CTSC:ARRA.012610

Back to CTSC:ARRA supplement

Agenda

i2b2 to XNAT to DICOM architecture and workflow

Harvard Catalyst Medical Informatics group Meeting Minutes January 19, 2010

In attendance:

• Valerie Humblet
• Mike Mendis
• Shawn Murphy
• Bill Tellier
• Mark Anderson
• Randy Gollub
• Yong Gao
• Wendy Plesniak
• Alex Zeitsev
• Charles McGow
• Dan Marcus
• Ron Kikinis
• Jesse Wei
• Karl Helmer

• DICOM conformance statement of each PACS
  • The goal is to make sure that each PACS has the right functionalities so we can be sure that the functions we are building are working against the different PACS.
  • Sometimes the statement might be something difficult to manipulate.
  • The persons responsible for this task are:
    • Steve Piper (mi2b2)
    • Jesse Wei (BIDMC)
    • Bill Tellier and Paul Lamonica (CHB)
    • Katie Andriole (BWH), to be confirmed
    • Thomas Schultz (MGH)

• Obtain samples from DICOM queries:
  • CFIND, CGET, CECHO and CCANCEL
  • Steve will be the one to understand what we want from the DICOM conformance statement
  • Important considerations:
    • Hierarchical versus relational CFIND. The first one is the standard, you need to know who the patient is before getting the images then the serie that has the modality in it. The second one is more flexible.
    • Understanding use of accession numbers in CFIND statement.

• Put phantoms into mi2b2 test PACS system on cloud computing infrastructure:
  • We want to get images from each PACS from the different institutions. In the interest of patient privacy, we will use phantoms.
  • We have to check that there are phantoms available and that the info in the DICOM header is intact:
  • The persons responisble for this task are:
    • Jesse Wei (BIDMC)
    • Bill Tellier (CHB)
    • Mark Anderson (BWH)
    • to be designated (MGH)

• STAR-D software
  • We must decide how we envision that STAR-D will understand who the users are, what the projects are. Mike is the specialist on this subject.

• Image data compression and encryption
  • Dan mentioned those two issues.
  • The PACS use compression, it can complicate our work because our system will need to be able to support then.
  • Not many PACS can support encryption. We may need to have encryption implemented, get a gateway between STAR-D and the PACS.
  • The IRBs are getting very aware that there might be problems during the transmission and hence they require encryption of the data.

• Next meeting:
  • How i2b2 manages users, password, token, users identification
  • Digital signature vs simple sign up