ROI Localization and Initialization Method for Left Ventricle Segmentation

This paper proposed a ROI localization method from cardiac short axis MR images and an initialization strategy for LV segmentation methods. To localize the ROI, we firstly apply cumulative value of the difference between adjacent temporalsequences during the whole cardiac cycle to estimate the motion region of the LV, namely the ROI we tend to locate. After localizing the ROI of the LV, we use Circle Hough Transform (CHT) to detect circle region in the ROI, basing on the shape characteristic of the LV. Then, the detected circle is used for initial contour of LV segmentation methods. Experimental results demonstrate that our proposed method is effective for ROI localizationand initialization ofLV segmentation methods.

[1]  Yi Su,et al.  Automatic segmentation of left ventricle cavity from short-axis cardiac magnetic resonance images , 2017, Medical & Biological Engineering & Computing.

[2]  Yue Zhao,et al.  A novel level set method for segmentation of left and right ventricles from cardiac MR images , 2014, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[3]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[4]  Olivier Ecabert,et al.  Adaptive Hough transform for the detection of natural shapes under weak affine transformations , 2004, Pattern Recognit. Lett..

[5]  Thi-Thao Tran,et al.  Active contour model and nonlinear shape priors with application to left ventricle segmentation in cardiac MR images , 2016 .

[6]  James C Moon,et al.  Distance regularized two level sets for segmentation of left and right ventricles from cine-MRI. , 2016, Magnetic resonance imaging.

[7]  Yue Zhao,et al.  Edge Based Segmentation of Left and Right Ventricles Using Two Distance Regularized Level Sets , 2015, ISVC.

[8]  Muhammad Sharif,et al.  Automatic segmentation of the left ventricle in a cardiac MR short axis image using blind morphological operation , 2018, The European Physical Journal Plus.

[9]  Shaoxiang Zhang,et al.  Simultaneous extraction of endocardial and epicardial contours of the left ventricle by distance regularized level sets. , 2016, Medical physics.

[10]  Chunming Li,et al.  Active contours driven by local Gaussian distribution fitting energy , 2009, Signal Process..

[11]  Maria A. Zuluaga,et al.  Automatic segmentation of right ventricle in cardiac cine MR images using a saliency analysis. , 2016, Medical physics.

[12]  Gustavo Carneiro,et al.  Fully Automated Non-rigid Segmentation with Distance Regularized Level Set Evolution Initialized and Constrained by Deep-Structured Inference , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[13]  Caroline Petitjean,et al.  A review of segmentation methods in short axis cardiac MR images , 2011, Medical Image Anal..

[14]  Yuanqi Su,et al.  Left ventricle segmentation via two-layer level sets with circular shape constraint. , 2017, Magnetic resonance imaging.