# CHANGES TO RELEASE

##2.0.3 (2015-05-19)
- BUG: **sct_register_to_template**: fixed issue related to appearance of two overlapped templates in some cases (issue #367)
- BUG: **sct_register_to_template**: now all input data are resampled to 1mm iso to avoid label mismatch (issue #368)
- OPT: **sct_create_mask**: improved speed

##2.0.2 (2015-05-16)
- BUG: **sct_fmri_compute_tsnr**: fixed issue when input path includes folder
- BUG: **sct_orientation**: now possibility to change orientation even if no qform in header (issue #360)
- BUG: **msct_smooth**: fixed error with small Hanning window (issue #363)
- BUG: **sct_straighten_spinalcord**: fixed issue with relative path (issue #365)
- NEW: **sct_label_utils**: added new function to transform group of labels into discrete label points
- NEW: **sct_orientation**: added a tool to fix wrong orientation of an image (issue #366)
- OPT: **sct_register_to_template**: twice as fast! (issue #343)

##2.0.1 (2015-04-28)
- BUG: **sct_extract_metric**: MAP method did not scale properly with the data. Now fixed (issue #348)
- BUG: fixed issue with parser when typing a command to see usage (it crashed)

##2.0 (2015-04-17)

- NEW: **sct_fmri_compute_tsnr**: new function to compute TSNR from fMRI data (performs moco before)
- OPT: **sct_straighten_spinalcord**: now MUCH faster and more accurate (issue #240)
- OPT: **sct_register_to_template**: allows more flexibility by allowing multiple steps for registration (flag -p).
  - N.B. flag "-m" has been replaced by "-s"
- OPT: **sct_register_multimodal**: allows more flexibility by imposing only one stage. Several stages can be run sequentially and then transformations can be concatenated.
  - N.B. flags "-s" and "-t" were replaced with "-iseg" and "-dseg" respectively
- OPT: **sct_extract_metric**: 
  - new methods for extraction: maximum likelihood and maximum a posteriori, which take into account partial volume effect
  - now possible to specify global regions for extraction with flag -l: wm, gm, sc
  - now possible to include a bunch of labels using ":". Example: 2:29
- NEW: **sct_get_centerline_from_labels**: obtain a centerline using a combination of labels and/or segmentations
  - N.B. sct_get_centerline was renamed for sct_get_centerline_automatic
- NEW: **sct_compute_ernst_angle**: new script to compute and display Ernst angle depending on T1 and TR
- OPT: **sct_process_segmentation**:
  - can compute average CSA across vertebral levels or slices
  - can compute length of segmentation
  - can compute CSA on non-binary images such as probabilistic gray/white matter maps
  - N.B. process names were simplified to: "csa", "length" and "centerline"
- OPT: **sct_crop_image**: now possible to crop an image based on a reference space
- OPT: new WM atlas: added gray matter and CSF for computing partial volume
- OPT: now use all available cores for ANTs and adjust variable when running dmri_moco (issue #238)
- INST: new installer in python, simpler to use and check for latest patches
- REF: msct_parser: new parser that generate documentation/usage
- REF: msct_image, sct_label_utils: smoothly converting the toolbox to objet-oriented, some scripts can be used as python module

##1.1.2_beta (2014-12-25)

- BUG: sct_dmri_moco: fixed crash when using mask (issue # 245)
- OPT: sct_create_mask: (1) updated usage (size in vox instead of mm), (2) fixed minor issues related to mask size.
- INST: links are now created during installation of release or patch (issue ).

