind me that http://fieldtrip.fcdonders.nl/example/testing_bem_created_leadfields should again be tested
Robert Oostenveld - 2012-07-02 22:22:15 +0200
I have updated all "see also" references, made the lists of headmodels consistent and attempted to use consistent names in the documentation mbp> svn commit Sending forward/ft_prepare_vol_sens.m Sending forward/ft_voltype.m Sending ft_dipolefitting.m Sending ft_prepare_headmodel.m Sending ft_prepare_leadfield.m Sending ft_prepare_sourcemodel.m Sending ft_sourceanalysis.m Sending utilities/ft_datatype_headmodel.m Transmitting file data ........ Committed revision 6195.
Robert Oostenveld - 2012-07-02 22:23:14 +0200
@Lilla, could you do an "svn update" and revisit the documentation?
Lilla Magyari - 2012-07-03 16:33:43 +0200
(In reply to comment #5) Hi Robert, I have checked the references of the headmodel functions. It looks great. But I have found still some inconsistencies, and I have some questions: (1) In the reference of ft_prepare_headmodel a method called "multisphere" is listed but the method in the script called by "localspheres" and the lower level function is called ft_headmodel_localspheres. I thought I could just rename the method as "localspheres" in the reference, but it is not that simple because vol.type="multisphere" is in the ft_headmodel_localspheres. (2) The reference doc (help) of ft_headmodel_simbio and ft_headmodel_fns contains still ft_headmodel_fem_simbio and ft_headmodel_fdm_fns. (3) there is method called "slab_monopole" in the script and a lower level function called ft_headmodel_slab belonging to it which is not listed in the reference of ft_prepare_headmodel. (4) I have a question: why ft_prepare_headmodel contains "methods" as asa and simbio? If I understant well these are reading in and transforming already made vols to FT style. Shouldn't be ft_read_vol used for these? (Actually, ft_headmodel_asa contains only one line: ft_read_vol). Or if these are implemented in this way, why aren't other type of vol-s read in with ft_prepare_headmodel as well (which can be read in only with ft_read_vol at the moment, e.g. neuromag)? What is the difference between those? (5) Isn't openmeeg a BEM method that is also developed for MEG? Isn't that also implemented in FT? (it is listed now only under EEG methods.) Can I correct the changes regarding point 2? I think that involves only the documentation. Lilla
Robert Oostenveld - 2012-07-03 16:38:41 +0200
(In reply to comment #6) re 1, good point, indeed needs more work to get consistent. re 2, has been fixed, it just takes some time for the ref doc to end up on the wiki re 3, this is a crimic specific model and I rather leave it undocumented as it probably got some issues with the algorithm. re 4, I'll explain in person > Can I correct the changes regarding point 2? > I think that involves only the documentation. No, you cannot change this (easily). The reference doc is automatically generated by fieldtrip/ft_documentation_reference function and then copied to the wiki. That is why the reference documentation pages are read-only on the wiki. So to make changes to th ereference documentation you actually have to change the fieldtrip code (and wait for 12 hours or so). I'll drop by.
Robert Oostenveld - 2012-07-03 22:01:29 +0200
related to testing, I just noticed on http://fieldtrip.fcdonders.nl/development/dashboard that the test_headmodel_xxx files do exist. They are located in fieldtrip/forward/test, rather than fieldtrip/test.
Lilla Magyari - 2012-07-04 14:17:56 +0200
(In reply to comment #8) In reply to your last comment: should be the test_headmodel_XXX moved under fieldtrip/test? And has been this fixed? (1) In the reference of ft_prepare_headmodel a method called "multisphere" is listed but the method in the script called by "localspheres" and the lower level function is called ft_headmodel_localspheres. I thought I could just rename the method as "localspheres" in the reference, but it is not that simple because vol.type="multisphere" is in the ft_headmodel_localspheres. If yes, I'd like to close this bug. Lilla
Robert Oostenveld - 2012-07-04 18:52:04 +0200
(In reply to comment #9) The test scripts should remain where they are. I think the multisphere -> localspheres rename is now completely done.
Lilla Magyari - 2013-07-23 22:28:08 +0200
I close this bug. Lilla