Changes between Version 21 and Version 22 of NIAID2022

Timestamp:

Sep 30, 2022, 12:49:30 PM (3 years ago)

Author:

Elaine Meng

Comment:

--

NIAID2022

@@ -64 +64 @@
 
 Zach has progress on DICOM, will put in daily build soon: 
-- Model panel shows hierarchy as Patient, Study, Image (series), and there will be a new DICOM browser and metadata browser (shows demo on his screen). Meghan and Phil: nice job!  Phil: do you anonymize the data?  Zach: we don't, so far have only dealt with data already anonymized.  
+- Model panel shows hierarchy as Patient, Study, Image (series), and there will be a new DICOM browser and metadata browser (shows demo on his screen). Meghan and Phil: nice job!  Phil: do you anonymize the data?  Zach: we don't, so far have only dealt with data already anonymized. Meghan: the DICOM browser is very helpful when multiple series are in the folder.  Eric: is that dialog shown in VR? TomG: yes.
+Elaine: maybe people would mostly use the DICOM browser to understand the data and decide on the series of interest before going into VR.  
 - Also have found and met with local collaborators who work with medical imaging and are interested in 3D applications.  Phil: I would be interested in further developments for segmentation in VR, are they interested in that?  TomG: one of the collaborators does segmentations and we will be viewing that in VR.  Collaborators are Jason Crane, Vanitha Sankaranarayanan; they showed Tom and Zach a video of liver segmentation by Jesse Courtier (at Mission Bay, also interested in collaboration): https://www.youtube.com/watch?v=Dy5h1482IzY
 - next steps are to polish up UI, meet with Vanitha et al., elicit feedback on interactions in VR
-- Eric mentioned our argument in group meeting earlier today about extracting all of DICOM functionality into a bundle or not; they see pros and cons, but Darrell agrees with me about keeping some DICOM functionality in the download for findability (which has implications for the broader issues of accessibility and pedagogical utility). TomG: this discussion came up because we would need a large package GDCM to read a specific format (lossless 16-bit JPEG that we cannot read currently)
+- Eric mentioned our argument in group meeting earlier today about extracting all of DICOM functionality into a bundle or not; they see pros and cons, but Darrell agrees with Elaine about keeping some DICOM functionality in the download for findability (which has implications for the broader issues of accessibility and pedagogical utility). Meghan: yes, having to explain two installations rather than one really increases the barrier to use in a classroom environment. TomG: this discussion came up because we would need a large package GDCM to read a specific format (lossless 16-bit JPEG that we cannot read currently)
 
 Phil: when your system was down, pipeline got stalled because it was trying to fetch presets from toolshed.  Eric: they should not be doing that every time it is used. They should install the wheel and only update it when an update is specifically available. Or, if they are creating a new virtual machine each time, they should cache the wheel locally and keep reusing it.  Eric: we also had a problem with slow startup when ChimeraX was trying to check the toolshed.  There is a ticket: #7659.  Greg: it only affects interactive use, however, not running in nogui mode.