ITK Release 4/Data Collection: Difference between revisions

From KitwarePublic
Jump to navigationJump to search
(Created page with "= Overview =")
 
 
(46 intermediate revisions by 7 users not shown)
Line 1: Line 1:
= Overview =
= Overview =
== Testing Data ==
* '''Testing Data'''
** (Slides Set 1:[[Media:ITKv4-Testing-Data-Nov-2010.ppt|PPT]],[[Media:ITKv4-Testing-Data-Nov-2010.odp|ODP]],[[Media:ITKv4-Testing-Data-Nov-2010.pdf|PDF]])
** (Slides Set 2: )
== MIDAS Structure ==
* Communities (recursive)
** Collection
*** Items
**** Resources
***** Bitstreams
**** Bitstreams
= Expected Data =
* Maintain data using original acquisition format (Bill's suggestion)?
{| border="1"
|- bgcolor="#abcdef"
! Group !! Expected Amount of Data !! File Formats !! Features
|-
| Raghu Machiraju || 10 Gb ||  TIFF, LSM, MetaImage ||
|-
| Ziv Yaniv ||  10 Gb || DICOM, MetaImage (e.g. segmentation label maps), bmp/jpg/avi (2D frame grabbed US and x-ray fluoroscopy images)||
|-
| Brad Lowekamp || 10 Gb || TIFF ? ||
|-
| Hans Johnson ||  4 Gb || GE45, Phillips, Siemens, Dicom, Analyze, Nifti, etc Migration from :  http://www.nitrc.org/projects/brainstestdata/  || Dictionary of expected meta-information (spacing, size, data-type, etc)
|-
| Sean Megason ||  10 Tb || TIFF, LSM [[File:Megason-A2D2proposal-SCORExx.pdf]]  ||
|-
| Marcel Pratwsaw ||  10 Gb || ??  ||
|-
| Marc Niethammer || 10 Gb || TIFF || Point spread functions (and image, or 12 (key,value) pair parameters
|-
|}
= MetaData =
==Medical==
* physical quantities associated with image acquisition apparatus (e.g. camera calibration for endoscopy and x-ray, US calibration).
* reason for scan (e.g. suspected liver tumor), should be in header but often is not.
* contrast phase (e.g. pre contrast, arterial phase).
* respiratory phase (e.g. end inspiration, end expiration).
* apparatus used to acquire images when not specified in DICOM (e.g. xyz frame grabber).
* rigid transformation data acquired with a tracker (pose of US/endoscope relative to tracking system, required for positioning multiple 2D images in 3D space)
* calibration data of other equipment used in experiment (calibration of tracked pointer)
= Structure =
== Suggested Organization ==
<graphviz>
digraph G {
ITK_DATA
ITK_DATA -> Microscopy_Community
Microscopy_Community -> Raghu_Machiraju
Microscopy_Community -> Sean_Megason
Microscopy_Community -> Brad_Lowekamp
Microscopy_Community -> Marc_Niethammer
ITK_DATA -> Radiology_Community
Radiology_Community -> Hans_Johnson
Radiology_Community -> Ziv_Yaniv
ITK_DATA -> Remote_Sensing_Community
Remote_Sensing_Community -> JPEG2000_Tests
}
</graphviz>
== Strawman of data organization ==
Filename scheme
  AllSubjects/Subject/ScanDateTime/Modality/Subject_ScanDate_Modality.nii.gz
ScanDate is in the form  YYYYMMDDTT  where we only keep 2 digits of time.
so you would have  e.g.
  ITKTestCTdata1/0001/2010122200/T1/0001_2010122200_T1.nii.gz
= MIDAS Installations =
* http://www.insight-journal.org/midas/
** http://www.insight-journal.org/midas/community/view/22 (Miscellaneous collection of ITK related data)
* http://www.insight-journal.org/rire/ (RIRE: Retrospective Registration Evaluation)
** http://www.insight-journal.org/midas/community/view/16
* http://midas.kitware.com/
** http://midas.kitware.com/community/view/7 (ITK Testing Data)
** http://midas.kitware.com/community/view/22 (Tutorials: Virtual Appliances > 2Gb)
= Integration with CTest =
* http://review.source.kitware.com/#change,301
* http://www.kitware.com/midaswiki/index.php/MIDAS%2BCTest
== BrainWeb Datasets ==
* http://mouldy.bic.mni.mcgill.ca/brainweb/
* http://midas.kitware.com/collection/view/44
== XNAT ==
* http://central.xnat.org/app/template/Index.vm
= Data Licenses =
* http://www.kitware.com/blog/home/post/72
** Public Domain Dedication and License (PDDL) — “Public Domain for data/databases” http://www.opendatacommons.org/licenses/pddl/
** Attribution License (ODC-By) — “Attribution for data/databases” http://www.opendatacommons.org/licenses/by/
** Open Database License (ODC-ODbL) — “Attribution Share-Alike for data/databases” http://www.opendatacommons.org/licenses/odbl/
* Designed by the Scientific Commons at http://www.opendatacommons.org/

Latest revision as of 16:44, 20 December 2010

Overview

Testing Data

  • Testing Data

MIDAS Structure

  • Communities (recursive)
    • Collection
      • Items
        • Resources
          • Bitstreams
        • Bitstreams

Expected Data

  • Maintain data using original acquisition format (Bill's suggestion)?


Group Expected Amount of Data File Formats Features
Raghu Machiraju 10 Gb TIFF, LSM, MetaImage
Ziv Yaniv 10 Gb DICOM, MetaImage (e.g. segmentation label maps), bmp/jpg/avi (2D frame grabbed US and x-ray fluoroscopy images)
Brad Lowekamp 10 Gb TIFF ?
Hans Johnson 4 Gb GE45, Phillips, Siemens, Dicom, Analyze, Nifti, etc Migration from : http://www.nitrc.org/projects/brainstestdata/ Dictionary of expected meta-information (spacing, size, data-type, etc)
Sean Megason 10 Tb TIFF, LSM File:Megason-A2D2proposal-SCORExx.pdf
Marcel Pratwsaw 10 Gb ??
Marc Niethammer 10 Gb TIFF Point spread functions (and image, or 12 (key,value) pair parameters

MetaData

Medical

  • physical quantities associated with image acquisition apparatus (e.g. camera calibration for endoscopy and x-ray, US calibration).
  • reason for scan (e.g. suspected liver tumor), should be in header but often is not.
  • contrast phase (e.g. pre contrast, arterial phase).
  • respiratory phase (e.g. end inspiration, end expiration).
  • apparatus used to acquire images when not specified in DICOM (e.g. xyz frame grabber).
  • rigid transformation data acquired with a tracker (pose of US/endoscope relative to tracking system, required for positioning multiple 2D images in 3D space)
  • calibration data of other equipment used in experiment (calibration of tracked pointer)

Structure

Suggested Organization

This is a graph with borders and nodes. Maybe there is an Imagemap used so the nodes may be linking to some Pages.

Strawman of data organization

Filename scheme

  AllSubjects/Subject/ScanDateTime/Modality/Subject_ScanDate_Modality.nii.gz

ScanDate is in the form YYYYMMDDTT where we only keep 2 digits of time.

so you would have e.g.

 ITKTestCTdata1/0001/2010122200/T1/0001_2010122200_T1.nii.gz

MIDAS Installations

Integration with CTest

BrainWeb Datasets

XNAT

Data Licenses