This repository is sharing docker images developped at the Euro-BioImaging (MED-Hub) and The Institute of Biostructures and Bioimaging of the National Research Council (IBB-CNR) for the XNAT platform hosted by University of Turin.
A suite of four Image Quality Assessment (IQA) metrics designed to evaluate the perceptual quality of preclinical and medical image datasets within the XNAT platform. It can also be run locally for testing or development.
It supports scan-level and subject-level analysis using the following metrics:
- Signal to Noise Ratio (SNR) is used to characterise image quality with higher values indicating better image quality
- Sharpness Index (SI) which refers to an image’s overall clarity and detail with higher values indicating sharper images.
- Mutual Information (MI) index calculates the mutual dependency in the images with higher index corresponding to higher similarity between two consecutive acquisitions.This tool is more dedicated to the sequential acquisition data such as the DCE images.
- Rician Noise Level estimation: is based on the Mean Absolute Deviation to estimate the Rician noise in MRI.
| Metric Name | Description |
|---|---|
| snrsi | Signal-to-Noise Ratio:characterizes image quality (higher = better) Sharpness Index (SI)[1]: measures clarity/detail (higher = sharper) |
| mi | Mutual Information (MI)[2]: measures dependency between images (higher = more similar) Dedicated for sequential acquisition data such as DCE images |
| riciannoiselevel | Rician Noise Level[3]: estimation based on the Mean Absolute Deviation to measure Rician noise in MRI |
Output: The tool generates both Excel files and PNG histogram plots for each metric:
- Excel Files (.xlsx): Contains the mean, median and stds for each metric and scan/subject.
- PNG Files (.png) Histogram plots corresponding to each metric.
The Quality Metric Tool is Docker-based and can run on any system that supports Docker (Linux, macOS, Windows). The Quality Metric Tool is Docker-based and can run on any system that supports Docker (Linux, macOS, Windows). No additional software is needed beyond Docker (Install from Docker Official Site)
To build and run the container from the root of this repository
- Pull the appropriate Docker image:
# For scan-level analysis
docker pull iqaxnat_scan:latest
# For subject-level analysis
docker pull iqaxnat_subject:latest
- Verify the Docker image:
docker images
- Running Locally
- Ensure Docker is installed on your local machine.
- Pull the appropriate Docker image as above.
- Prepare input and output directories on your local system.
- Command Format
docker run -v path_of_input:/input -v path_of_output:/output [IMAGE_NAME] [METRIC_NAME]
| Parameter | Description |
|---|---|
| path_of_input | Directory containing scan/subject data |
| path_of_output | Directory to store analysis results |
| IMAGE_NAME | Docker image (iqaxnat_scan:latest or iqaxnat_subject:latest) |
| METRIC_NAME | IQA metric(s), space-separated for multiple metrics |
- Examples
- Scan-Level Analysis
docker run -v /data/xnat/scans:/input -v /data/xnat/results:/output iqaxnat_scan:latest snrsi
- Multiple Metrics
docker run -v /data/xnat/scans:/input -v /data/xnat/results:/output iqaxnat_scan:latest snrsi mi riciannoiselevel
The Quality Metric Tool is designed to run seamlessly inside an XNAT environment through the XNAT Container Service.
Below are the steps required to deploy and run the tool within your XNAT instance.
Log into your XNAT server (or the machine hosting XNAT Container Service) and pull the correct image.
👉 Check this Getting started for more details
For detailed instructions on how to run Docker-based tools inside XNAT (using already pulled images), please refer to the following guide: 👉 [The XNAT Usage Guide]
[1] Erteza A. Sharpness index and its application to focus control. Appl Opt. 1976 Apr 1;15(4):877-81. doi: 10.1364/AO.15.000877. PMID: 20165091.
[2] Coupé P, Manjón JV, Gedamu E, Arnold D, Robles M, Collins DL. An object-based method for Rician noise estimation in MR images. Med Image Comput Comput Assist Interv. 2009;12(Pt 2):601-8. doi: 10.1007/978-3-642-04271-3_73. PMID: 20426161.
[3] McLaughlin PW, Narayana V, Kessler M, McShan D, Troyer S, Marsh L, Hixson G, Roberson PL. The use of mutual information in registration of CT and MRI datasets post permanent implant. Brachytherapy. 2004;3(2):61-70. doi: 10.1016/j.brachy.2004.06.001. PMID: 15374537.
This repository is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0). You may use and share the content non-commercially, with proper attribution, but you may not modify or create derivative works.
If you have any comment, suggestion, or question, please do contact preclinicaldata@eurobioimaging.eu