Difference between revisions of "DBP:Harvard:Software:Minutes:8"

From NAMIC Wiki
Jump to: navigation, search
m (Update from Wiki)
 
m (Update from Wiki)
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
−
'''DTI Meeting Minutes 8-24-05'''
+
==== DTI Meeting Minutes | August 9, 2006 ====
  
−
''Present:'' Marty, Scott, Lauren, CF, Marek, Mark, KangUk, Min, Gordon
+
<br />'''Present:''' Doug, Marc, Marek, Gordon, Sylvain, Katharina, Toshiro
  
−
<br />'''Minutes on the NAMIC Wiki Page'''
+
<br />'''Scanning Update:'''
  
−
Note that a copy of these meeting minutes will be kept on the NAMIC Wiki page
+
* 7 subjects were scanned yesterday. Biggest logistical hurdle is reconstruction time--holds up scanning for 25 minutes
 +
* Data is on the network, but needs to be assigned case numbers
 +
* Sylvain made tensors for the new cases
 +
** One case had FAT artifact, may be bad enough that we will have to bring the subject back
 +
** There is a bug in Slicer where the 3D view and the slice view are decoupled
 +
** 2 cases are sheared strangely--the oblique acquisition may be causing this, may be measurement frame.
 +
* Gordon suggested that we experiment with scanning a subject with head rotated at 45deg.
 +
* Also, if possible the scheme of gradient directions should be changed (need to talk to NanQuei)
  
−
<br />'''New DTI Meeting Schedule -- IMPORTANT!!'''
+
<br />'''DWI inspection for new data:'''
  
−
These 10:00 meetings will be at the Hollenberg on the 1st and 3rd Wednesday of the month, which means that in rare cases when there are 5 Wednesdays in a month (for example, this month), the 5th Wednesday meeting has to be at 1249. NOTE: This means that next week's meeting 8/31 will be at 1249 again!! I'll send reminders next week.
+
* Gordon presented a method for inspecting sequences of DWIs to look for drops in signal or other artifacts. Baseline images and gradient images in a composite, cycling through slices to see abrupt changes in intensity.
−
 
+
* With the new data, a lot of problems are obvious: zipper lines, dropped DWIs, uneven drops in intensity within a single DWI, motion artifacts, FAT artifact (shows up well in AvGradient image)
−
<br />'''Lauren's Fiber Tracking Paper (Tentatively) Accepted'''
+
* PNL can use this method to inspect every new DTI case that comes in
−
 
 
−
Lauren will meet with Marek, Mark, and CF to decide which figures should be replicated in more subjects. Lauren will handle most of the reviewers' technical questions, and will recruit help from others as needed.
 
−
 
 
−
<br />'''Coronal AP Tensors'''
 
−
 
 
−
Although it is ok to use these cases (coronal AP instead of PA) for measures like FA, ADC, etc., there is still confusion as to whether fiber tracking will work on these cases. Mark will look at the glyphs with Sylvain to see if renumbering the images alters the glyph pattern. Remember: renumbering the I files was necessary to give proper coregistration between tensors and structural data.
 
−
 
 
−
<br />'''Fiber Tracking with 2 ROI Method'''
 
−
 
 
−
It appears that there are two Slicer modules for using 2 ROIs to isolate fiber tracts in Slicer, one created by Lauren and the other by Raul. Lauren's appears to have a minor bug involving the 2nd ROI not paying attention to the requested label (Mark will talk to Lauren about this), and Raul's module does not work. There is also the question of once we have these isolated fiber tracts, can we find mean FA/ADC along those tracts. CF will work on this, possibly with a Matlab script.
 
−
 
 
−
<br />'''VolumeMath/Cast Problems'''
 
−
 
 
−
CF will contact Steve Pieper regarding issues we've been having with the float vs. short file types in VolumeMath. Currently we have to cast Scalar Measure volumes as shorts for VolumeMath to work properly, but there is concern that this alters the data.
 
−
 
 
−
<br />'''Background Noise in Scalar Measures'''
 
−
 
 
−
Scalar measures created in the DTMRI module have background noise around the brain. This background noise is not present in Stephan's reconstruction script.
 
−
 
 
−
<br />'''Stephan's LSDIrecon Code'''
 
−
 
 
−
Scott has been looking at Stephan's code to see how compatible it is with Slicer. There was talk of whether to use Stephan's code in Slicer, or come up with a separate "library" to be used both by Stephan and Slicer developers. Scott and CF will meet with Stephan about this. Also, Scott is going to check with Mark to make sure Scott is using valid cases to test these methods.
 
−
 
 
−
<br />'''Higher Resolution DTI Scans Using the 3T'''
 
−
 
 
−
KangUk and Marek have been experimenting with higher resolution scans using the 3T Magnet, and KangUk will work on comparing quantitative measures between these new protocols.
 

Latest revision as of 13:55, 18 December 2006

Home < DBP:Harvard:Software:Minutes:8

DTI Meeting Minutes | August 9, 2006


Present: Doug, Marc, Marek, Gordon, Sylvain, Katharina, Toshiro


Scanning Update:

  • 7 subjects were scanned yesterday. Biggest logistical hurdle is reconstruction time--holds up scanning for 25 minutes
  • Data is on the network, but needs to be assigned case numbers
  • Sylvain made tensors for the new cases
    • One case had FAT artifact, may be bad enough that we will have to bring the subject back
    • There is a bug in Slicer where the 3D view and the slice view are decoupled
    • 2 cases are sheared strangely--the oblique acquisition may be causing this, may be measurement frame.
  • Gordon suggested that we experiment with scanning a subject with head rotated at 45deg.
  • Also, if possible the scheme of gradient directions should be changed (need to talk to NanQuei)


DWI inspection for new data:

  • Gordon presented a method for inspecting sequences of DWIs to look for drops in signal or other artifacts. Baseline images and gradient images in a composite, cycling through slices to see abrupt changes in intensity.
  • With the new data, a lot of problems are obvious: zipper lines, dropped DWIs, uneven drops in intensity within a single DWI, motion artifacts, FAT artifact (shows up well in AvGradient image)
  • PNL can use this method to inspect every new DTI case that comes in
Retrieved from "https://www.na-mic.org/w/index.php?title=DBP:Harvard:Software:Minutes:8&oldid=4432"