Clone this repository to adapt the Dockerfile and tool script and rapidly build QMENTA tools. It contains a simple example with all the required files to build and test the tool. If you want to know more about how to use the SDK and all its features, take a look at the documentation.
In order to add tools to the QMENTA platform you need to have developer privileges. If you are interested in this feature, please contact us at info@qmenta.com.
This file contains the sequence of instructions to build a new tool image. You can setup environment paths, run commands during the image building stage and copy files (see Dockerfile commands).
The main script that is executed when a tool is launched on the QMENTA platform. This script typically performs the actions shown below using the QMENTA SDK functions where suitable:
- Download the input data to the container.
- Process it (call any third party tool).
- Upload the results.
The naive example shown in this repository computes the histogram for a range of intensities in a T1-weighted image. The range of intensities can be specified from outside the tool thanks to the tool settings specification.
Feel free to contact us if you have any doubt or request at sdk@qmenta.com! We are happy to hear from you and expand the capabilities of the platform to fit new useful requirements.
This tool code uses this HTML template to populate some fields with the patient data and the analysis results to generate a PDF report.
Use Docker to build a new image using the Dockerfile:
docker build -t image_name .
Where image_name should conform to the syntax my_username/my_tool:version.
The first build may take several minutes since it will need to generate the image layer containing the software installation.
Alternatively, take a look at the standalone.Dockerfile to see how to install the SDK in an image based on Ubuntu.
Optionally, the test_tool.py script can be used to locally launch your tool image if you specify the input files and the required values for you settings (see mock_settings_values.json):
mkdir analysis_output
python test_tool.py image_name example_data analysis_output \
--settings settings.json \
--values mock_settings_values.json
Add the tool to the QMENTA platform
To add a tool image to your list of tools you will first need to login push it to your Docker Hub registry:
docker login
docker push image_name
To register the tool to the QMENTA platform, access the Analysis menu, and go to My Tools. You will need to enter your credentials of your registry, the name and a version number for the tool, its settings configuration file and a description. You can find an example of the settings configuration file in this repository (settings.json).