On the High Pass Filtering Properties of the Intensity-Curvature Functional: Applications in Magnetic Resonance Imaging of the Human Brain

Web proceedings papers

Authors

Carlo Ciulla , Ustijana Rechkoska Shikoska , Filip A. Risteski and Dimitar Veljanovski

Abstract

This research presents the use of the intensity-curvature functional (ICF) as k-space filter. The k-space magnitude of the following four image space filters was calculated: two high pass filters (termed as HPF and PSOHPF) and two gradients (termed as GRADX and GRADY). Two dimensional images of the human brain were collected with Magnetic Resonance Imaging (MRI) and they were processed using the following k-space filtering technique. The images were fitted with the bivariate linear model function. The real and the imaginary parts of the k-space of the ICF were subtracted from the k-space of the filters and the k-space of the gradients. The ensuing k-space was inverse Fourier transformed so to obtain four k-space filtered images. The major contribution of this paper is thus the comparison between four k-space filtering techniques. Data confirms the aptitude of the ICF to behave as high pass filter in image space. Moreover, the k-space filtered images present a level of demarcation of the boundaries of the vessels which is slightly superior to the details observable in the image space filters. In conclusion, vessels demarcation in MRI can be obtained using the ICF as k-space filter

Keywords

K-space · Image space filters · K-space filtered images · Intensitycurvature functional ·Demarcation of the boundary · High pass filter.