Brain vascular image segmentation based on fuzzy local information C-means clustering

Chaoen Hu; Xia Liu; Xiao Liang; Hui Hui; Xin Yang; Jie Tian

doi:10.1117/12.2251725

16 February 2017 Brain vascular image segmentation based on fuzzy local information C-means clustering

Chaoen Hu, Xia Liu, Xiao Liang, Hui Hui, Xin Yang, Jie Tian

Proceedings Volume 10068, Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues XV; 100680Q (2017) https://doi.org/10.1117/12.2251725
Event: SPIE BiOS, 2017, San Francisco, California, United States

Abstract

Light sheet fluorescence microscopy (LSFM) is a powerful optical resolution fluorescence microscopy technique which enables to observe the mouse brain vascular network in cellular resolution. However, micro-vessel structures are intensity inhomogeneity in LSFM images, which make an inconvenience for extracting line structures. In this work, we developed a vascular image segmentation method by enhancing vessel details which should be useful for estimating statistics like micro-vessel density. Since the eigenvalues of hessian matrix and its sign describes different geometric structure in images, which enable to construct vascular similarity function and enhance line signals, the main idea of our method is to cluster the pixel values of the enhanced image. Our method contained three steps: 1) calculate the multiscale gradients and the differences between eigenvalues of Hessian matrix. 2) In order to generate the enhanced microvessels structures, a feed forward neural network was trained by 2.26 million pixels for dealing with the correlations between multi-scale gradients and the differences between eigenvalues. 3) The fuzzy local information c-means clustering (FLICM) was used to cluster the pixel values in enhance line signals. To verify the feasibility and effectiveness of this method, mouse brain vascular images have been acquired by a commercial light-sheet microscope in our lab. The experiment of the segmentation method showed that dice similarity coefficient can reach up to 85%. The results illustrated that our approach extracting line structures of blood vessels dramatically improves the vascular image and enable to accurately extract blood vessels in LSFM images.

Conference Presentation

Citation Download Citation

Chaoen Hu, Xia Liu, Xiao Liang, Hui Hui, Xin Yang, and Jie Tian "Brain vascular image segmentation based on fuzzy local information C-means clustering", Proc. SPIE 10068, Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues XV, 100680Q (16 February 2017); https://doi.org/10.1117/12.2251725

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available