====== NeuroElf feature requests ======

===== Motivation =====
Many (if not most) of the features being implemented in NeuroElf have been requested by colleagues. While I enjoy the process of developing algorithms and translating them into Matlab, I usually require that the product of my work is actually used. Hence I would say that if you think a feature you are missing could be of use for several labs/researchers, you are welcome to submit it for inclusion into this list (which I will update whenever a feature becomes requested or available).

//Comment: the fact that most features on the list were requested on May 18th 2010 is because I presented the toolbox to the SCN lab meeting that day...//

===== Procedure =====
In case you have a feature (or file format) request, please follow these steps:

  * check whether (and if so since when) the feature/format you wish to request is already on the list below
  * write an email to **//info (AT) neuroelf (DOT) net//**, stating the request (or asking to prioritize the request)
  * wait for a reply by email and also...
  * check back on this page to see if it's on the list or the status has been updated

===== Feature requests =====
I have decided to create separate lists for
  * features of use which seem either relatively easy to implement (short time) or of high interest (important additions for the toolbox)
  * features that seem useful but might take very long to implement or where the use is limited to a very specific application
  * features which I consider not suitable or of too little use for the toolbox (in which case I would also give a short comment as to why I believe so)
  * features that were requested and which have been implemented

==== To be implemented ====
^ Feature name              ^ Requested by       ^ Request date  ^ Description  ^ Status  ^
| HDR menu -> segmentation  | Jochen Weber       | 4/12/2011     | adding segmentation (SPM-based) to the HDR menu  | in pipeline  |
| ComputeFormula also for SMP  | Jochen Weber    | 4/12/2011     | add support for SMP::ComputeFormula (either different or combined implementation  | in pipeline  |
| physio confounds          | Lauren Atlas       | 4/18/2011     | support to automatically create confound regressors from physio (ECG) data, still needs specs!  | in pipeline  |
| better subject-specific VOI support  | Hakwan Lau  | 4/18/2011  | support for subject specific VOIs in ne_mdmvoicondavg -> MDM::VOITimeCourses, and throughout  | in pipeline  |
| VOI time course plot      | Jochen Weber       | 4/20/2011     | option to plot VOI time courses (probably from within ne_mdmvoicondavg dialog), incl. PRT information  | in pipeline  |
| multi-study (MDM) dialog enhancements  | SCAN Unit  | 4/20/2011  | add RunTimeVars.Discard inspection/update feature into dialog; also feature to see condition names from models, options for single-trial GLM and PPI  | in pipeline  |
| fmriqasheet enhancements/fix  | SCAN Unit      | 4/20/2011     | 1. percentage of outlier volumes is incorrect; 2. UI should get a dialog to assess multiple scans, give results for multiple scans, have options (e.g. realignment); 3. link fmriqasheet output to object.RunTimeVars.Discard field  | in pipeline  |
| SDM::CalcBetas            | SCAN Unit          | 4/20/2011     | make it optional to first project out confounds (filters, motion) of both data and remaining model and then perform regression (thorough testing after implementation!!)  | in pipeline  |
| SDM::CalcBetas check/fix  | Jochen Weber       | 5/20/2010     | after code inspection, I realized that some features only exist in the help (masking/thresholding/global signal as nuisance); adding that code!!  | in pipeline  |
| local storage of configs  | Nick Ketz          | 5/18/2010     | in a network-based software setup (central installation of NeuroElf), the configuration files are currently shared (as they live in the installation folder); either make the filenames local (e.g. based on username) or store locally (in user folder)  | in pipeline  |
| spm5_prepro enhancement   | Jen Silvers        | 6/8/2010      | option to set a filename pattern for the VTCs that are created after preprocessing  | in pipeline  |
| spm5_prepro improvements  | Hedy Kober et al.  | 5/18/2010     | more flexibility w.r.t. subject normalization (support for EPI template, DARTEL toolbox), text field for number of skipped volumes, setting origins/orientation of source data (fewer problems during normalization/coregistration), option to purge intermediate preprocessing steps (less storage on disk required)  | in pipeline  |
| importvmp dialog enhancement  | Hedy Kober     | 5/18/2010     | allowing the user to specify Analyze/NIftI files containing maps from several folders  | in pipeline  |
| normalization check       | Jochen Weber       | 5/18/2010     | automatic generation (update) of a multi-subject/multi-session VTC (or other format) to quickly verify (via eyeballing) that the normalization has worked alright  | in pipeline  |
| weighted data             | Jochen Weber       | 4/12/2011     | start with data acquisition, weights = 1, then, during each processing/analysis step, carry over weights as estimate of plausibility of data; use weights for all subsequent steps  | not sure yet  |
| VTC/VOI-network analysis  | Jochen Weber       | 4/12/2011     | different options to assess suspected network nodes of VOIs (correlation matrices, differences in correlations across conditions, Granger, SEM?, robust option, etc.)  | not sure yet  |


==== Potential candidates ====
^ Feature name              ^ Requested by       ^ Request date  ^ Description  ^ Comment  ^
| Granger causality         | Jochen Weber et al  | 5/18/2010    | computing of Granger causality, both in the BrainVoyager pair-wise style, but also in a network-like way (comparable to the toolbox of Anil Seth in Matlab)  | needs a lot of research and coding effort, decision pending  |
| improvement of PPI        | Ajay Satpute at al  | 5/18/2010    | the new PPI-like feature (MDM::ComputeGLM) should be improved (to allow the SPM-typical deconvolution and reconvolution)  | will require quite some additional coding  |

