Difference between revisions of "DBP:Harvard:Software:Minutes:5"

From NAMIC Wiki
Jump to: navigation, search
m (Update from Wiki)
 
m (Update from Wiki)
Line 1: Line 1:
−
== DTI Meeting Minutes | May 3, 2006 ==
+
== DTI Meeting Minutes | May 10, 2006 ==
  
−
<br /> Present: Marty, CF, Gordon, Mark, Marc, Doug, KangUk, Sylvain, Marek
+
<br /> Present: CF, Marty, Marc, Mark, Toshiro, KangUk, Doug, Sylvain, Gordon
  
−
<br />'''Scanning'''
+
<br />'''New Sequence:'''
  
−
* Marek is continuing to test out the new sequence
+
* Nan-kuei will be helping with testing of the new sequence
−
* New rule that a licensed MD must request MRI even if for research
+
* Once sequence is decided, planning on bringing in all previous DTI subjects during month of July
−
* Problem with reconstruction with multiple b-value images; causing FA to decrease significantly
+
* Test images
−
* At the scanner, unable to arbitrarily decide the number of baseline images; Marek will try to manually change whatever parameter this is
+
** 1.7^3 resolution
−
* Can get resolution of scans down to 1.5x1.5x1.5, but this may increase noise in the images
+
** Problem with a missing DWI for one slice--causes major differences in tensor estimation
 +
** Strange cross-section artifact superimposed on DWI images (doesn't seem to affect FA maps as much)
 +
** Forebrain appears dimmer than hindbrain
 +
** Gordon will discuss these issues with Marek and Nan-Kuei
  
−
<br />'''Tensor Registration'''
+
<br />'''KangUk - White matter atlas coregistration:'''
  
−
* Goal of this is coregistration of segmented white matter labelmaps with our data
+
* Both linear and nonlinear coregistration still not that great for a problematic case
−
* In test of new tensor registration module from Mattan, linear registration is working better than nonlinear, which is way off
+
* KangUk will test on a better case; will try to do dual-channel coregistration in AG using FA and Mode
−
* KangUk will test this with scans of the same resolution
+
* Will continue to use scalar maps for coregistration; using structural would add an additional registration step and more error
−
* This registration is done at scalar level, and is not actually using the tensors; we can try the same thing using AG and multiple scalar maps
 
−
* Marc will try to troubleshoot this on Friday
 
  
−
<br />'''Anterior Commissure, Mode/FA histograms'''
+
<br />'''DTMRI scalar testing:'''
  
−
* Gordon has sent scripts for creating AC boundary maps
+
* Needs to be an estimation error map in the pipeline; but this is sequence-dependent, so will have not be implemented until new sequence is set
−
* Gordon has sent script for generating histograms; will continue to work on this using labelmaps PNL sends to him.
 

Revision as of 13:19, 18 December 2006

Home < DBP:Harvard:Software:Minutes:5

DTI Meeting Minutes | May 10, 2006


Present: CF, Marty, Marc, Mark, Toshiro, KangUk, Doug, Sylvain, Gordon


New Sequence:

  • Nan-kuei will be helping with testing of the new sequence
  • Once sequence is decided, planning on bringing in all previous DTI subjects during month of July
  • Test images
    • 1.7^3 resolution
    • Problem with a missing DWI for one slice--causes major differences in tensor estimation
    • Strange cross-section artifact superimposed on DWI images (doesn't seem to affect FA maps as much)
    • Forebrain appears dimmer than hindbrain
    • Gordon will discuss these issues with Marek and Nan-Kuei


KangUk - White matter atlas coregistration:

  • Both linear and nonlinear coregistration still not that great for a problematic case
  • KangUk will test on a better case; will try to do dual-channel coregistration in AG using FA and Mode
  • Will continue to use scalar maps for coregistration; using structural would add an additional registration step and more error


DTMRI scalar testing:

  • Needs to be an estimation error map in the pipeline; but this is sequence-dependent, so will have not be implemented until new sequence is set
Retrieved from "https://www.na-mic.org/w/index.php?title=DBP:Harvard:Software:Minutes:5&oldid=3267"