Optimization of 3-D Active Appearance Models Using the Inverse Compositional Approach: Applications to Cardiac MRI Segmentation
Alexander Andreopoulos and John K. Tsotsos
Technical Report CS-2004-08
York University
October 28, 2004
Abstract
We present an efficient algorithm for the fitting of three dimensional (3-D) Active Appearance Models (AAMs) on short axis cardiac MRI, using an extension of the inverse compositional image alignment algorithm that was recently introduced by Matthews and Baker [8]. We demonstrate its applicability for the segmentation of the left ventricle in short axis cardiac MR images. We perform experiments to evaluate the speed and segmentation accuracy of our algorithm on a total of 477 MR images, acquired from 11 patients. We observe a 60 fold increase in fitting speed compared to a brute force Gauss-Newton optimization and a segmentation accuracy which agrees well with an independent standard. We conclude that this is an efficient and robust algorithm for left ventricle segmentation using 3-D Active Appearance Models.
Download paper in PDF format.
The documents distributed by this server have been provided by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.