What's next
A roadmap to the future developments of Brainstorm.
Current topics
Documentation
- Merging the 12+3 introduction tutorials to illustrate better the latest developments
Source modeling
- Implementation of a new unified minimum norm/beamformer framework
Functional connectivity
- Significance thresholding of the connectivity matrices
Computation
- Removing the dependence to the Java interface to run in headless mode
- Interacting directly with distributed computing systems
Compatibility
Reading and writing FieldTrip structures
Calling FieldTrip functions from the pipeline
Recordings
Nicer 2D topographies, standardized (using FieldTrip .lay files?)
- 2DLayout (recordings + TF): Use the same standard positions, too much space between sensors
- RAW file viewer:
- Pre-load next page of recordings
- Documentation: Add definition of bad segments
2DLayout: Doesn't work when changing page => need refresh of GlobalData.Preferences.TopoLayoutOptions.TimeWindow
- EEG reference/storage:
- Intracranial electrodes: Define in the MRI viewer
Bad channels that can be specified at the program level (for sites that have permanently bad channels) => AS Dubarry
- RAW processing:
- Make it work for all the file formats (at least bandpass filter + sin removal)
Events: advanced process for recombining. Example: http://www.erpinfo.org/erplab/erplab-documentation/manual/Binlister.html
- Colormaps:
- Create a colormap similar to MNE, where extrema are bright
- NIRS:
- Add new data type
- Display of sensors by pairs oxy/deoxy (red/blue), overlaid
- Images of amplitude: [sensor x time], [trial x time], scout: [trial x time]
Can be done with Matrix > View as image: extract cluster, concatenate for all trials
- 2D Layout for multiple conditions
Filtering: Use short FIR filters instead of IIR for bandpass, to limit the ringing
Or allow the users to edit the LowStop parameter in bst_bandpass.- Show easily recordings maximum/values in the file viewer
- Copy figures to clipboard (with the screencapture function)
Interface
- Global colormap max: Should get the maximum across all the open files
- Open new windows as tab
Connectivity
- Tutorial coherence [1xN]
- Thresholding the connectivity matrices
- t-tests on connectivity measures
- Graph view:
- Fixed scales for intensity sliders
- Fix zoom in one region
- Text bigger
- Too much data in appdata
- Other metrics:
- Coherence by bands: bst_coherence_band_welch.m
- Granger by bands: bst_granger_band.m
- Inter-trial coherence
- Work on progress bars
Processes
- Distributed processing: Brainstorm that can run without Java
- Use Matlab Coder to optimize some processes: Wavelets, bandpass filter, sinusoid removal
- Allow processes in Python and Java
Call FieldTrip functions from pipeline editor
- Optimize opening time ofr for the menu "Add process"
- SSP:
- Make average the default option
Display warning if changing the ChannelFlag while there is a Projector applied
Show where the attenuation is projected:
(sum(IK,2)-sum(SSP(k,:)*IK,2)./sum(IK,2)
Bandpass:
- Offer option: bst_bandpass_fft / bst_bandpass_filter
- Rewrite without the force low-pass filter at Fs/3
- Show warning when using inappropriate high-pass filter (precision too high)
- Use FIR filter
- Spectral flattening (John):
- Apply on the signal before any frequency/connectivity/PAC analysis
- Using ARIMA(5,0,1)
Example code:
arm = mean(lpc(diff(detrend(Fbase')),5)); % detrend and diff PRE = filter(arm,1,diff(detrend(Fbase')))'; % apply to the data| X = filter(arm,1,diff(detrend(F')))'; X(:,1:length(arm)) = []; Time = tndx((length(arm)+2):end)/1000 - SOT;
- PSD: Rewrite to have the same input as coherence
Remove line noise: http://www.nitrc.org/projects/cleanline
- Interpolate the bad channels
- ICA for cleaning eye movements in EEG low density (SSP don't work well)
- Band-limited power envelope
- Reports:
- Save as HTML / PDF
- Do not display the intermediate files
- Pipeline editor:
- Add loops over subjects/conditions/trial groups
- Time-frequency:
- Frequency bands: extended syntax (ex: [2 3 4], 10:5:90, ...)
