Directional histogram ratio at random probes: A local thresholding criterion for capillary images

Abstract

With the development of micron-scale imaging techniques, capillaries can be conveniently visualized using methods such as two-photon and whole mount microscopy. However, the presence of background staining, leaky vessels and the diffusion of small fluorescent molecules can lead to significant complexity in image analysis and loss of information necessary to accurately quantify vascular metrics. One solution to this problem is the development of accurate thresholding algorithms that reliably distinguish blood vessels from surrounding tissue. Although various thresholding algorithms have been proposed, our results suggest that without appropriate pre- or post-processing, the existing approaches may fail to obtain satisfactory results for capillary images that include areas of contamination. In this study, we propose a novel local thresholding algorithm, called directional histogram ratio at random probes (DHR-RP). This method explicitly considers the geometric features of tube-like objects in conducting image binarization, and has a reliable performance in distinguishing small vessels from either clean or contaminated background. Experimental and simulation studies suggest that our DHR-RP algorithm is superior over existing thresholding methods.

Clustering Tree-Structured Data on Manifold

“…The input data are vascular tree images. Several pre-processing steps (e.g., denoising, binarizing, and gap filling [46,47]) are applied to these images before the medial axis thinning algorithm [48] is employed to extract the vessel tree skeleton. The skeletonization results enable us to extract tree branches by detecting the junctions and end points in the skeleton using the pixel connection property.…”

“…T-A matrices in reality can be generated from, e.g., angiographic images after a number of pre-processing steps such as segmentation [46] and skeletonization [51]. However, we start from the simplest case by directly generating the T-A matrices so the errors introduced by image pre-processing can be ruled out when evaluating the algorithm performance.…”

Capillary extraction by detecting polarity in circular profiles

“…While such remarkable technical improvements enable us to gain new insights into vascular architecture, quantitative characterisation of vessel image data presents a unique challenge. It has also been recognised that the reliability of such characterisation heavily depends on the outcome of image preprocessing methods such as thresholding [5]. Therefore, the development of efficient preprocessing algorithms is of significant interest.…”

“…A variety of approaches have been developed for vascular image thresholding or extraction [5, 9]. Methods for vessel extraction can be divided into three categories based on the information employed: intensity‐based approaches (e.g.…”

“…[7, 12, 13]), and multi‐cue methods that incorporate both geometric features and intensity (e.g. [5, 14]). The intensity‐based approaches mainly explore the local statistics of pixel intensity [8, 14, 15].…”