System matrix modelling of externally tracked motion.

Journal Article (Journal Article)

BACKGROUND AND AIM: In high-resolution emission tomography imaging, even small patient movements can considerably degrade image quality. The aim of this work was to develop a general approach to motion-corrected reconstruction of motion-contaminated data in the case of rigid motion (particularly brain imaging) which would be applicable to any PET scanner in the field, without specialized data-acquisition requirements. METHODS: Assuming the ability to externally track subject motion during scanning (e.g., using the Polaris camera), we proposed to incorporate the measured rigid motion information into the system matrix of the expectation maximization reconstruction algorithm. Furthermore, we noted and developed a framework to incorporate the additional effect of motion on modifying the attenuation factors. A new mathematical brain phantom was developed and used along with elaborate combined Simset/GATE simulations to compare the proposed framework with the cases of no motion correction. RESULTS AND CONCLUSION: Clear qualitative and quantitative improvements were observed when incorporating the proposed framework. The method is very practical to implement for any scanner in the field, not requiring any hardware modifications or access to the list-mode acquisition capability.

Full Text

Duke Authors

Cited Authors

  • Rahmim, A; Cheng, J-C; Dinelle, K; Shilov, M; Segars, WP; Rousset, OG; Tsui, BMW; Wong, DF; Sossi, V

Published Date

  • June 2008

Published In

Volume / Issue

  • 29 / 6

Start / End Page

  • 574 - 581

PubMed ID

  • 18458606

Pubmed Central ID

  • PMC2914313

International Standard Serial Number (ISSN)

  • 0143-3636

Digital Object Identifier (DOI)

  • 10.1097/MNM.0b013e3282f5d2de


  • eng

Conference Location

  • England