- How to combine 3 orientations for unconstrained sources
- Display logs as negative
- 2D Layout in spectrum
- Make much faster and more memory efficient (C functions coded by Matti ?)
- TF scouts: should display average of TF maps
- Artifact detection:
- Detection of bad segments in the RAW files (Beth)
- Artifact rejection like SPM: if bad in 20%, bad everywhere
- Test difference between adjacent samples
- Average:
- Remember how many trials were used per channel
- Save standard deviation
- Display standard deviation as a halo around the time series
- Co-registration of MEG runs:
- SSP: Group projectors coming from different files
- Finish validation of the method
- Apply to continuous recordings for correcting head movements
- Resample continuous files
Current Source Density (CSD) => Ghislaine
http://psychophysiology.cpmc.columbia.edu/software/CSDtoolbox/index.htmlOther processes:
- Moving average
- Max
- Median
- Significance test (Dimitrios, Leo)
- Spatial smoothing: check / document parameters
Contact sheets & movies: use average of time windows instead of single instants, for each picture.
- Optical flow
- Simulation:
Fix units in simulation processes => no *1e-9 in "simulate recordings"
- Use "add noise" process from Hui-Ling (in Work/Dev/Divers)
Database
- Optimize startup speed
- Start Brainstorm without Java (-nodesktop)
- Sort files by comment
- MEG protocols: More flexible organization of the database; sub-conditions to allow different runs X different conditions.
Group matrix files => allow to process matrix files by trial types
- Add notes in the folders (text files, visible as nodes in the tree)
- Screen captures: save straight to the database
- Rename multiple files
- Allow matrix files to be attached to other files
Distributed computing
Options from FieldTrip:
Loose collection of computers: https://github.com/fieldtrip/fieldtrip/tree/master/peer
Alternative, with less limitations: http://research.cs.wisc.edu/htcondor/
Single multicore machine: https://github.com/fieldtrip/fieldtrip/tree/master/engine
Batch system: https://github.com/fieldtrip/fieldtrip/tree/master/qsub
Documentation: http://fieldtrip.fcdonders.nl/faq#distributed_computing_with_fieldtrip_and_matlab
Source modeling
- Finish dipole scanning (allow the tab to control multiple figures separately)
- Panel Get coordinates: Add button "find maximum"
- Create scout form maximum: Not available yet for mixed/volume head model.
- Stenroos 2014 paper: Include the following methods
Inner and outer skull surfaces generator from FieldTrip (needs SPM, probably not so different from BST)
- Nolte corrected-sphere model (good model re:Alex)
- Fast BEM models
- Visualize Beamformer results:
- Read CTF SAM .svl
- Display as layers in the MRI viewer
- Unconstrained sources:
- Compute unconstrained and then project on the normal ?
- Difference and stat should be: norm(A) - norm(B)
- Stat and connectivity: what to do? (re-send email John+Sylvain)
- Overlapping spheres: improve the estimation of the spheres for the frontal lobes
- Volume grid:
- Scouts 3D
- Test volume sources with all the subsequent processes (timefreq, stat...)
- Optimize: 3D display (better than 9x9 cubes)
- Optimize: vol_dilate (with 26 neighbors)
- Magnetic extrapolation: Do the same thing with EEG
- Noise covariance matrix:
- Display with figure_image()
- Storage of multiple noise covariance matrices (just like the head models)
- Always save as full, then at inversion time, we can decide between full, heteroskedastic (diagonal) or homoskedastic (i.i.d, scalar)
Problem of having inividual trials + averages in the condition => Display warning or not?
