Meeting Banner
Abstract #1528

Automated Segmentation of Liver Parenchyma and Blood Vessel with in-vivo Radial Gradient and Spin-Echo (GRASE) Datasets for Characterization of Diffuse Liver Disease

Abhishek Pandey1, 2, Ali Bilgin1, 3, Sindhu Cumar2, Bobby T. Kalb2, Diego R. Martin2, Maria I. Altbach2

1Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States; 2Medical Imaging, University of Arizona, Tucson, AZ, United States; 3Biomedical Engineering, University of Arizona, Tucson, AZ, United States

The analysis of imaging parameters in the liver has become of increased importance for the evaluation of pathologies such as fibrosis, inflammation and iron deposition. Although parametric imaging techniques have been developed, the analysis of parameters maps is mainly restricted to an ROI within the liver. Whole liver analysis should yield a better representation of the disease and if the analysis is automated it can be used routinely in the clinic. In this work, we present a combined liver and vessel segmentation technique that is used to evaluate full liver T2 and Fat Fraction maps in an automated fashion.

Keywords

abdominal acknowledgment acquisition although analyzed anatomical applied automated automatic avoid axial become best biomedical blood blue bobby boundaries boundary breath canny characterization clearly clinic clinical close closest cluster collection combined common component computer conditions consisted consultation contrast corrected created datasets deposition developed diffuse dilation disease dots echoes edge electrical engineering estimation evaluating evaluation every evolution experienced facilitate field final fitting flow fraction generate generated give gives gradient grant green highly histogram histograms hold homogeneities identify important in vivo included inflammation initial intensity iron limited liver mainly manual manually maps maria martin mask masks matrix medical merging methodology noise operation operator opposed parametric perfectly period phantom pixels plot process processed properties proposed providing radial radiologist reconstruct registered removal removed restricted routinely sample scanner scheme segmentation segmented series simulated since slice slices snake spin step steps strategy subjects surrounding table threshold tissue transferred truth unique various vector vessel vessels views vision water whole yield yields