Could be fixed by determining the total number of slices and start and end slice position, then calculating average slice spacing and interpolating the original image data accordingly. A bit messy. Good on 3D Slicer for detecting, alerting users and sorting it!
October 27, 2025 at 11:53 AM
Could be fixed by determining the total number of slices and start and end slice position, then calculating average slice spacing and interpolating the original image data accordingly. A bit messy. Good on 3D Slicer for detecting, alerting users and sorting it!
A closer read of this paper makes it seem that some of the scans were done in chunks of a few slices each, and that the chunks had different slice spacing from each other. That's a bit of a mess: making a composite assuming equal z-spacing throughout the volume will lead to distortions.
October 27, 2025 at 11:46 AM
A closer read of this paper makes it seem that some of the scans were done in chunks of a few slices each, and that the chunks had different slice spacing from each other. That's a bit of a mess: making a composite assuming equal z-spacing throughout the volume will lead to distortions.
The CPU implementation has a problem of thickness spheres overhanging the input mask by a pixel, introducing some inaccuracy. This GPU implementation has a more accurate Euclidean distance transform giving the sphere-fitting higher fidelity to the input geometry and less orthogonal grid bias.
October 13, 2025 at 4:30 PM
The CPU implementation has a problem of thickness spheres overhanging the input mask by a pixel, introducing some inaccuracy. This GPU implementation has a more accurate Euclidean distance transform giving the sphere-fitting higher fidelity to the input geometry and less orthogonal grid bias.
Small structures are also problematic as ever. But what's the thickness of a 1-pixel-wide line? The answer is much more meaningful for features represented by at least several pixels' thickness (what is several you may ask, well, try it and see...).
October 13, 2025 at 4:28 PM
Small structures are also problematic as ever. But what's the thickness of a 1-pixel-wide line? The answer is much more meaningful for features represented by at least several pixels' thickness (what is several you may ask, well, try it and see...).
Smaller simpler images now complete in a few seconds or less, depending on image size and feature size.
Unconstrained 'outside' is still challenging, but is now handled by treating it as spheres seeded from the image borders with a radius equal to the distance to the nearest feature in the image.
Unconstrained 'outside' is still challenging, but is now handled by treating it as spheres seeded from the image borders with a radius equal to the distance to the nearest feature in the image.
October 13, 2025 at 4:24 PM
Smaller simpler images now complete in a few seconds or less, depending on image size and feature size.
Unconstrained 'outside' is still challenging, but is now handled by treating it as spheres seeded from the image borders with a radius equal to the distance to the nearest feature in the image.
Unconstrained 'outside' is still challenging, but is now handled by treating it as spheres seeded from the image borders with a radius equal to the distance to the nearest feature in the image.
Large structures could make Tb.Sp in particular seem to run forever, or least longer than you would reasonably wait. 😴
Try it now. Big structures are still challenging, but should be done in like 1 hour instead of a week.
Try it now. Big structures are still challenging, but should be done in like 1 hour instead of a week.
October 13, 2025 at 4:22 PM
Large structures could make Tb.Sp in particular seem to run forever, or least longer than you would reasonably wait. 😴
Try it now. Big structures are still challenging, but should be done in like 1 hour instead of a week.
Try it now. Big structures are still challenging, but should be done in like 1 hour instead of a week.