Limits...
Parameter-free binarization and skeletonization of fiber networks from confocal image stacks.

Krauss P, Metzner C, Lange J, Lang N, Fabry B - PLoS ONE (2012)

Bottom Line: The size and intensity pattern of the template is automatically adapted to the input data so that the method is scale invariant and generic.Furthermore, the template matching threshold is iteratively optimized to ensure that the final skeletonized network obeys a universal property of voxelized random line networks, namely, solid-phase voxels have most likely three solid-phase neighbors in a 3 x 3 x 3 neighborhood.This optimization criterion makes our method free of user-defined parameters and the output exceptionally robust against imaging noise.

View Article: PubMed Central - PubMed

Affiliation: Department of Physics, Biophysics Group, Friedrich-Alexander University, Erlangen, Germany.

ABSTRACT
We present a method to reconstruct a disordered network of thin biopolymers, such as collagen gels, from three-dimensional (3D) image stacks recorded with a confocal microscope. The method is based on a template matching algorithm that simultaneously performs a binarization and skeletonization of the network. The size and intensity pattern of the template is automatically adapted to the input data so that the method is scale invariant and generic. Furthermore, the template matching threshold is iteratively optimized to ensure that the final skeletonized network obeys a universal property of voxelized random line networks, namely, solid-phase voxels have most likely three solid-phase neighbors in a 3 x 3 x 3 neighborhood. This optimization criterion makes our method free of user-defined parameters and the output exceptionally robust against imaging noise.

Show MeSH

Related in: MedlinePlus

Matching template.Example for an automatically generated matching template in the x-z plane. The blurring is largest in the z direction.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3351466&req=5

pone-0036575-g006: Matching template.Example for an automatically generated matching template in the x-z plane. The blurring is largest in the z direction.

Mentions: In the following we focus on the scan in the x-direction. The x-scan can be imagined by a y-z-plane (the search plane) that moves through the 3D image stack. The algorithm detects sections of fibers with the search plane by comparing small 2D -regions of the search plane (the search sections) with a predefined template (Fig. 6) of the same size . The automatic generation of this template and the optimum choice of its size will be described below. The template, as well as the search sections, are represented as vectors with components that correspond to the intensities of the pixels. From all vector components, the global mean intensity of the 3D image stack is subtracted. Finally, all vectors are normalized to magnitude 1 to become independent from absolute intensities. To quantify the mismatch between search section and template we use the Euclidean distance of the corresponding vectors. If this distance exceeds a predefined mismatching threshold (for details see below), the central pixel of the corresponding search section is set to 0 (fluid phase). After this operation, the search plane contains in general numerous localized clusters of solid phase pixels, corresponding to cross sections of individual fibers. Within these clusters, local minima of mismatch (representing the medial axis of fiber cross sections) are determined and set to 1 (solid phase), while all others are set to 0. Analogous scans are performed through the y- and (in case of confocal fluorescence microscopy) z-directions. The binary image stacks resulting from each scan are combined using a logical OR-operation, yielding the final reconstruction result.


Parameter-free binarization and skeletonization of fiber networks from confocal image stacks.

Krauss P, Metzner C, Lange J, Lang N, Fabry B - PLoS ONE (2012)

Matching template.Example for an automatically generated matching template in the x-z plane. The blurring is largest in the z direction.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3351466&req=5

pone-0036575-g006: Matching template.Example for an automatically generated matching template in the x-z plane. The blurring is largest in the z direction.
Mentions: In the following we focus on the scan in the x-direction. The x-scan can be imagined by a y-z-plane (the search plane) that moves through the 3D image stack. The algorithm detects sections of fibers with the search plane by comparing small 2D -regions of the search plane (the search sections) with a predefined template (Fig. 6) of the same size . The automatic generation of this template and the optimum choice of its size will be described below. The template, as well as the search sections, are represented as vectors with components that correspond to the intensities of the pixels. From all vector components, the global mean intensity of the 3D image stack is subtracted. Finally, all vectors are normalized to magnitude 1 to become independent from absolute intensities. To quantify the mismatch between search section and template we use the Euclidean distance of the corresponding vectors. If this distance exceeds a predefined mismatching threshold (for details see below), the central pixel of the corresponding search section is set to 0 (fluid phase). After this operation, the search plane contains in general numerous localized clusters of solid phase pixels, corresponding to cross sections of individual fibers. Within these clusters, local minima of mismatch (representing the medial axis of fiber cross sections) are determined and set to 1 (solid phase), while all others are set to 0. Analogous scans are performed through the y- and (in case of confocal fluorescence microscopy) z-directions. The binary image stacks resulting from each scan are combined using a logical OR-operation, yielding the final reconstruction result.

Bottom Line: The size and intensity pattern of the template is automatically adapted to the input data so that the method is scale invariant and generic.Furthermore, the template matching threshold is iteratively optimized to ensure that the final skeletonized network obeys a universal property of voxelized random line networks, namely, solid-phase voxels have most likely three solid-phase neighbors in a 3 x 3 x 3 neighborhood.This optimization criterion makes our method free of user-defined parameters and the output exceptionally robust against imaging noise.

View Article: PubMed Central - PubMed

Affiliation: Department of Physics, Biophysics Group, Friedrich-Alexander University, Erlangen, Germany.

ABSTRACT
We present a method to reconstruct a disordered network of thin biopolymers, such as collagen gels, from three-dimensional (3D) image stacks recorded with a confocal microscope. The method is based on a template matching algorithm that simultaneously performs a binarization and skeletonization of the network. The size and intensity pattern of the template is automatically adapted to the input data so that the method is scale invariant and generic. Furthermore, the template matching threshold is iteratively optimized to ensure that the final skeletonized network obeys a universal property of voxelized random line networks, namely, solid-phase voxels have most likely three solid-phase neighbors in a 3 x 3 x 3 neighborhood. This optimization criterion makes our method free of user-defined parameters and the output exceptionally robust against imaging noise.

Show MeSH
Related in: MedlinePlus