Refraction-compensated motion tracking of unrestrained small animals in positron emission tomography

Andre Kyme, Steven Meikle, Clive Baldock, Roger Fulton

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)


Motion-compensated radiotracer imaging of fully conscious rodents represents an important paradigm shift for preclinical investigations. In such studies, if motion tracking is performed through a transparent enclosure containing the awake animal, light refraction at the interface will introduce errors in stereo pose estimation. We have performed a thorough investigation of how this impacts the accuracy of pose estimates and the resulting motion correction, and developed an efficient method to predict and correct for refraction-based error. The refraction model underlying this study was validated using a state-of-the-art motion tracking system. Refraction-based error was shown to be dependent on tracking marker size, working distance, and interface thickness and tilt. Correcting for refraction error improved the spatial resolution and quantitative accuracy of motion-corrected positron emission tomography images. Since the methods are general, they may also be useful in other contexts where data are corrupted by refraction effects.
Original languageEnglish
Pages (from-to)1317-1328
Number of pages12
JournalMedical Image Analysis
Issue number6
Publication statusPublished - 2012
Externally publishedYes


  • Positron emission tomography
  • Refraction
  • Stereo motion tracking
  • Triangulation

Fingerprint Dive into the research topics of 'Refraction-compensated motion tracking of unrestrained small animals in positron emission tomography'. Together they form a unique fingerprint.

Cite this