Journal of Northeastern University Natural Science ›› 2017, Vol. 38 ›› Issue (10): 1383-1387.DOI: 10.12068/j.issn.1005-3026.2017.10.004

• Information & Control • Previous Articles     Next Articles

Segmentation Method of Base of Left Ventricle in Cardiac Magnetic Resonance Images

XU Li-sheng1,2, GUO Zeng-zhi1, TAN Wen-jun2, WANG Lu3   

  1. 1. School of Sino-Dutch Biomedical & Information Engineering, Northeastern University, Shenyang 110169, China; 2. Key Laboratory of Medical Image Computing, Ministry of Education, Northeastern University, Shenyang 110169, China; 3. School of Computer Science & Engineering, Northeastern University, Shenyang 110169, China.
  • Received:2016-04-15 Revised:2016-04-15 Online:2017-10-15 Published:2017-10-13
  • Contact: XU Li-sheng
  • About author:-
  • Supported by:
    -

Abstract: A novel method was proposed for segmenting the base of the left ventricle in cardiac magnetic resonance imaging (MRI) images based on local intensity clustering (LIC) model and watershed algorithm. First, the cardiac MRI images were segmented by LIC model to detect the tissues and organs. Then, the connected tissues and organs were separated by using watershed algorithm to make up for the missing edges. The seed points were artificially selected to carry growing for the preliminary extraction of left ventricle. Finally, whether the preliminary extraction of the left ventricle contains the aorta will be judged by priori knowledge of the shape features of the left ventricle, if the preliminary extraction of the left ventricle contain the aorta, the effect of the missing edge caused by the aorta will be removed to get an accurate segmentation result of the base of left ventricular. Experimental results demonstrated that the proposed method can effectively remove the effect of weak edges and edge leakage of the base of the left ventricle in MRI images to obtain an accurate segmentation result of the base of left ventricular.

Key words: cardiac magnetic resonance images, left ventricle, image segmentation method, LIC model, watershed algorithm

CLC Number: