Web based access to a DICOM image archiving systems (Medical image database and PACS)

Web based access to a DICOM image database

• The Patient level is the top level; single Patient may have multiple Studies.
• A Study comprises one or several Series; a single Study may have multiple Series, but a Series always belongs to a definite Study.
• A Series comprises one or several Images; a single Series may have multiple Images, but an Image always belongs to a definite Series.
• The lowest level is the Image level. The Images coming from a single imaging modality belong to the same Series.

To use this system, the user first activates a WWW browser and specifies the URL of a server. This URL refers to a PHP script (see below) that produces an HTML file with query forms. This query form contains two fields: Patient name and Patient ID (see Fig. 1).

Fig. 1. Query form to select a Patient ( HTML  &  JPEG  files)	Fig. 2. Query form to select a Study ( HTML  &  JPEG  files)

Once a form is completed, the user presses a button to submit the request. The PHP script submits the query to the relational database, to determine those patients that match the search criteria. The user can choose a Patient, which, in turn, causes a list of his/her Studies (see Fig. 2), then Series, and eventually of Images for this Patient to be displayed.

To maintain and support the DICOM archiving system, a set of web-based tools was created. This set includes the following tools:

• a tool to upload client DICOM files to the DICOM database WWW server;
• a tool to edit an item in the DICOM image database;
• a tool to add a DICOM file to the DICOM image database.

The tools allow an authorized remote client to add new data to the database and to update the old ones.

Web based access to a DICOM PACS

• "Patient List"
• "Study List"
• "Series List"
• "Image View"

On-line analysis of DICOM images over the Internet

• DICOM-to-JPEG format conversion;
• Linear and non-linear histogram transformations ( piecewise-linear one and equalization);
• Image sharpening ;
• Isolines and equidensities detection (see Fig. 3);
• Edge detection (see Fig. 4)

Fig. 3. DICOM image view: equdensities ( HTML  &  JPEG  files)	Fig. 4. DICOM image view: edge detection ( HTML  &  JPEG  files)

The set of image processing programs is written in C. These programs can be invoked from a PHP script by client request. Additional image processing and medical analysis facilities can be added to the system in a simple manner, without recompiling the whole application.

Web tools to build remote medical analysis applications

Recently a lot of tools for interfacing a database with the WWW have been developed by users and database vendors. We have chosen to use PHP, a server-side HTML-embedded script language, for this purpose. In syntax this CGI wrapper language is similar to C one, this language supports access to a number of databases, file upload and allows creating complex applications. A number of PHP scripts have been constructed for the Joensuu telemedicine platform, which allow items in the database to be created, retrieved, updated and deleted. The scripts will receive requests from the WWW client via the WWW server, manipulate the relational database, invoke image processing programs and dynamically generate HTML pages to be sent back to the client via the WWW server.

Reference

• Alexander Kolesnikov, Olavi Kelle, Tuomo Kauranne "Remote medical analysis with web technology" Proc. of Finnish Symposium on Signal Processing - FINSIG'97, May 22, 1997, Pori, Finland, p.87.

---

APPENDIX II: List of the demo static HTML pages

mSQL database query	Image processing
"Password" "Patient List" - I "Patient List" - II "Study List" "Series List" "Image View"	Sharpening Equidensities Edge detection Equalization Piecewise linear transform
PACS server query	Dababase management
"Patient List" "Study List" "Series List"	"Add data to Database " "DICOM mSQL Database Editor" "Upload DICOM file"

---

Alexander Kolesnikov