Abstract
Gradient-oriented profiles are presented as a novel method for boundary parameterization and unsupervised boundary classification. Each profile is created at locations of high gradient magnitude by sampling an ellipsoidal neighborhood of voxels oriented along the image gradient. Every profile is analyzed via nonlinear optimization to fit the best cumulative Gaussian, directly parameterizing the boundary to yield estimates of (1) extrapolated intensity values for voxels located far inside and outside of the boundary and (2) boundary location and width. For these parameter estimates, intrinsic measures of confidence are established to eliminate the low-confidence parameters. Gradient-oriented profiles are demonstrated on artificially generated three-dimensional test data and shown to accurately parameterize and classify the boundary, establishing them as a high-quality replacement for simpler methods of boundary detection. Towards shape analysis gradient-oriented profiles are applied to medial primitives known as core atoms. The additional intensity information delivered by gradient-oriented profiles improves the function of core atoms by eliminating dependence on the absolute direction of the local image gradient and the background intensity. The performance of core profiles is demonstrated on test data simulating a three-dimensional ultrasound scan of the heart improving the previous capabilities of core atoms.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
