Releases: MICA-MNI/micapipe
v0.2.3
micapipe v0.2.3
- Updated RtD
- Added tSNR volumetric and mapping, matched names on all scripts, and updated
QC_subj
accordingly - Center-aligned images before affine registration.
- Rename transformation json before it was being overwritten by multiple func acquisitions (e.g. task)
- Quiet percentage of completition of MRtrix
- Center align b0 and T1 for the first affine registration
- New flag
-upsample_dwi
for proc_dwi - Fixed MPC cleanup
- Upgraded flair preprocessing with WM-peak, GM mode normalization
- Fixed SINGLE session issues (e.g
proc_func
andQC
) - Fixed R libraries issues for fls-fix in new
Dockerfile
New modules
Both modules include: json files, subject QC and cleanup integration.
-SWM
: superficial white matter generation and mapping in nativepro space.-SWM_MPC
: Superficial white matter MPC in native qMRI space.
v0.2.2
micapipe v0.2.2 'Northern Flicker'
General
- FIX issue with multiple acquisitions when
-proc_dwi
was used with-dwi_acq
. Including the following modulesSC
,QC
andQC_subj
. - FIX instance when multi echo processing has no spike regression file now calculates the nuisance regression correctly.
- FIX an issue with the dimensions of the tSNR file in
proc_func
.
QC group
- Set higher thresholds for GD and SC in the similarity subjects matrices for better outlier detection.
- NEW vertex-wise and ROI variance per feature.
- NEW histograms of features and variances per ROI and vertex-wise.
QC_subj
- NEW a surface plot of the tSNR to the
proc_func
module inQC_subj
.
v0.2.1
v0.2.0
micapipe v0.2.0 'Northern Flicker'
👥 Compatible with multiple acquisitions (func, dwi, MPC). multiple quantitative maps, DWI acquisitions and func acquisitions including task
🧠 We improved the registration between between modalities to optimize the performance.
🌐 Surface mapping from native space in all modules.
🔬 Optimize to process ultra high resolution 7T datasets, with new CNN based processing tools. (tested on 0.5-1mm3, e.g 7T)
💦 MP2RAGE processing workflow (3T and 7T) for structural images, including a cleaning of the salt and pepper background noise.
✨ Connectomes are NOW stored as GIFTI instead of txt. these files use less storage and are faster to load.
⚡️ NEW FLAIR processing module
🔍 New QC subject and group level more informative with a lot of new features and pdf (new json files for QC)
👾 New algorithms (brain mask, surface reconstructions - fastsurfer, registration)
v0.1.5
General changes
- Update print help
- Removed unused flags
- Update the documentation (partially)
- UPDATE jsons to increase
pyBIDS
compatibility - FIX bug on MPC output directory
proc_func
and FC
- Rename flags on
proc_func
-regress_WM_CSF
to-NSR
- FIX bug on glm regression matrix of cofounding (higher precision)
proc_dwi
and SC
upgrade:
- Manage multiple acquisitions
-dwi_acq
micapipe_cleanup
is up to date to erase multiple acquisitions-b0thr
allows the user to set a threshold to determine the volumes that correspond to b=0, default=61-no_bvalue_scaling
disable the diffusion b-values scaling of the corresponding DWI norm (see Mrtrix3 for further info)
v0.1.4
Update FSL from 6.0.0 to 6.0.3
mica-pipe
command goes to micapipe
To print help should specify it with the flag micapipe -h
or micapipe -help
post_structural
will always run schefer-400 by default if is not included in the -atlas
list
proc_rsfmri
is deprecated and replaced by proc_func
proc_func
handles more than one functional acquisition (e.g. tasks), as well as multi echo data (tedana https://github.com/ME-ICA/tedana).
proc_func
Added option to drop the first 5 TRs -dropTR (by default is not dropped)
proc_func
Added option to not run the functional connectomes -noFC (only func surface data)
proc_func
Added 6 parameters of motion to the regression of -regress_WM_CSF (func~spikes+6motion+wm+csf)
proc_func
Added 6 parameters of motion to the regression of -GSR (func~spikes+6motion+wm+csf+gs)
MPC
can processes more than one quantitative map (at a time) with the flag -mpc_acq <qMRI_name>
micapipe_cleanup
can be called from micapipe
command: micapipe -cleanup
micapipe -cleanup
uses the string -acqStr
to erase multiple acquisitions of -proc_func
and -MPC
Improved comments and print logs of -proc_func
and FC.py
proc_func
generates new jsons files of each acquisition with metadata about processing and completion status
proc_func
exit status when Melodic/FIX fail
-QC_subj
is not compatible with proc_func
yet.... or with MPC multiple acquisitions but it is still with the old proc_rsfmri
outputs
NOTE: the read the docs is not updated yet.