One major problem in using magnetic resonance imaging (MRI) is that its intensities are not standardized. This hinders different steps of quantitative image analysis such as segmentation or registration. More specifically, there is a lack of research done on the standardization of diffusion-weighted images (DWI). In this paper, we present a novel application of a whole body MRI intensity standardization method to DWI images. The approach uses the joint properties of DWI and T2-weighted with fat saturation (T2WI) images through their combined histogram. It is based on acquiring the deformation model between a reference and a current joint histogram through non-rigid registration. Since groups of histograms are used together, it is independent from the spatial information of the images. The evaluation of the method was accomplished by getting the distances between the reference and current histogram before and after registration. The registration resulted into a 60% reduction of the SSD distance between reference and current histogram.
[1]
Joachim Hornegger,et al.
Nonrigid Registration of Joint Histograms for Intensity Standardization in Magnetic Resonance Imaging
,
2009,
IEEE Transactions on Medical Imaging.
[2]
Patrick Clarysse,et al.
Correction of bias field in MR images using singularity function analysis
,
2005,
IEEE Transactions on Medical Imaging.
[3]
Bostjan Likar,et al.
A Review of Methods for Correction of Intensity Inhomogeneity in MRI
,
2007,
IEEE Transactions on Medical Imaging.
[4]
Jan Modersitzki,et al.
Numerical Methods for Image Registration
,
2004
.
[5]
G Adam,et al.
Diffusion-weighted MR imaging for differentiation of benign fracture edema and tumor infiltration of the vertebral body.
,
2001,
AJR. American journal of roentgenology.
[6]
Jayaram K. Udupa,et al.
New variants of a method of MRI scale standardization
,
2000,
IEEE Transactions on Medical Imaging.