search

PlusToolkit / PlusLib

Software library for data acquisition, pre-processing, and calibration for navigated image-guided interventions.
http://www.plustoolkit.org
Other
136 stars 103 forks source link

Reverse Volume Reconstruction - Going from reconstructed US volume to 2D images #1005

Closed prash-p closed 12 months ago

prash-p commented 2 years ago

Is it possible to 'extract'/sample the 2D location-tagged ultrasound frames from a reconstructed US volume, given the reconstructed volume and the time-stamped ImageToReference transforms? Is this possible using Slicer/SlicerIGT? I understand that the sampled frames would not be the original resolution that US was captured in.

ungi commented 2 years ago

If your 3D reconstruction is in the Reference coordinate system, and you have a series of ImageToReference transforms, then you use the vtkImageReslice filter in Slicer's python interactor to obtain a cross section of the volume where Image would be. It should be possible to write a Python script that extracts the cross sections using a sequence of transforms. As you said, there will be loss of information due to resolution, compounding, tracking errors, etc. so it's not an ideal solution. If you decide to use vtkImageReslice, here is an example code to get started: https://gist.github.com/ungi/80442ac03b5ea9ed6ae680a800837ece

prash-p commented 2 years ago

Thank you Tamas! The application is for extracting segmentation masks from a CT scan registered to the tracked US :)