[IGSTK-Developers] HSMDSS Position Paper for Review

[Brian Blake]

These are excellent comments!! Thus, showing the fact that a neophyte should probably not write a position paper (smile).  I am going to attempt to incorporate all of your comments.  However, I have some quick responses below.  Regards.

-BB

"Stephen R. Aylward" wrote:

> Hi Brian
>
> I am reading over the paper - one quick note: Julien and I are in the
> Department of Radiology at UNC - not comp sci.

Oops, this is an easy change.


>
>
> What about including a organization paragraph at the start of or just
> before section 2?
> Something the introduces the major software challenges and/or themes of
> the paper and provides an outline to the paper.

  The CFP for this forum was very detailed in how they want the paper. So the format is mostly set, but I do believe you are right, a roadmap section would assist the reader.


>
>
> Section 2.2 and 2.3 are weak.   I am not certain what is the goal of
> these sections.  DICOM is the medical image/data communication standard.
>    Various extensions support overlays, CAD results, patient data
> including medical exams being requested, pathology results, etc.
> Tracker data communication isn't part of this standard - but I am not
> certain if this was your focus.   Seems like you need a problem domain:
> for intra-operative guidance there is a need for maintaining a scene,
> spatial relationship between objects, etc - that isn't well addressed by
> existing standards.   Perhaps that can be the focus.   That should then
> reference SpatialObjects and Scene graph theory - The Insight
> SoftwareGuide has a chapter on spatial objects and could be referenced.
>   XML is a file format (kind of) that could be used to transfer a scene...

As stated in the earlier response, the paper submission guidelines are somewhat stringent. They want you to discuss 3 challenges in a particular area, 3 research needs to address the challenges, and 3 next steps in the 5-10 years
timeframe.  The position paper has an Introduction (Section 1), Then
2.1.  3 Challenges
2.2  3 Research Needs to address the challenges
2.3 3 Next Steps
*** This is duplicated in 3.1,3.2, & 3.3 ***

OK, I think this "imposed" format may help you understand the sections/flow of the paper.  But where your comments really pay dividends  is in the content.  If I hear you right, you suggest that DICOM is a sufficient and well-established
interface between components and/or software modules, but there are areas where it may not be sufficient perhaps in capturing meta-information such as specific changes, versioning, diffs, etc. Then perhaps I can build on this with other
non-scene interfaces like the tracker ?

Most of my thoughts were built on my interactions with the IGSTK project and limited medical knowledge (smile), so these discussions are helpful not only for this paper.



>
>
> To me, the major contribution is section 3.   Presents challenges to
> software design, logging, and perhaps goes a bit against norms of reuse
> (you don't necessarily want to make code general - it should be specific
> to ensure validation/verification - but then it isn't good for re-use).
>   etc.
>

Good, I will make this case in Section 3.


>
> Just my thoughts...   I think you write very well and that there is a
> huge need to document this experience as you are doing!
>

Thanks for the input, I am going to try another version prior to Monday, perhaps Sunday and if you have time to look it over again along with Kevin's comments, that would be extremely helpful.

>
> THANKS!
> Stephen
>
> M. Brian Blake wrote:
> > Hi all,
> >
> > I have attached a draft of the HSMDSS position paper that I put together (very rough still). Please read and add comments. Send it back to me and I will merge for submission on Monday. So please try to get me comments by this weekend.
> >
> > Best regards,
> >
> > Brian
> >
> > M. Brian Blake, Ph.D
> > Assistant Professor
> > Georgetown University
> > Washington, DC 20057
> > mb7 at georgetown.edu
> > http://www.cs.georgetown.edu/~blakeb/
> >
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > IGSTK-Developers mailing list
> > IGSTK-Developers at public.kitware.com
> > http://public.kitware.com/cgi-bin/mailman/listinfo/igstk-developers
>
> --
> ===========================================================
> Dr. Stephen R. Aylward
> Associate Professor of Radiology
> Adjunct Associate Professor of Computer Science and Surgery
> http://caddlab.rad.unc.edu
> aylward at unc.edu
> (919) 966-9695

--
M. Brian Blake, Ph.D
Assistant Professor
Department of Computer Science
Georgetown University
Washington, DC 20057
(202) 687-3084
http://www.cs.georgetown.edu/~blakeb