Abstract

A major problem with non-rigid image registration techniques in many applications is their tendency to reduce the volume of contrast-enhancing structures [10]. Contrast enhancement is an intensity inconsistency, which is precisely what intensity-based registration algorithms are designed to minimize. Therefore, contrast-enhanced structures typically shrink substantially during registration, which affects the use of the resulting transformation for volumetric analysis, image subtraction, and multispectral classification. A common approach to address this problem is to constrain the deformation. In this paper we present a novel incompressibility constraint approach that is based on the Jacobian determinant of the deformation and can be computed rapidly. We apply our intensity-based non-rigid registration algorithm with this incompressibility constraint to two clinical applications (MR mammography, CT-DSA) and demonstrate that it produces high-quality deformations (as judged by visual assessment) while preserving the volume of contrast-enhanced structures.

Full Text
Published version (Free)

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 call