MBIRN: Reproducibility of T2-based multispectral data & derived morphometry

From NAMIC Wiki
Jump to: navigation, search
Home < Reproducibility of T2-based multispectral data & derived morphometry

UPDATES

  • May 17, 2005 (J. MacFall, J. Jovicich)
    • Reproducibility of white matter lesions with T2/PD/FLAIR/T1: controls and depressed subjects with white matter lesions are being recruited for acquisition of test-retest data on three platforms: GE 1.5T, Siemens 3T, GE 4T. Expect to scan first subject within a month. Although nice, it's unlikely that we can fit an additional DTI scan in this 1-hour protocol. Analysis plan is to start using Guido Gerig's lesion detection tool. We could use the NAMIC drive to push the distribution of this tool. NEED: recommendation for T1 protocol, and RF coils.

  • Feb 11, 2005 (J. MacFall, J. Jovicich)
    • James & Jorge met over the phone for updates
    • The Duke scanners that will be used for lesion reproducibility are: GE 1.5T, GE 4T, Siemens 3T
    • Subjects: (number?) subjects will be scanned at each of the three scanners twice in a 6-month period. I expect to begin recruitment soon and imaging in late March or early April. That means that by early October we should have the whole data acquired.
    • Acquisition Protocols:
      • 3D T1: FLASH 30deg/5deg for 1.5T. Needs feedback from Anders re 3T & Allen re 4T
      • 2D FSE T2/PD
      • 2D FLAIR
    • Analysis: G. Gerig's multi-spectral segmentation tool (lesion, CSF, WM, GM) will be used to look at reproducibility across platforms. Who will be responsible for running this analyses and pushing for a publication?
    • Tests: wouldn't it be prudent to have a normal control be scanned at least on the high field systems to make sure that the acquisition protocols have some preliminary tuning before the subjects with lesions come in?
    • T2 goals for Oct 2005: Presumably we should be able to have acquired all data (two time-points per subject on the three platforms). We should have also completed the segmentations for all first-time lesion brains and completed across-platform reproducibility analysis of these.
    • We have been focusing on the administrative matters of getting the contract finalized. This could not be done without the calibration IRB being finalized and I am happy to report that we now have the IRB in place (Biomedical Informatics Research Network (BIRN) Imaging Reproducibility Study, Duke IRB # 6789-05-2R0). This now allows the Duke contract to be finalized.
    • We reviewed Duke's Active Collaborations
    • We also reviewed MIRIAD progress
    • It was clarified that the Morph BIRN grant 'cuts' reflect simply the new grant cycle that NCRR made for us, so that the budget period is from 09/30/2004 - 05/31/2005 (8 months instead of 12, so it looks like a 33% cut if you think that it corresponds to a full calendar year).

  • Jan 6, 2005 (J. MacFall)
    • Budget cut is being discussed and we are meeting Friday to figure out how to modify Duke's objectives
    • IRB used for the grant submission is not suitable for grant so we have submitted a new one that will require full board review. It may be approved in February 2005.
    • We have a large group of suitable subjects but we have to get permission from the IRB to use the list generated for another study.