The metadata repository application and the community
The purpouse of this community is to display and share resources and knowledge on the meta data in the DICOM image file. Our own contribution is an application that can be used in a general PACS environment. The program collects DICOM header meta data and stores it in a separate MySQL data repository. The result is easy access to data describing simple production statistics to quality parameters such as exposure index and patient doses for digital images. An interface with a simple java applet chart program makes display of the data on the internet a simple task.
The application has been on line since november 2004 and today processes and stores 35 000 000 DICOM images annually in 3 large scale image archives
Technical data in digital medical images
The diagnostic image has become more and more well described over the years due to the continous evolution of the DICOM standard in parallel with the revolution in detector technologies för diagnostic imaging. At the same time viewing and processing tool require more and more extensive descriptions of the image file content. Altogether, the description of the image, that is present in every image produced, has become more and more complete. Unfortunately the data is hard to recover and process in a rational way in the clinical environment.
Summary of the information flow in the application
1.
The images are routed to the application by the PACS
2. The information in the DICOM image header is parsed and inserted in a MySQL database with the field names strictly in accordance with the DICOM standard
3. Structured reports are made by scheduled SQL querys from the MySQL metadata respository or online queries are made using the MySQL workbench
4. The application is interfaced to the R statistics application (http://cran.r-project.org/) for online monitoring, see below
5. The data is automatically displayed in the internet explorer via a set of ASP templates and a simple applet chart program (Easy Charts)
Current developments are oriented towards patient dose calculations and image analysis.
Automated monitoring of database parameters
Numerous timelines are created in any large scale radiology institution, indicating developments in patient or detector dose values, e.i. exposure index values. Manual monitoring is time consuming. The recently added possibility to monitor different exposure values automatically simplifies detection of errors in the radiological workflow indicating level changes in detector dose time series with high sensitivity and specificity. The statistical analysis is made using the Robust filter package in the R environment. For details please see reference below.
Read more
DICOM Metadata repository for technical information in digital medical images
Hans-Erik Källman a ; Erik Halsius a ; Magnus Olsson a ; Mats Stenström
Acta Oncologica 2008
Automated detection of changes in patient exposure in digital projection radiography using exposure index from DICOM header metadata.
Källman HE ; Halsius E ; Folkesson M ; Larsson Y ; Stenström M ; Båth M .
Acta Oncol. 2011
Some examples
|
Production statistics Simple production statistics is easily obtained from the metadata repository as the database structure is simple. With the patient identification removed from the data, anyone can work with the data.
 |
Storage space Since the size of every image is reported in the image file the total content in the archive is easily calculated for different modalities.
|
|
Exposure index The dose to the detector is a good indicator for the noise level, thus a parameter much in use in dose optimisation. It can be easily obtained and displayed. Comparisons between different Index values can be made.
 |
Much much more... The image file contains information on dose area product, dose length product for CT, filter thickness, geometry and naturally all the interesting generator parameters. The possibility to process this data rationally greatly enhances the possibility to use the data. Consult the DICOM conformance statement for your imaging device.
|
Download
Download the software at Sourceforge
Installation
Installation support can be attained through the community. Please mail ddsc@ddsc.se for further details.
If you want to install on your own:
Start by downloading and configuring MySQL, for best performance use MySQL 5.
Add the database scheme dcmstat and a user called dcmimport with insert and select rights to this scheme. The dcmimport user should be without a password.
Then run the sql scripts in the DDSC-package to create the database structure needed.
Then you'll need a Java Runtime Environment, the compiled version of the project is built using version 1.6.0, any later version should work. Download a JRE of choice from Sun.
For the application to work you'll need two support libraries: dcm4che2 and mysql-connector-java. Extract them in a directory, for example c.\ddsc. These libraries are included in the ddsc download, you won't need to download them separately.
Extract the files from the DDSC-package mentioned above in the same folder. Start the program with java -jar ddsc.jar from the folder you extracted it to. The first time the application is run you will be asked to complete the configuration of local and remote AE titles, ports and ip addresses. These configuration settings are then stored into the file DDSC.properties. To change them later on, you can either edit this file manually or remove it and restart the application to run the guide again. All modifications to the configuration file requires you to restart the application.
For exended logging change the line in DDSCLoggin.lcf from log4j.rootLogger=warn, R to log4j.rootLogger=debug, R. The information that's displayed with the debug-logging could help us help you with potential problems!
Use of the automated monitoring functionality requires the installation of the R software.