##1.1.1 (2014-11-13)

- FIX: updated ANTs binaries for compatibility with GLIBC_2.13 (issue: https://sourceforge.net/p/spinalcordtoolbox/discussion/help/thread/e00b2aeb/)

##1.1 (2014-11-04)

- NEW: sct_crop: function to quickly crop an image.
- NEW: sct_extract_metric (replaces the old sct_estimate_MAP_tracts.py). New functionalities added (maximum likelihood estimation and tract grouping). More flexible with label files.
- NEW: sct_convert_mnc2nii
- NEW: sct_create_mask: create mask of different shapes (cylinder, box, gaussian). Useful for moco.
- NEW: sct_fmri_moco: motion correction function for fMRI data. Uses regularization along z.
- NEW: sct_compute_mtr: compute MTR
- NEW: sct_otsu: OTSU segmentation (usefull for DWI data)
- NEW: sct_resample: quick upsample/downsample 3D or 4D data
- NEW: sct_segment_greymatter: function to segment the grey matter by warping that one from the atlas
- OPT: sct_orientation can now be applied to 4d data
- OPT: sct_register_multimodal now using the new antsSliceReg method that regularizes along z.
- OPT: new version of the white matter atlas: more accurate, deformation accounting for internal structure (use BSplineSyN instead of SyN).
- OPT: sct_dmri_moco now using the new antsSliceReg method that regularizes along z.
- OPT: removed all .py extensions for callable functions (created links)
- OPT: sct_label_utils: now possible to create labels. Also added other useful features.
- INST: now possible to specify installation path for the toolbox
- INST: conda dependences are now automatically installed by the installer.
- INST: added pillow (fixed issue #117)
- INST: "getting started" now provided via example commands in batch_processing.sh
- REF: sct_straighten_spinalcord (fixed issues #56, #116)
- TEST: major changes on the testing framework for better modularity with Travis. Now using separate small dataset.

##1.0.3 (2014-07-30)

- BUG: fixed bug in sct_process_segmentation.py related to import of scipy.misc imsave,imread in miniconda distrib (issue #62)
- BUG: fixed bug in sct_process_segmentation.py related to import of PIL/Pillow module (issue #58)
- OPT: sct_register_multimodal now working for images with non-axial orientation (issue #59)
- OPT: sct_register_straight_spinalcord_to_template has now been replaced by sct_register_multimodal in sct_register_to_template.
- OPT: major improvements for sct_dmri_moco, including spline regularization, eddy correction, group-wise registration, gaussian mask.
- OPT: sct_check_dependencies.py can now output log file and do extensive tests (type -h for more info)
- NEW: sct_apply_transfo.py: apply warping field (wrapper to ANTs WarpImageMultiTransform)
- NEW: sct_concat_transfo.py: concatenate warping fields (wrapper to ANTs ComposeMultiTransform)
- NEW: batch_processing.sh: example batch for processing multi-parametric data

##1.0.2 (2014-07-13)

- NEW: virtual machine
- BUG: fixed sct_check_dependencies for Linux
- BUG: fix VM failure of sct_register_to_template (issue #41)
- OPT: sct_register_to_template.py now registers the straight spinal cord to the template using sct_register_multimodal.py, which uses the spinal cord segmentation for more accurate results.

##1.0.1 (2014-07-03)

- INST: toolbox now requires matplotlib

##1.0 (2014-06-15)

- first public release!

##0.7 (2014-06-14)

- NEW: dMRI moco
- INST: libraries are now statically compiled
- OPT: propseg: now results are reproducible (i.e. removed pseudo-randomization)

##0.6 (2014-06-13)

- Debian + OSX binaries
- BUG: fixed registration2template issue when labels were larger than 9
- BUG: fixed bug on PropSeg when the image contains a lot of null slices
- INST: now installer write on bashrc and links bash_profile to bashrc
- BUG: removed random parts in PropSeg

##0.5 (2014-06-03)

- now possible to get both template2EPI and EPI2template warping fields
- fixed major bug in registration (labels were cropped)
- NEW: probabilistic location of spinal levels
- NEW: binaries for Debian/Ubuntu

##0.4 (2014-05-28)

- NEW: installer for ANTs (currently only for OSX)
- fixed bugs

##0.3 (2014-05-26)

- major changes in sct_register_multimodal
- fixed bugs

##0.2 (2014-05-18)

- NEW: nonlocal means denoising filter
- NEW: sct_smooth_spinalcord --> smoothing along centerline
- fixed bugs

##0.1 (2014-05-03)

- first beta version!