Difference between revisions of "Engineering:TCON 2007"

From NAMIC Wiki
Jump to: navigation, search
Line 10: Line 10:
 
==2007-04-19 ==
 
==2007-04-19 ==
 
Agenda:
 
Agenda:
* http://www.na-mic.org/Bug/view.php?id=38  A bug in the TIFF IO was identified at the AHM.  Given a high-priority rating by Ron.
+
* http://www.na-mic.org/Bug/view.php?id=38  A bug in TIFF IO.  Given a high-priority rating by Ron.  Volunteer needed.
  
 
==2007-04-12 ==
 
==2007-04-12 ==

Revision as of 14:36, 17 April 2007

Home < Engineering:TCON 2007

Back to TCON:Main

2007-05-03

Agenda:

  • Summer Project Week Kickoff

2007-04-26

Agenda:

2007-04-19

Agenda:

2007-04-12

Agenda:

  • Slicer2 RuleBasedSegmentation Module (for the DLPFC semi-automatic segmentation)
  • Further discussion of large scale experiment control (Dan, Steve, Sebastien)
  • Process for annual report (Tina, Jim)
  • Collaboration grant updates
  • fBIRN / BIRN informatics update (Steve)

Attendees: Jim, Dan, Sebastien, Brad, Tina, Vince, John, Ramsey, Nicole, Stephen, Jeff Notes:

  • (Pre meeting discussion of sunspots)
  • Rule-Based Segmentation Module
    • Marek would like to use it for his data
    • GATech provided info on work so far, Brad has reviewed
    • VTK/ITK implementation issues that can be easily resolved
    • Slicer2 module needs to be integrated with actual filter
    • Delphine and Ramsey wrote the interface and filters based on Jim Fallon's rules
    • Slicer2 implementation relies on subvolume and draw capabilities (discussion of slicer3 implementation).
      • possible new subvolume selector to add to slicer3
      • could use new editor in slicer3 for ROI
      • also need some kind of wizard or detailed instructions to help guide
      • then the rule based code could be a command line module
    • GATech is planning to do a project with Jim Fallon using the Slicer
    • At project week we can look at getting Marek up to speed on Slicer2 and the engineering group can work on a slicer3 implementation.
  • Issue with NAMICSandBox ability to commit? John may be specific to John.
  • Update on Neil's work on the GWiz (Grid Wizard) and discussions with Jeff and Steve at fBIRN meeting
  • Stephen update
    • Starting to do 'batchboard' tests on bspline registration
    • planning public face for registration optimized methods
    • integration of BatchMake and CMake for running tests on multiple platforms
  • Sebastien is doing timing tests on various machines at BWH
    • ITK timer class does the clocking (wall clock timing) but could look at performance timers at the processor API (build in the CPU).
    • Process priority is set high to get best guess of fastest behavior
    • may need an administrative approach to avoid multiple jobs running during night performance tests
    • new features being added to cmake/ctest/batchmake to help determine machine environment, could be extended
  • Progress report mechanics
    • The report describes just the current year (not the cumulative sum of all the years).
    • Tina is going to send email about the themes for the progress report
    • Need to be sure the pdf has clickable URLs for wiki page entries
    • Discussion of Impact for the highlights section
      • non-na-mic people at the programmer events and AHM
  • Update on XCEDE work at fBIRN AHM

2007-04-05

Agenda:

  • Collaboration proposal Peter Y.

Minutes:

  • Attendees: Peter, Dan, Jim, Katie, Steve, Wendy, Nicole, Vince, Tina, Will, Sebastien, Stephen, Brad

Discussion of Peter Yim's plans for disseminating algorithms through the NA-MIC Kit.

  • Biomedical Engineering, UNC, NIH
  • Algorithms for MRA, vessel reconstruction
    • Segmentation, skeletonization
    • University of Medicine and Dentistry of NJ
    • Vascular diagnostics, quantifying angiogenesis, stroke...
  • Algorithms could be valuable, but aren't written in a way that can be disseminated (written in C for IRIS Explorer)
    • Peter needs to have collaborators to re-write the code
    • Peter can provide the clinical background and potential user community
  • Could be good candidates for integration into ITK
    • Need to review what it means for the NIH patented technology to be deployed
    • Some of the algorithms require close interaction and others are more automated
  • Discussion of the possible allocations of effort and collaboration

2007-03-29

Agenda:

Highlights:

Attendees: Will, Jim, Wendy, Csaba, Marek, Nicole, Katie, Dan, Steve, Sebastien, Brad, Ron, Tina

  • Marek - vcsf data on 5 patients. familiar with software and algorithms available in na-mic. segmentation and fine tuning of algorithms is needed in slicer. Would like to use GeorgiaTech algorithm for automated segmentation of DLPFC.
  • Csaba putting details of JHU project on wiki. Will schedule himself into the tcon in about 4 weeks.

