Scale-Space and Edge Detection Using Anisotropic Diffusion
IEEE Transactions on Pattern Analysis and Machine Intelligence
Building skeleton models via 3-D medial surface/axis thinning algorithms
CVGIP: Graphical Models and Image Processing
CVRMed-MRCAS '97 Proceedings of the First Joint Conference on Computer Vision, Virtual Reality and Robotics in Medicine and Medial Robotics and Computer-Assisted Surgery
An Experimental Comparison of Min-Cut/Max-Flow Algorithms for Energy Minimization in Vision
IEEE Transactions on Pattern Analysis and Machine Intelligence
What Metrics Can Be Approximated by Geo-Cuts, Or Global Optimization of Length/Area and Flux
ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1 - Volume 01
The use of a projection method to simplify portal and hepatic vein segmentation in liver anatomy
Computer Methods and Programs in Biomedicine
MICCAI'10 Proceedings of the 13th international conference on Medical image computing and computer-assisted intervention: Part III
Hi-index | 0.00 |
The segmentation and classification of the major intra-hepatic blood vessels along with the segmentation and analysis of hepatic tumors are critical for patient specific models of the diseased liver. Additionally, the accurate identification of liver anatomical segments can assist in the clinical assessment of the risks and benefits of hepatic interventions. We propose a novel 4D graph-based method to segment hepatic vasculature and tumors. The algorithm uses multi-phase CT images to model the differential enhancement of the liver structures and Hessian-based shape likelihoods to avoid the common pitfalls of graph cuts with undersegmentation and intensity heterogeneity. A hybrid classification step based on post-order walks of a graph identifies the right, middle and left hepatic, and portal veins. Veins are tracked using the graph representation and planes fitted to the vessel segments. The method allows the detection of all hepatic tumors and identification of the liver segments with 87.8% accuracy.