Sign inSign up

biomedia/landmark_detection

By biomedia

•Updated almost 10 years ago

Cardiac landmark detection pipeline

Image
0

421

biomedia/landmark_detection repository overview

⁠Landmark detection pipeline

Original author: Ozan Oktay⁠ Docker version: Jonathan Passerat-Palmbach⁠

⁠Pre-requisites

  • Install docker⁠
  • Download the input data files: (or find a subset here -> /homes/jpassera/playground/dummy_data_biobank) scp -r [email protected]:/homes/jpassera/playground/dummy_data_biobank .

⁠How to run

# change to downloaded directory
pushd dummy_data_biobank
# retrieve latest image
docker pull biomedia/landmark_detection
# run the pipeline on a single subject (results written back to `pwd`/data)
docker run --rm [volumes mappings] biomedia/landmarkdetection <input_file> <output_dir> <path_to_model> <flip input image: true|false>
# concrete example for subject 1000376
docker run --rm -v `pwd`/data/1000376:/tmp/input_data -v `pwd`/results:/tmp/results -v `pwd`/compModel.mat:/tmp/model.mat biomedia/landmark_detection /tmp/input_data/sa.nii.gz /tmp/results /tmp/model.mat true

⁠Information

Subjects data originating from the UK Biobank project⁠

Methodology paper [1] O. Oktay, W. Bai, M. Rajchl, M. P. Heinrich, B. Glocker, D. Rueckert Stratified Decision Forests for Accurate Anatomical Landmark Localization⁠ In IEEE Transactions on Medical Imaging (TMI) 2016.

Tag summary

Content type

Image

Digest

Size

333.2 MB

Last updated

almost 10 years ago

docker pull biomedia/landmark_detection