A program for estimating the intracranial volume of a patient from brain CT scans. The program uses a python script to integrate a dicom to nii converter with the Ants tools-set for obtaining a brain mask using OASIS templates.
The following steps demonstrate how to use the program:
Use the following command to create the container:
docker-compose up -d icv_app
A folder containing CT scan files in the DICOM format should be inserted in the input volume which will be mounted by docker to the container from the following locations at the local driver :
-
For Windows:
\\wsl$\docker-desktop-data\version-pack-data\community\docker\volumes -
For Linux:
/var/lib/docker/volumes
Use the following command to get the output:
docker-compose exec icv_app bash -c "python main.py --pixel_spacing 0.5 --slice_thickness 1"
where pixel_spacing and slice_thickness are optional parameters that determine the voxel unit volume of the scan.
The output result which is the following will be saved in the mounted output volume of the container:
- A file with the IcV estimation
- The brain mask file in NifTI format
- A PNG image of the overlay of the mask in three orthogonal sections
Run a flask web based api of the program with the following command:
docker-compose up -d icv_flask
And send to the api a GET or POST request with a minio server endpoint, access key, server key and optionally pixel spacing and slice thickness. A local minio sever will also be automatically created however the application can integrate with any minio server if it has its endpoint and authorization. The following is an example of a GET request to a local version of the application flask api:
e.g: http://127.0.0.1:5000/run_icv?minio_url=172.17.0.1:9000&minio_access_key=minioadmin&minio_secret_key=minioadmin
The program will search there for a bucket with the name 'dicom-input-data' to obtain DICOM input data and after computing the output will upload the results back to another bucket in the server called 'output-data'.