Classical nonrigid registration for prostate MR images

Note: Please read the DeepReg Demo Disclaimer.

Source Code

This is a special demo that uses the DeepReg package for classical nonrigid image registration, which iteratively solves an optimisation problem. Gradient descent is used to minimise the image dissimilarity function of a given pair of moving and fixed images, often regularised by a deformation smoothness function.

Author

DeepReg Development Team

Application

Registering inter-subject prostate MR images may be useful to align different glands in a common space for investigating the spatial distribution of cancer.

Data

Data is an example MR volumes with the prostate gland segmentation from MICCAI Grand Challenge: Prostate MR Image Segmentation 2012.

Instruction

Installation

Please install DeepReg following the instructions and change the current directory to the root directory of DeepReg project, i.e. DeepReg/.

Download data

Please execute the following command to download and pre-process the data.

python demos/classical_mr_prostate_nonrigid/demo_data.py

Launch registration

Please execute the following command to register two images. The optimised transformation will be applied to the moving images, as well as the moving labels. The results, saved in a timestamped folder under the project directory, will compare the warped image/labels with the ground-truth image/labels.

python demos/classical_mr_prostate_nonrigid/demo_register.py

Visualise

The following command can be executed to generate a plot of three image slices from the the moving image, warped image and fixed image (left to right) to visualise the registration. Please see the visualisation tool docs here for more visualisation options such as animated gifs.

deepreg_vis -m 2 -i 'demos/classical_mr_prostate_nonrigid/logs_reg/moving_image.nii.gz, demos/classical_mr_prostate_nonrigid/logs_reg/warped_moving_image.nii.gz, demos/classical_mr_prostate_nonrigid/logs_reg/fixed_image.nii.gz' --slice-inds '4,8,12' -s demos/classical_mr_prostate_nonrigid/logs_reg

Note: The registration script must be run before running the command to generate the visualisation.

plot

Contact

Please raise an issue for any questions.

Reference

[1] Litjens, G., Toth, R., van de Ven, W., Hoeks, C., Kerkstra, S., van Ginneken, B., Vincent, G., Guillard, G., Birbeck, N., Zhang, J. and Strand, R., 2014. Evaluation of prostate segmentation algorithms for MRI: the PROMISE12 challenge. Medical image analysis, 18(2), pp.359-373.