Status%20and%20way%20forward%20with%20%20Publishing%20DICOM%20in%20XML

1
Status and way forward with Publishing DICOM in
XML

• Cor Loef
• Chair ad hoc group on Publishing DICOM in XML

2
Ad hoc WG on DICOM Standard publication and
maintenance in XML

• Started in WG 10, work item approved September
  2002
• Formally moved as ad hoc group to WG 6 in 2003
• Decision on way forward to DSC in Dec 2003

3
Rationale

• DICOM standard big and complex
• Difficult to maintain coherence
• Difficult to implement

4
Objectives

• Create architecture for an XML representation of
  standard
• Devise procedures for maintenance and publication

5
Benefits

• Improve consistency of standard
• Multiple publication formats
• Word, PDF, HTML, etc.
• Improve efficiency of updates
• Improve efficiency and consistency of
  implementation, test tools, DICOM toolkits
• Establish more powerful method of information
  modeling

6
Plan

• Draft framework
• Convert Part 6
• Convert significant portions of Part 3 and Part 16

7
Decision to be taken, as written in workitem
An formal evaluation of the validity of the XML
approach will take place the moment part 6, part
16, and a significant part of part 3 are
completed in draft, with a plan and concept for
the complete DICOM standard. At that moment, and
based on the evaluation, the DICOM Standards
Committee will take a decision whether or not to
continue with the approach. The group is
expected to deliver in 2003 the deliverables of
the first phase. The goal is then to have part
6, 16 and 3 published in 2004.
8
(No Transcript)
9
Directory Structure
Root directory
???
Individual parts
Folders for documents, stylesheets, schemas,
common entity definitions, etc.
??
10
Part Directory Structure
main
chapter
? ? ?
appendix
? ? ?
Entity definitions
Other files
Images
11
main.xml
lt!-- entity definitions, including part title
--gt ltbook label"6" id"Part06"gt title-matter
chap01 chap02 chap03 chap04 chap05
chap06 chap07 chap08 appenA lt/bookgt
Boilerplate text, with part title as parameter
Each chapter is in a separate file (Better done
with XInclude and catalog)
So is each appendix (annex)
12
DocBook

• Widely used DTD for technical content
• Actively developed and maintained by OASIS
• Default stylesheets
• Customizable
• One content ? multiple renderings
• XSL-FO
• HTML

13
Output (for each Part)

• FO ? PDF
• Single HTML page
• Chunked HTML pages
• HTML Help
• Benefit
• Cross-references
• Within a part
• Across parts

14
Current Status

• Complete conversion of Part 6
• Majority of Parts 3 16 converted
• To do
• P3 Tables A.1-1, A.1-2, Annex K (SR encoding
  example), Annex P (Index)
• P16 some context groups, Annex F (Japanese)
• Decide how to represent constraints on IODs,
  macros, templates
• Complete conversion of Part 12
• For process verification

15
Deliverables

• Part 3
• Part 6
• Part 16
• Part 12

16
Plan

• Continue with this effort
• Target whole DICOM standard in XML as draft
  available for WG6 review June 2004
• Each part will be assessed on the need to
  structure the text

17
Conversion from Word

• Tool Majix
• http//www.tetrasix.com
• Converts RTF to sdocbook DTD (Simplified DocBook)
• Works mostly, but needs tweaking
• Prose and simple tables converted with this
• Other tools are available, but none can do this
  perfectly

18
Conversion of Templates

• Special processing required to recover structure
  from Word table
• Java program written to do this
• Works well
• Caught typos

19
Figures

• Currently
• Majix outputs them from Word as WMF
• Use as is, or convert to GIF, JPEG
• Eventually, should be redrawn in SVG
• Some can be produced programmatically
• E.g. Mammo CAD SR IOD template hierarchy

20
Math

• DocBook has minimal support for mathematical
  expressions
• Equations should be rendered as graphics and
  included inline
• Until MathML gets wider support

21
Tools

• For general XML editing, any text editor can be
  used
• emacs
• vi
• XML editors
• XML Spy
• Morphon (free)
• XML Mind (xxe)

22
XSLT

• Xalan
• http//xml.apache.org/xalan
• Saxon
• http//saxon.sourceforge.net
• Version 7 ( beyond) supports XSLT 2.0
• xsltproc
• http//xmlsoft.org/XSLT
• In C
• Much faster than the others (in Java)

23
XSL-FO

• Needed to produce PDF
• XEP
• http//www.renderx.com
• XSL Formatter
• http//www.antennahouse.com
• FOP
• http//xml.apache.org/FOP
• Free, but less coverage than the others

24
XML Catalog XInclude

• XML catalog allows better management of the
  collection of files
• XInclude allows inclusion of (fragments of) XML
  files in others
• Included files can be validated

25
What can be done with XML content?

• Fewer typos
• Consistency checking
• Better configuration management
• Generation of tables figures

26
What Else?

• Process for conversion of all Parts
• How to do updates
• Develop utilities
• Generate A.1-1 and A.1-2 of P3
• Generate template hierarchies
• etc.
• Customization of stylesheet

27
What about new content

• How to incorporate new supplements and CPs?
• Transitional period
• No change for authors
• NEMA staff will convert FT to DocBook
• Eventually
• New process
• Do it at granularity level of sections