- Save nAvg in noisecov file, to make it easier to scale to other recordings
When deploying to other conditions: Apply destination SSP (NoiseCov = SSP . NoiseCov . SSP' )
Sources on surface: Display peak regions over time (time = color) => A.Gramfort
- Simulation: synthesize pseudo data-files from a cortex patch (duration, amplitude, noise)
Calculate ImagingKernel * Gain for a scout
- EEG Source modeling: Manage references and bipolar montages properly (maybe not necessary)
- MEG source modeling: Do reconstruction only for a subset of sensors for estimating dipoles?
- Processes compute head model and sources: Additional option to set the file comment
- Time-frequency beamformers:
- Band-pass everything in different frequency bands + Source estimation + TF
- Ask data to Sarang where he sees effects that cannot be extracted with MN followed by TF
- Dipole fitting: We will not implement non-linear fits, recommended use of dense volume grids
- Display volume scouts and surface scouts at the same time
- Process "Extract scouts time series": Add PCA option (replace isnorm with choice PCA/Norm)
Anatomy
- Project all sub-cortical structures to default anatomy
- Warping: Scale option has to be fixed, it is currently very unstable
- Scouts:
- Display edges in the middle of the faces instead of the vertices
- Project scouts betweens subjects and between hemispheres
- Display scouts in a tree: hemisphere, region, subregion
- Downsample to atlas: allow on timefreq/connect files
- Sort scouts by region in process options
- Co-register MRI and CT for electrodes marking in the MRI Viewer
- Generate mixed density surfaces
- Import / registration:
- Major bug when importing surfaces for an MRI that was re-oriented manually
- Use mid-gray instead of pial surface?
- Smooth surface: Fix little spikes to irregularities in the mesh
Menu head model > Copy to other conditions/subjects (check if applicable first)
- Add cerebellum to default model generated with "Import FS anatomy"
- Add eyes models to attract eye activity
Statistics
- ANOVA: Use LENA functions
- Output = 1 file per effect, all grouped in a node "ANOVA"
- Display several ANOVA maps (from several files) on one single figure, using a "graphic accumulator", towards which one can send any type of graphic object
- Permutation tests:
- t-test only (wilcoxon? sign-test?): paired, equal var, unequal var
- nb permutations ~ 1000
- maximum statistic over "time" or "time and space"
- Permutations / clustering: cf fieldtrip
Threshold in time: keep only the regions that are significative for contiguous blocks of time, or over a certain number of time points
=> Process that creates a static representation of a temporal window- t-test on volume sources
Paired t-test on unconstrained sources: (convert to flat + Z-score) => AnneSo
- Question of Gaussianity of the samples: take a subset of samples + Kolmogorov-Smirnov / Shapiro-Wilk test
http://www.mathworks.fr/fr/help/symbolic/mupad_ug/perform-shapiro-wilk-test.html
http://www.mathworks.fr/fr/help/symbolic/mupad_ref/stats-swgoft.html
http://stackoverflow.com/questions/14383115/shapiro-wilk-test-in-matlab
- Create icons for Stat/PAC, Stat/Sprectrum, etc.
- One sample t-test across subjects
Input / output
- Finish MINC/CIVET integration (finir lecture MINC2: P Bellec)
- Send email to CIVET mailing list when done
FieldTrip structures: In / Out (see fieldtrip/utilities/ft_datatype_*)
- EEG File formats:
EEG CeeGraph
- EGI: Finish support for epoched files (formats 3,5,7)
- BCI2000 Input (via EEGLAB plugin)
- EEGLAB import:
- Selection of conditions in script mode
- Support for binary AND epoched files (now it's one or the other)
- Allow epoched files with recordings saved in external files (now external files implies continuous recordings)
- BST-BIN: Add compression
- Review raw on all the file formats (ASCII EEG and Cartool missing)
Distribution & documentation
- Document iEEG in "What's new"
- Rewrite basic 12+3 tutorials: group in one series
- Illustrate 2D Layout
- Illustrate unconstrained sources
- Cleaning threads on the forum
Add Help buttons and menus (in popups, dialog windows...) => Links to the website.