2007-03-22

Agenda:

  • Progress Reports - the folks who lead the sections last time will be invited to do the same again this year, or to nominate someone else who they will help.
  • DBP-II - algorithms contacts added by Ross. (UNC: Guido, Harvard: Polina, MIND: Bruce Fischl, JHU: Allen)
  • Invitation sent to DBP2 for this call, with the following agenda suggestion:
    • Notes: Marek has conflict on 3/22 and will attend 3/29 tcon.
    • reminder that he would like your pages to be updated in March. Example of an updated page is: http://www.na-mic.org/Wiki/index.php/DBP2:MIND
    • introducing you to your namic engineering contact points and the role they will be playing. (we expect to introduce you to your namic algorithms core contact points at a later point as well).
    • summer project week at MIT (June 25-29). Ron's expectation is that by then (1) you will have identified a problem that can be solved with the na-mic kit in a 6-month timeframe and the results for which you will use in your work, (2) you will send an engineer and data to this project week, and the engineer will be able to articulate to the larger group what the problem is, and use that week to work on it with the na-mic team.

Jim has notes and attendees to add.

2007-03-15

Agenda:

  • Heads up: 2007 Progress Report. Recommendations needed from Ron on section/theme leads.
  • Dissemination and Outreach for the rest of us (Will, Tina)
  • DBP2 Points of Contact (DBP, Engineering, Algorithm). Need contact information for DBP2s. Algorithm contact points still need to be added. (Update: email sent to Ross 3/1 and 3/15 with request to provide names in march.)
  • Performance analysis tools and strategies: ITK_Registration_Optimization, Slicer3:Performance_Analysis (Steve Pieper, Sebastian Barre, Stephen Aylward, Katie Hayes)
  • Slicer release schedule (Steve)
  • Question from Douglas Alan: At the January NA-MIC All Hands meeting in SLC, Mike Halle and I met with Bill Lorensen and Jim Miller to discuss how in Slicer we might pass a FITS world coordinate system transform (and its inverse) from an ITK FITS image reader to the resulting Slicer VTK image so that Slicer can then be enhanced to use the transform to properly display the world coordinates. The proposed solution was for me to define a subclass of itk::Transform (i.e., FITSWCSTransform) that would handle the WCS transform, and once I had completed this (which I have) for me to then modify my FITSImageIO reader so as to include a pointer to a FITSWCSTransform object in the FITSImageIO's MetaDataDictionary. As it turns out, there seems to be a snafu with this plan: the MetaDataDictionary of an ITK image appears to be stripped from the image whenever it passes through an ITK filter. Or so it is the case, at least, for the filters that I have tried. Any ideas on how we should proceed, in light of this?

Notes:

Attendees: Jim, Steve, Will, Katie, Nicole, Raymundo, Stephen, Sebastien, Wendy, Tina, Jeff

  • Schedule for the progress report in on the wiki. Last year's report is also available. Assignments are imminent.
  • Need to update the NA-MIC web presence. Wiki to web will be used to author and publish the content. Ron will set the vision for the web presence statement (DBP focus? Theme focus? Core focus?). Everyone is responsible ensuring all NA-MIC activities and contributions are advertised.
  • Tina will get the contact info for the DBP2's
  • Stephen is pulling together a test application to run across systems to evaluate registration performance.
    • Note that only one optimizer is relevent to B-Spline registration (LBFGS-B?)
    • Torsten has a hierarchical B-Spline algorithm that Killian uses as the yardstick.
    • Brad developing proposal to evaluate EM pipeline soup-to-nuts for sources of errors.
  • Steve and Katie looking at Callgrind and Callgrind Control to track the overhead of observers and potentially the over-calling of callbacks.
  • Tracker performance okay but the images update slowly
  • Slicer3 release at the end of March
    • DTI tools availability gated by MICCAI deadline
    • Launchers work with getbuildtest but may have impacted CPack installations.
    • ModelMaker needs MRML read/write (Jim!)
    • Filebrowser will be available
    • Any features to suppress?
    • Any additional documentation needed?
  • Slicer 2.7 release at the end of March
    • Mostly bug fixes but a few new features
  • Jim will talk to Mike about the FITS needs for the MetaDataDictionary and the pipeline.

