How to Interpolate Between Slices Using Gaussian Kernel - SlicerITKUltrasound Module

Support
1

Hi,

I am trying to interpolate between the slices of my 3D volume. The desired interpolation method must be Gaussian Kernel weighted average. Upon my research online, there are lots of sources on Gaussian Kernel filtering and smoothing, but that is not what I am trying to achieve. The aim of my project is to reconstruct the kidney by stacking 2D ultrasound images and interpolating between slices to create a denser volume. During my research, I found this module in 3D slicer that relates to ultrasound imaging and has my desired interpolation methods (ITKGaussian, VTKGaussianKernel).

However, when I try to resample the 3D volume I created using ImageStacks, there is no image/volume displayed. I tried playing around with the settings and did get something, but not the full image I was hoping for. Does anyone know how to troubleshoot the settings in this module to get the desired volume?

Step 1: Stack 57 .jpg ultrasound images into 3D volume using ImageStacks.
Step 2: Visualized volume using VolumeRendering.
Step 3: Applied resampling methods using ScanConvertCurvilinearArray.

Default settings
Default settings2560×1400 235 KB

Arbitrary settings
Arbitrary settings2560×1400 332 KB

2

If you use freehand ultrasound with tracking then you can use Volume Reconstruction module in SlicerIGT extension. You can use the “sticks” method for interpolation between slices.

If you have images acquired by a 4D ultrasound probe then you can use SlicerITKUltrasound extension for reconstructing 3D volumes. B-mode conversion is for creating B-mode voxel values from RF data, so you probably do not need this module. Scan conversion is for converting scanlines or slices into a 2D image fan or 3D volume. There are many variations for scan conversion (if your inputs are scanlines or slices, if your transducer slides, rotates, etc.), so I would recommend to check with ITK developers if the current modules support your data.