Image fusion between whole body FDG PET images and whole body MRI images using a full-automatic mutual information-based multimodality image registration software

Yoshitaka Uchida, T. Kazama, H. Ito, Y. Nakano, T. Fujibuchi, T. Isobe

Research output: Contribution to journalArticle


We attempted image fusion between whole body PET and whole body MRI of thirty patients using a full-automatic mutual information (MI) -based multimodality image registration software and evaluated accuracy of this method and impact of the coregistrated imaging on diagnostic accuracy. For 25 of 30 fused images in body area, translating gaps were within 6mm in all axes and rotating gaps were within 2degrees around all axes. In head and neck area, considerably much gaps caused by difference of head inclination at imaging occurred in 16 patients, however these gaps were able to decrease by fused separately. In 6 patients, diagnostic accuracy using PET/MRI fused images was superior compared by PET image alone. This work shows that whole body FDG PET images and whole body MRI images can be automatically fused using MI-based multimodality image registration software accurately and this technique can add useful information when evaluating FDG PET images.

Original languageEnglish
Pages (from-to)1743-1750
Number of pages8
JournalJapanese Journal of Clinical Radiology
Issue number12
Publication statusPublished - Dec 12 2006
Externally publishedYes


All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging

Cite this