CTSC:ARRA.011811

Back to CTSC:ARRA supplement

Agenda

• Update on Phase 3 and Phase 4 mi2b2 software
• Policies and Procedures for PACS access at each institution
• Organization for Human Brain Mapping (OHBM) Conference Abstract

Harvard Catalyst Medical Informatics group Meeting Minutes January 18, 2011

In attendance:

• Bill Wang
• Shawn Murphy
• Alex Zeitsev
• Chris Herrick
• Vincent Roch
• Charles McGow
• Steve Pieper
• Randy Gollub
• Wendy Plesniak (phone)
• Mark Anderson (phone)
• Yong Gao

Update on Phase 3 and Phase 4 mi2b2 software/hardware

• This week Jesse made good progress on the BIDMC mi2b2 installation, using the VMWare. We learned from tests using his research PACS that not all PACS support patient level query roots. mi2b2 currently uses patient level queries to validate medical record numbers against a PACS and return patient demographic information. The alternative to using a patient level query is to query by study and "clean up" the returned data in the mi2b2 server software. Raises the question of what mi2b2 should use as its standard so that it will be most flexible in being able to interface with as many PACS as possible. Would need to change from study level to patient level query root. Dave is thinking through how this would effect the subsequent data query, retrieve and display steps. Chris will contact Dan Marcus to find out best strategy for mi2b2 at Wash U and other sites.

• The Windows 2008 software on the new MGH 64 bit server continues to cause problems for the mi2b2 software. The JAI (Java Advanced Imaging) library software in particular was troublesome so we need to find alternative options. Also ran into problems using Axis2 1.1 - web services were not being registered properly.

• Running into problem retrieving large image data sets (>2000 images/study). Currently, retrieve stops at 2000 images exactly. Not sure source of problem. This is being investigated by Chris and Bill.

Policies and Procedures for PACS access at each institution

We will ask each of our institutional representatives (hopefully at our next meeting) to give us their recommendations for these parameters to use for development and initial implementation. Shawn will send out an email to them asking, Do you want us to exclude archived images? If so, how will we know what images are archived and hence off limits? Are there limitations necessary with respect to the rates and times of day images are retrieved? Any other restrictions that we need to consider? There answers will be used to draft a first proposal to the External Advisory Board indicating:

1. Any limitations by dates of study being retrieved
2. Any limitations by types of studies being retrieved (interaction of first two)
3. Any limitations rates of pull
4. Any limitations as to time of day (interaction of second two)
5. Need for emergency pause capability

Organization for Human Brain Mapping (OHBM) Conference Abstract

Randy and Vincent submitted, with input from many of our team, an abstract for OHBM based on Vincent's analysis of the first preliminary mi2b2 data. She will circulate the final draft to the author list and post the abstract at the time of the meeting in June 2011.