Tubeness is an algorithm for enhancing filamentous structures of a specified thickness (blood vessels, neurites, and other tube-like structures)described by Sato et al 1998.1. It can be applied to 2D and 3D images. It can be a faster alternative to Frangi. The algorithm is explained in more detail here. There are three Tubeness implementations in ImageJ:
-
Legacy plugin (In Fiji registered under Plugins › Analyze › Tubeness (source code). This is now deprecated.
-
An ImageJ Op. This is a modernized, multithreaded version of the initial implemenation that has been validated/benchmarked more carefully (details)
-
An SNT implementation (source code) that extends the ImageJ Op approach to multiple scales (similarly to Frangi) for better detection of neurites. SNT’s Secondary Layer Wizard can also be used to preview the effect of scale size in the result.
Tubeness: Left: Original image. Right: Filtered image.
