This is the repo for "high resolution mouse brain registration" project. Current implementation is based on section contour matching.

http://seed.ucsd.edu/mediawiki/index.php/High-resolution_Mouse_Brain_Section_Images_Registration

https://www.evernote.com/pub/mistycheney/brain

https://www.dropbox.com/sh/quhklvnr8o5m0qk/p-UiWtPVLq

Raw data in ndpi format in /oasis/projects/nsf/csd183/yuncong

Decompressed images are stored on Gordon /oasis/projects/nsf/csd183/yuncong/tif

/oasis/scratch/csd181/yuncong

https://github.com/mistycheney/registration

• gordon_scripts: contains bash scripts used on the cluster (for running hadoop streaming)

• runSegmentation.sh hadoop streaming script for segmenting multi-section slides

• runAlign.sh hadoop streaming script for doing the alignment

• local_scripts: contains bash scripts used in local machine

• downloadScores.sh bash script for downloading scores pickle files from Gordon

• pickle: stores pickle files

• epydoc: api documentation generated by epydoc (see index.html)

• java_classes: customized InputFormat classes

• src/registration: source code for the registration module

• aligner.py: Aligner class. Main class for registration, provides method for managing the optimization workflow.

• allen.py: Functions for downloading and querying the Allen atlas images

• config.py: Configuration variables

• contour.py: Functions for extracting contours and applying image transformations

• preprocess.py: Functions for segmenting multi-section slides

• scoring.py: ScoreReader Class, with function for computing Hessians

• util.py: Utility functions for plotting, etc.

• viewer.py: Various viewer classes.

• run.py: Main function for doing the alignment locally.

• run_map.py: Mapper for doing the alignment using hadoop streaming on the cluster.

• segment_map.py: Mapper for segmenting multi-section slides using hadoop streaming on the cluster.

• slide_seg_info.txt: text file containing information for multi-section slide segmentation, each row specifies the x,y,width,height of the bounding box for a section. This is created manually.

git clone https://github.com/mistycheney/registration.git
git clone https://github.com/mistycheney/brain_data.git
cd registration
mkdir scores
chmod 766 scores
# upload brain_data/Section/4 to hdfs
source gordon_scripts/runAlign.sh 4 dummy_out src/registration/run_map.py