Live-Vessel: Extending Livewire for Simultaneous Extraction of Optimal Medial and Boundary Paths in Vascular Images

Abstract

Abstract. This paper incorporates multiscale vesselness filtering into the Livewire framework to simultaneously compute optimal medial axes and boundaries in vascular images. To this end, we extend the existing 2D graph search to 3D space to optimize not only for spatial variables (x, y), but also for radius values r at each node. In addition, we minimize change for both scale and the smallest principle curvature and incorporate vessel boundary evidence in our optimization. When compared to two sets of DRIVE expert manual tracings, our proposed technique reduced the overall segmentation task time by 68.2%, had a similarity ratio of 0.772 (0.775 between manual), and was 98.2% reproducible.

Automatic Extraction of Mandibular Nerve and Bone from Cone-Beam CT Data

“…In this work we chose a conceptually simple approach for nerve detection, yet other methods for tracing tubular structures may be considered, e.g. as described for vessel detection, see for instance [12,13] and references therein. In future work, the statistical shape model shall be extended to distinguish cortical and trabecular bone as well.…”

VascuSynth: Simulating vascular trees for generating volumetric image data with ground-truth segmentation and tree analysis

“…r le f t is then obtained by multiplying f le f t by r per f in (10). Note that the ratio r right /r le f t must be known in order to calculate f le f t (and similarly for f right ).…”

“…This strong need has been the catalyst for algorithms for segmentation and analysis of vasculature. This has been reflected by the large number of proposed algorithms for enhancement, segmentation, and analysis of vasculature and perfusion data from 3D medical images in the past few decades [3,4,5,6,7,8,9,10,11,12]. However, what has been lacking is the existence of benchmarks for the validation, evaluation, and comparison of all these approaches.…”

Extraction of tubular structures over an orientation domain

“…The notion of shortest path has proved to be efficient in order to perform global segmentation of tubular structures, in images, see for instance [5,16]. These geodesic curves can also be used to extract tubular structure centerlines, as proposed by Deschamps and Cohen [6] and by Santamaria et al [19].…”