Mutual Information-Based CT-MR Brain Image Registration Using Generalized Partial Volume Joint Histogram Estimation

Hua Mei Chen, Pramod K. Varshney

Research output: Contribution to journalArticlepeer-review

161 Scopus citations


Mutual information (MI)-based image registration has been found to be quite effective in many medical imaging applications. To determine the MI between two images, the joint histogram of the two images is required. In the literature, linear interpolation and partial volume interpolation (PVI) are often used while estimating the joint histogram for registration purposes. It has been shown that joint histogram estimation through these two interpolation methods may introduce artifacts in the MI registration function that hamper the optimization process and influence the registration accuracy. In this paper, we present a new joint histogram estimation scheme called generalized partial volume estimation (GPVE). It turns out that the PVI method is a special case of the GPVE procedure. We have implemented our algorithm on the clinically obtained brain computed tomography and magnetic resonance image data furnished by Vanderbilt University. Our experimental results show that, by properly choosing the kernel functions, the GPVE algorithm significantly reduces the interpolation-induced artifacts and, in cases that the artifacts clearly affect registration accuracy, the registration accuracy is improved.

Original languageEnglish (US)
Pages (from-to)1111-1119
Number of pages9
JournalIEEE Transactions on Medical Imaging
Issue number9
StatePublished - Sep 2003


  • Image registration
  • Interpolation-induced artifacts
  • Joint histogram estimation
  • Mutual information
  • Registration of brain CT and MR images

ASJC Scopus subject areas

  • Software
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Electrical and Electronic Engineering


Dive into the research topics of 'Mutual Information-Based CT-MR Brain Image Registration Using Generalized Partial Volume Joint Histogram Estimation'. Together they form a unique fingerprint.

Cite this