A fully automatic cardiac segmentation method using region growing technique

Automatic images segmentation is a crucial phase from image processing to image analysis; it is a subset of an expansive field of Computer Vision. In the medical field, this operation allows achieving efficient analysis of the contractile heart function, which is the main criterion used to determine the prognostic of different cardiopathies. This paper presents a fully automatic cardiac segmentation of the right ventricle (RV) based on region growing (RG) technique, which can segment the area of the RV from the basal to the apical slices at End-systolic (ES) and Enddiastolic (ED) phases. In an initial step, an adaptive histogram equalization (AHE) has been introduced as a pretreatment to improve contrast and broaden the regions having a bad distribution in image. In the next steps, the proposed method uses two algorithms in order to automate the segmentation process. The first one is the generalized Hough transform technique (GHT) applied to select the initial seed pixel, so that the RG can start growing, and the second one is an iterative threshold selection algorithm to compute the optimal threshold value for the best segmentation. The experiment results show that the proposed method is efficient in both selecting seed point and segmenting region of the RV.

[1]  Ahmed S. Fahmy,et al.  Active Shape Model with Inter-profile Modeling Paradigm for Cardiac Right Ventricle Segmentation , 2012, MICCAI.

[2]  Tao Jianhua,et al.  A Fast Implementation of Adaptive Histogram Equalization , 2006, 2006 8th international Conference on Signal Processing.

[3]  Daoudi Abdelaziz,et al.  Automatic Segmentation of the Right Ventricle by Active Shape Model and a Distance Transform , 2015 .

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

[5]  Yi Wang,et al.  Left Ventricle: Fully Automated Segmentation Based on Spatiotemporal Continuity and Myocardium Information in Cine Cardiac Magnetic Resonance Imaging (LV-FAST) , 2015, BioMed research international.

[6]  Rüdiger Dillmann,et al.  Segmentation of the left and right cardiac ventricle using a combined bi-temporal statistical model , 2006, SPIE Medical Imaging.

[7]  Tai-Hoon Cho Object Matching Using Generalized Hough Transform and Chamfer Matching , 2006, PRICAI.

[8]  Douglas G. Altman,et al.  Measurement in Medicine: The Analysis of Method Comparison Studies , 1983 .

[9]  Cristian Lorenz,et al.  A Hybrid Method for Automatic Anatomical Variant Detection and Segmentation , 2011, FIMH.

[10]  John K. Tsotsos,et al.  Efficient and generalizable statistical models of shape and appearance for analysis of cardiac MRI , 2008, Medical Image Anal..

[11]  Pierre Croisille,et al.  Quantification of Right and Left Ventricular Function in Cardiac MR Imaging: Comparison of Semiautomatic and Manual Segmentation Algorithms , 2013, Diagnostics.

[12]  Jianhua Tao,et al.  A Fast Implementation of Adaptive Histogram Equalization , 2006 .

[13]  Dana H. Ballard,et al.  Generalizing the Hough transform to detect arbitrary shapes , 1981, Pattern Recognit..

[14]  Salem Saleh Al-amri,et al.  Image Segmentation by Using Threshold Techniques , 2010, ArXiv.

[15]  Antoine Manzanera,et al.  A coronary artery segmentation method based on multiscale analysis and region growing , 2016, Comput. Medical Imaging Graph..

[16]  Fredrik Orderud,et al.  Real-Time Active Shape Models for Segmentation of 3D Cardiac Ultrasound , 2007, CAIP.