A STRATIFIED REGISTRATION FRAMEWORK FOR DSA ARTIFACT REDUCTION USING RANDOM WALKER

Sundarapandian, Manivannan and Kalpathi, Ramakrishnan (2016) A STRATIFIED REGISTRATION FRAMEWORK FOR DSA ARTIFACT REDUCTION USING RANDOM WALKER. In: 10th Indian Conference on Computer Vision, Graphics and Image Processing (ICVGIP), DEC 18-22, 2016, Indian Inst Technol, Guwahati, INDIA.

PDF 2016_10th_In_Co_Co_Vi_Gr_Im_Pr.pdf - Published Version Restricted to Registered users only Download (665kB) | Request a copy

Abstract

In Digital Subtraction Angiography (DSA), non-rigid registration of the mask and contrast images to reduce the motion artifacts is a challenging problem. In this paper, we have proposed a novel stratified registration framework for DSA artifact reduction. We use quad-trees to generate the non-uniform grid of control points and obtain the sub-pixel displacement offsets using Random Walker (RW). We have also proposed a sequencing logic for the control points and an incremental LU decomposition approach that enables reuse of the computations in the RW step. We have tested our approach using clinical data sets, and found that our registration framework has performed comparable to the graph-cuts (at the same partition level), in regions wherein 95% artifact reduction was achieved. The optimization step achieves a speed improvement of 4.2 times with respect to graph-cuts.

Item Type:	Conference Proceedings
Additional Information:	Copy right for this article belongs to the ASSOC COMPUTING MACHINERY, 1515 BROADWAY, NEW YORK, NY 10036-9998 USA
Department/Centre:	Division of Electrical Sciences > Electrical Engineering
Date Deposited:	15 Jul 2017 07:41
Last Modified:	15 Jul 2017 07:41
URI:	http://eprints.iisc.ac.in/id/eprint/57432

Actions (login required)

View Item