==== Unlikely candidates ====
^ Feature name              ^ Requested by       ^ Request date  ^ Description  ^ Reason for exclusion from list of likely candidates  ^

==== Already implemented ====
^ Feature name              ^ Requested by       ^ Request date  ^ Description  ^ Implemented in version  ^
| GUI: define ROI feature   | Jochen Weber       | 5/18/2010     | the actual code to create a VOI from the marked voxels was missing  | v0.9b  |
| GUI: local maxima checkbox  | Kevin Ochsner    | 5/18/2010     | prior to the implementation, users needed to press the "L" key to disable/enable local maxima splitting of clusters (and the state of that flag was invisible)  | v0.9b  |
| GUI: goto nearest maximum  | Kevin Ochsner     | 5/18/2010     | the GUI should allow users to set the current position of the cursor to the closest voxel that was determined as a maximum of an earlier cluster table call  | v0.9b  |
| GUI: flexible contrast weights  | Ajay Satpute  | 5/18/2010    | allowing users to edit contrast weights by entering a number instead of the less flexible +1/-1/0 and subsequent balancing logic  | v0.9b  |
| ClearObjects (xff method)  | Nick Ketz         | 5/18/2010     | giving users an option to clear several objects at once that match a common filename pattern  | v0.9b  |
| any2ascii overload        | Jochen Weber       | 5/18/2010     | overloading of other classes (xfigure, xff) to allow passing of objects in ASCII format (e.g. for GUI elements)  | v0.9b  |
| tff optional as binary    | Jochen Weber       | 5/30/2010     | for large text-based files (FBR), there is a high speed-up potential if the file is read and parsed as binary content -- implemented for testing!  | v0.9b  |
| masks from VMPs in GUI    | Ethan Kross        | 5/29/2010     | giving the user a simple option to convert the current statistical selection (height and size threshold of currently displayed map/s) into a BrainVoyager compatible MSK object  | v0.9b  |
| echo mode in GUI          | Nick Ketz          | 5/18/2010     | GUI should give the names of functions/methods called (to allow users to script their actions in a simple way)  | v0.9b  |
| GSR data processing       | Rich Lopez         | 5/19/2010     | processing of GSR data with automatic filtering and signal detection  | v0.9b SNAPSHOT  |
| flexinterpn multiple kernels  | Jochen Weber   | 6/19/2010     | option to pass multiple kernels (one per dimension) to flexinterpn, which then allows smoothing in one dimension only  | v0.9b SNAPSHOT  |
| plotcurves improvements  | Jochen Weber    | 6/19/2010     | adding type cue; making spots "fixed" (certain types, such as onset/cue); adding key-commands for "find max in window", "find min in window"  | v0.9b SNAPSHOT  |
| GUI-based origin setting  | Jochen Weber       | 5/19/2010     | enhanced way to specify the origin and orientation of a dataset by allowing the user to drag (and shift+drag) the dataset for translation (and rotation) //feature implemented differently// | v0.9b SNAPSHOT  |
| showspmrparams enhancement  | Hedy Kober       | 6/28/2010     | legend next to plots and keeping the colors the same as SPM  | v0.9b SNAPSHOT  |
| mediation                 | Jochen Weber et al  | 5/18/2010    | support (or integration) of Tor Wager's mediation toolbox so as to have it work with all supported file formats //implemented for the RFX-GLM format, currently only single-level mediation// | v0.9b SNAPSHOT  |
| add GUI code for MDM::ComputeGLM  | Hedy Kober  | 5/18/2010    | add the code for the GUI dialog to locate VTC/SDM/PRT/RP files for MDM::ComputeGLM  | v0.9b SNAPSHOT  |
| GLM::VOITimeCourses  | Jochen Weber  | 6/30/2010  | ability to extract timecourses from referenced VTCs in a GLM  | v0.9b SNAPSHOT  |
| combining VMPs (in GUI)   | Hedy Kober         | 5/18/2010     | dialog that allows to flexibly combine maps from several VMP containers  | v0.9b SNAPSHOT  |
| importrfxglm bug fix      | Jochen Weber       | 5/18/2010     | currently the condition selection in importrfxglmfromspms does not allow to specify the conditions to be imported; while they can be easily removed after the import is complete, it would of course be nicer for that feature to work!  | v0.9c  |
| CreateSMP from VMP to AFT  | Jochen Weber      | 4/12/2011     | change the code so that it works for all map type objects (incl. CMP, GLM, HEAD, HDR, VMP)  | v0.9c  |
| NII export support for GLM/VMP formats  | Hakwan Lau  | 4/18/2011  | so far, the output method is limited to HDR/IMG, also support NII (checking that exported files contain valid transformation matrix!)  | v0.9c  |
| mutliple scatters         | Jen Silvers        | 5/18/2010     | ability to plot several, independent scatter plots in the GUI (e.g. for multiple groups of subjects or to compare conditions; interface still requires specification! probably define separate scatter types/colors per group with default being using the same; if split perform regression per group  | v0.9c  |
| legend on bar plot        | SCAN Unit          | 4/20/2011     | add optional legend to bar plot  | v0.9c SNAPSHOT  |
| full support for 3D/4D HDR/NII MAT files  | Jochen Weber  | 4/12/2011  | support for SPM's 4D (and older 3D) MAT files with (different) quaternion matrices for each volume (sampling, updating, etc.)  | v0.9c SNAPSHOT  |