[Insight-developers] Proposed agenda for developer's meeting

[Terry S. Yoo]

The agenda looks pretty good.  I am concerned that there are some holes
in our education especially in the hybrid deformable segmentation work.
There are also some basic issues such as nonlinear and linear filtering
that are FAQs from people outside the consortium that we should all
know a bit better.

Folks, that's a hint... I think the the Columbia and UPenn folk should
give a tutorial on the *use* of fuzzy connectedness and hybrid
segmentation techniques.  Celina in particular had strong opinions
on what were appropriate segmentation tools for pelvic data.  I believe
that we need to get a "how-to" on some of these other techniques.

Peter will be joining us on Friday.  He would like to start Friday with
a discussion on the June meeting in Boston.  Be prepared to
let him know some basic planning information like how many rooms
he needs to set aside and how many people each group is likely to
bring.

Validation presentations should also contain an overview (brief, folks)
of what is to be done, how far along you are, and items of interest
that might be shared (data, statistical tools, etc.) between sites.  This
is a progess report, folks, rather than a scientific presentation.

I look forward to seeing everyone there.

Terry

========================

Will Schroeder wrote:

> Hi Folks-
>
> Based on feedback on discussion from the ITK developers, here's a 
> meeting agenda that I'd like to consider. What I am particularly 
> interested in is 1) educating the community on the various subsystems 
> of ITK, and 2) designing / redesigning / brainstorming parts of the 
> system  that need it. Of course, there's a lot more information to 
> disseminate about the beta, etc.
>
> I'd like to make a request: if you are educating us, please include 
> examples, source code, etc. so we have something concrete to look at. 
> Don't try to impress, please try to communicate.
>
> I've put names next to these tasks. Please look carefully and see if 
> your name appears, or does not appear, and  make sure that it is 
> appropriate. The moderator is responsible for keeping things on track 
> and breaking up fights :-)
>
> We can talk about it tomorrow on the TCON.
>
> Will
> ----------------
>
> Day 1
> =====
> + Introduction (1 hr Terry)
>
> + Beta Release (1 hr Terry & Will)
>   - advertising
>   - packaging (e.g., stlib) (Will)
>   - contribution process (formal contribution process) (Will)
>
> + Education (15/30/60 minutes per topic - Jim moderator)
>   - Spatial function hierarchy (Damion)
>   - BLOX (Damion/George)
>   - Statistics framework (Jisung Kim and Sayan)
>   - FEM (Jim Gee & Aljaz Noe)
>   - ImageIO (Bill Hoffman)
>   - Neighborhood and Neighborhood Iterators (Josh & Jim Miller)
>   - Image Metrics (Luis, Brian, Lydia)
>
> Day 2
> =====
> + Design (4 hrs. Bill L. moderator)
>   - Exception hierarchy (Bill Hoffman)
>   - Templating process objects (Jim Miller)
>     . removing template parameters (Will Schroeder)
>     . calculations (accuracy of) in filters (Jim Miller)
>     . mutex on light object (Jim Miller)
>     . Extending pixel traits (Luis)
>   - Concept checking (Lydia)
>   - Pipeline update mechanism (Josh/Ross/Will/Jim Miller)
>   - Segmentation framework (Sayan)
>   - Image API (Will)
>   - Missing functionality (Vikram)
>   - Using typedefs better (Jim Miller)
>
> + Using ITK (4 hrs. Terry moderator)
>   - performance
>   - working with templates in non-templated systems
>   - other observations
>   - are we meeting our objectives?
>     is it useful in the way we meant it to useful?
>     target algorithms? (segmentation and registration)
>     are people going to find it useful?
>     the matlab of segmentation and registration?
>
> Day 3
> =====
> + Validation (4 hrs. Terry moderator)
> + Conclusions (Terry)
>
> _______________________________________________
> Insight-developers mailing list
> Insight-developers@public.kitware.com
> http://public.kitware.com/mailman/listinfo/insight-developers