Page 1 of 1
Changing the time slider
Posted: Tue Aug 13, 2019 4:28 pm
by justintso1199
Hi, I'm working with a set of 4D images (3D volume + cardiac phase-resolved). I'm looking to create a mesh for each frame in the cardiac cycle, and SimVascular allows visualization of this with the time slider in the image navigator. However, I've found that I can only plot paths or create segmentations when the time slider is at 0. If I change it to anything else, none of the points of the path show up. In addition, I've found that if I save the project with these images, close it, and reopen the project, the time slider no longer can move past 0. Would you know how to resolve this?
Re: Changing the time slider
Posted: Fri Aug 16, 2019 10:17 pm
by nwilson
Dear Justin,
Unfortunately, the most common use case of SimVascular is a single volume. The MITK core the GUI is built on supports visualizing multiple time points, but I don't think the SimVascular internals will properly handle it. The only current work around would be to save the multiple time points as independent volumes and work on them independently.
As for the image data visualization component, are you converting the image data to a .vti? If so, if you leave it as DICOM instead and don't convert to a vti you might be able to save a project and move the time slider between time points when you reload. That still probably will not fix the paths and segmentations issue, however.
Nathan
SimVascular Support Team
Re: Changing the time slider
Posted: Mon Aug 19, 2019 1:17 pm
by vvedula22
Hi Justin,
Based on your other post,
viewtopic.php?f=188&t=10689&sid=e75f93a ... 3cd5ca3e60
it appears that you are interested in segmenting heart chambers from time resolved 3D DICOM data. As Nathan pointed out, SimVascular doesn't yet process multiple image sets. For this, I'd recommend you to start by exporting 4D DICOM into 3D+time data sets, stored in separate folders for each cardiac phase. You can write a simple shell script for doing this based on the information from DICOM meta data. You may then segment heart chambers at any one cardiac phase, use image registration techniques to extract motion of the ventricles from image data at other cardiac phases and morph the initial segmentation. We have published the workflow for zebrafish data in PLOS Comp Bio journal
https://journals.plos.org/ploscompbiol/ ... bi.1005828
The registration tools are not part of SV yet but we plan to integrate them in future.
For segmenting the heart chambers, you could first use 2D (paths, groups, etc.) or 3D (fast marching 3D, region growing, etc.) segmentation tools in SV and use MeshMixer (free from AutoDesk) to process the segmentations as discussed in the paper above.
Vijay