2007-03-08

Agenda:

  • Question from Steve: best way to handle "neurological" analyze files (where right-is-right unlike radiological files where "left -is-right"). The oaisis brains data collection from mBIRN stores their data this way and the ITK IO factory doesn't have a way to know which kind of file these are.
  • Python in Slicer - project overview: Dan
  • Google Summer of Code - are there any project takers?

Notes:

  • Update - it turns out oasis-brains.org analyze files are stored in radiological convention but displayed in neurological convention, so everything is as 'correct' as it can be for this file format. But the ambiguity remains about the real meaning of analyze files, so in the future oasis will supply nifti formated files.

2007-03-01

Agenda:

  • Adding Engr+Algorithms core representatives to each DBP-II team: DBP2:MIND
  • Google Summer of Code
  • SIfTI effort to standardize spatial transform files. (login info will be given on the phone)
  • Technical Slicer3 item: command line executables

Notes:

  • Eng contacts for DBP-II:
    • Katie Hayes for the JHU project
    • Jim Miller for UNC
    • Brad Davis for Harvard (Confirmed)
    • Steve Pieper for MIND
  • Summer of Code
    • Generally agreed to be a good idea if there's time
    • Check with Luis
  • SIfTI
  • Command Line Interface to EMSegmenter
    • Jim will work on making MRML scenes come as arguments to command line modules and give an example that can be used for EMSegmenter and other projects like the ModelMaker.

2007-02-22

Attendees: Wendy, Jim, Nicole, Dan, Andy, Katie, Tina, Ron

Highlights:

  • match dbp-IIs to engineering core members (Jim and Tina to do offline)
  • contact dpbs regarding participation in project week (Tina)

2007-02-15

Agenda:

Attendees: Jim, Nicole, Tina, Katie, Jags

2007-02-08

Agenda:

  • Please note that we now use a new TCON#. If you don't have it, please send an email to namic-eng@na-mic.org.
  • Review remaining Project Half-Week Updates

2007-02-01

Agenda:

2007-01-25

Attendees: Bill, Jim, Steve, Wendy, Nicole, Katie, Brad, Andy & Jags

Agenda:

  • Last call with Bill as the PI...
  • Next TCON (Feb 1) to review 4-block from project half week. Will use Jim's tcon #.

Highlights:

  • New phone number for TCON next week; email Jim for the number
  • Problem with region growing?
  • Bye, Bill! We'll see you in three weeks when you get bored with retirement.

2007-01-18

Attendees: Bill, Tina, Steve, Katie, Sebastien, Zack, Jags

  • Slicer3 release schedule discussion
  • Which teem are we on?
  • Status of filebrowser?
    • Current unix browser does not allow for making a new directory
    • Does the new one support specifying a list of file types? (Yes)
  • Feb 1/8 for project half-week results 4-blocks review?

Highlights:

  • Feb 1 and 8 for project half-week results 4-blocks review
  • 12 days till Bill's retirement...
  • OHMB workshop agenda info by Steve: DTI tutorial for Slicer3 is a candidate topic.

2007-01-11

No tcon this week. AHM_2007 in progress.

2007-01-04

Agenda and Highlights:

  • Project Half-Week Activities - Focus on Slicer3 Projects. And any others that need resolution of title/team. 4-blocks should appear asap. Each project lead will use these to introduce their project on Wednesday, Jan 10th.
  • There are pages in google cache that are not yet on the namic-wiki. What is the process for pulling those in? Should each person who finds a page just cut and paste the text as best as they can? How should we use the "wanted pages" list on the wiki?

Highlights:

  • Slicer3 Projects have been added to the wiki.
  • Wanted page list is being pared down as a high priority item,.
  • Jim and Bill will be on Daniel Goldberg's project half week project.

Attendees: Bill, Will, Ron, Steve, Jim, Stephen, Neil, Andy, Luis, Nicole, Katie, Wendy, Jags, Tina