Unsupervised Correction of Respiratory-induced Motion in 4D Short-axis Cardiac Cine MRI |
Mikkel B. Stegmann, Dorthe Pedersen, Henrik B. W. Larsson
|
| Abstract | Purpose: To rapidly and automatically correct for respiratory-induced motion in dynamic three-dimensional cardiac magnetic resonance images acquired using breath-hold.
Methods: Statistical models are employed to localise the left ventricle in a set of short-axis images. Subsequently, translation estimates are derived and employed for correction of respiratory motion.
Results: Qualitatively and quantitatively validation was carried out on 12 subjects. Slice displacements were significantly reduced (p<0.001). Mean slice displacement before and after correction: 3.40+/-1.44 mm and 1.60+/-0.34 mm. Estimation and correction took <10 seconds per subject.
Conclusion: The presented method provides unsupervised, rapid and purely image-based correction of respiratory-induced motion in short-axis cardiac MRI. |
| Type | Journal paper [Submitted] |
| Journal | Journal of Cardiovascular Magnetic Resonance (submitted) |
| Year | 2006 Month March |
| Publisher | Taylor & Francis |
| BibTeX data | [bibtex] |
| IMM Group(s) | Image Analysis & Computer Graphics |