- Publication list: Fold by years
- Finish existing tutorials:
- Dipoles
- Auditory: Extend to MNE/EEGLAB/SPM
- Auditory: Finish scripts
- Group MEM/Epilepsy + Epilepsy tutorials
- New tutorials:
- MEG connectome
- Scrambled faces (SPM/MNE-Python)
- Describe all the processes
- Statistics
- Coherence (cortico-muscular ?)
- Intra-cranial recordings (Average ref by groups using Comment field)
- Co-register MEG runs (Beth)
- Missing in the introduction tutorials:
- Volume scouts
- First steps: Brainstorm preferences
- First steps: Temporary folder
- Exploration: Clusters
- Headmodel: explain the fields + how to get the constrained leadfield
- Sources: Model evaluation (by simulating recordings)
- Sources: Theshold min. size (not documented yet)
- Time-frequency: Description of "log freq scale" option
- Modify a structure manually: Export to Matlab/Import from Matlab
- Missing in tutorial "Export to SPM": Add section "Compare with Brainstorm"
- Missing in page "Cite Brainstorm": Add all the methods used in the software
- Reference tutorials on Google scholar
Play videos on wiki with <video> tag (save videos in .ogg)
Current bugs
- Workshops:
- Create scouts doesn't work: scout created on the other side of the brain
- Import anatomy folder: Out of memory on Win 32bits (restart Matlab)
- Compute head model: Out of memory on Win 32bits
- Bandpass filter: Out of memory (Auditory workshop)
- Bug workshop Michael (Mint 12/gnome3/linux 3.0.0-12, KWin/Mutter/Compiz)
- Cannot import two .ds folders at once
- Colors inverted in the PSD/TF plots (power vs. log)
- Cannot type the name of a channel in "Detect ECG"
- Image viewer: Difficult to get to 100%
- Canolty maps computation: Fix progress bar
- Smooth TF windows that contain multiple plots
- Record tab: Text of epoch number is too big on MacOS
- in_bst_data_multi: If trials have different sizes, output is random (the one of the first file)
- tree_dependencies: sources files, reprojected on default anatomy; If based on data files that are bad trials, they should be ignored by tree_dependencies, and they are not
- Image viewer has some bugs on some systems
- Screen capture when there is a fading effect in the window manager: captures the window
- Close figure with coherence results should hide the "frequency" slider
- Edit scout in MRI: small modifications cause huge increase of the scout size
- Reports: Text size is too small with Java 1.5 (2006b-2007a)
- Optimize MRI viewer with patch() instead of image()
- Matlab 2014b bug with rmdir/movefile: Enter but never returns from the call
- 2DLayout (time series):
- Sometimes the lines are not visible
- Does not work when DC offset is not removed
- 2DLayout (TF): Images are too far apart with EEG 20 channels
- Colormap menus: Do not work well on compiled MacOSX 10.9.5 and 10.10
- Event markers are not visible anymore with the sequence: Open MEG, open EOG, close MEG.
Geeky programming details
- Removing all the CTRL and SHIFT in the keyboard shortcuts
- Hide Java panels instead of deleting them
- mri2scs: convert arguments to meters
- Interpolations: Use scatteredInterpolant, griddedInterpolant, triangulation.nearestNeighbor (Matlab 2014b)
- bst_warp and channel_project: Use tess_parametrize_new instead of tess_parametrize
- Shared kernels: do the "get bad channels" operation in a different way (reading all the files is too slow)
- Optimize bst_get:
- Now study and subject have necessarily the same folder name
- Replace big switch with separate functions
- Progress bar:
- Add different levels (to handle sub-processes)
- Make work correctly with RAW on resting tutorial
- Uniformize calls in bst_process/Run
- Add a "Cancel" button
- Line smoothing / anti-aliasing (time series figures)
- Fix all the 'todo' blocks in the code
- Replace handle "0" with bst_get('groot')
- At the end of bst_startup in compiled mode, replace loop with waitfor(jFrame)
- Error message: Add a link to report directly the bug on the forum
in_bst_results (isFull=1): DataFile should be relative