hongkunge / image-registration-by-learning-common-feature

Image Registration by Learning Common Feature

This is part of the research project for medical image registration. The paper of this project is accepted by Machine Learning in Medical Image.

Paper

My paper based on the project got published! Check it here.

Structure of the project directory

Functions in the folder ./5-fold_experiment/utility/ written by Hongkun Ge and All the codes cited by other authors have been given clear indication for copyright.

./

• demo_1.m Test the feature of CT and MR images.
• demo_2.m Overall training and testing process of CT and MR IMAGES in a 5-fold manner.
• demo_3.m Kernel CCA training demo.
• demo_4.m Kernel CCA testing demo.
• demo_5.m Training the alpha1 and alpha2.
• test.m Training and testing process in one-fold experiment.

utility/

The folder includes all the functions I used when training and testing kernel CCA. All the important functions are annotated with explanatory notes.

Script/

The folder includes the experiments of all the methods for comparison.