Paired brain MR-ultrasound registration¶
This demo aims to register pairs of brain MR and ultrasound scans. The dataset consists of 22 subjects with low-grade brain gliomas who underwent brain tumour resection . The main application for this type of registration is to better delineate brain tumour boundaries during surgery and correct tissue shift induced by the craniotomy.
The dataset for this demo comes from Xiao et al.  and can be downloaded from: https://archive.sigma2.no/pages/public/datasetDetail.jsf?id=10.11582/2020.00025.
Please install DeepReg following the instructions and
change the current directory to the root directory of DeepReg project, i.e.
Please execute the following command to download/pre-process the data and download the pre-trained model. By default, the downloaded data is only a partial of the original one. However the access to the original data is temporarily unavailable.
Launch demo training¶
Please execute the following command to launch a demo training. The training logs and
model checkpoints will be saved under
Here the training is launched using the GPU of index 0 with a limited number of steps
and reduced size. Please add flag
--full to use the original training configuration,
python demos/paired_mrus_brain/demo_train.py --full
Note: The number of epochs and reduced dataset size for training will result in a loss in test accuracy so please train with the full dataset and for a greater number of epochs for improved results.
Please execute the following command to run the prediction with pre-trained model. The
prediction logs and visualization results will be saved under
demos/paired_mrus_brain/logs_predict. Check the CLI documentation
for more details about prediction output.
Optionally, the user-trained model can be used by changing the
demo_predict.py. Note that the path should end with
.ckpt and checkpoints are
logs_train as mentioned above.
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/paired_mrus_brain/logs_predict/<time-stamp>/test/<pair-number>/moving_image.nii.gz, demos/paired_mrus_brain/logs_predict/<time-stamp>/test/<pair-number>/pred_fixed_image.nii.gz, demos/paired_mrus_brain/logs_predict/<time-stamp>/test/<pair-number>/fixed_image.nii.gz' --slice-inds '190,128,96' -s demos/paired_mrus_brain/logs_predict
Note: The prediction must be run before running the command to generate the
<pair-number> must be entered by the user.
 Y. Xiao, M. Fortin, G. Unsgård , H. Rivaz, and I. Reinertsen, “REtroSpective Evaluation of Cerebral Tumors (RESECT): a clinical database of pre-operative MRI and intra-operative ultrasound in low-grade glioma surgeries”. Medical Physics, Vol. 44(7), pp. 3875-3882, 2017.