Smart-Upload

Script for uploading dicom files to xnat. This script will clean and decompress the images before uploading them to the ctp. Then, the ctp anonymizes and uploads the files to the xnat platform.

Previous steps

Before launching the script there are some points to take in account. First, you will need to make a file to describe the info that is independent of the project:

• src_name: this is the root folder where the images are found. This folder will be searched in /mnt
• dst_name: this is the folder where all the logs will be saved. Same as before (in process to move the searched folder to /mnt)
• port: port of the ctp
• pipelines: ctp folder where its own scripts are found

Second, you will need to copy the images to the source folder. There are some ways to organize these folders that will be explained later.

Usage

Basic usage:

./maestro.sh -m machine -p project [-d department]

This will all search the images in the source directory. The department is only needed for organizational purposes. If no specified, this will save the logs in destination/project. Otherwise, project will be ignored and department will be used as the project name and saving the logs in destination/project/department

Options

• -a, --all: upload ALL the projects that are in source folder. The projects need to be in a folder tree like source/project/department
• -b, --batches: upload in batches of the specified size. This will make the script wait until the batches are fully uploaded before upload the next batch
• -d, --department DEPARTMENT: Specifies the department name to be used
• -f, --folder: search the DICOM images in source/project/department
• -m, --machine FILE: Specifies the machine name to be used
• -p, --project PROJECT: Specifies the project name to be used
• --help: print the help

Anonimization Script

We also need an anonimization script to specify wich fields we want to remove from the DICOM header. To generate it, we can follow the instructions provided in the anonimization-script folder. Also, to use that generate script with our script, we need to use "PROJECT_NAME_TEMPLATE" as project name. Finally, we need to rename the file to "template_anonymization.script"