Abstract
Rotational motion artifacts are notably prevalent in magnetic resonance (MR) images, such as of head and brain. Even a slight rotation during the data acquisition can have a significantly adverse effect on the quality and the diagnostic value of the image. This paper presents a post-processing technique for reducing the effects of rotation at a constant angular velocity. The rotation is constrained to be within the imaging plane. The correction algorithm is based on the separable nature of the 2D Fourier transform, and uses the symmetric properties of the motion model to effectively reduce the reconstruction time. A method of estimating the constant angular velocity from the corrupted data is also incorporated. The performance of the algorithm is verified using a simulated phantom and motion. It has also been demonstrated to be robust against the presence of noise. © 1998 Elsevier Science B.V. All rights reserved.
| Original language | English |
|---|---|
| Pages (from-to) | 103-114 |
| Journal | Signal Processing |
| Volume | 70 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - 30 Oct 1998 |
| Externally published | Yes |
Research Keywords
- Image reconstruction
- Magnetic resonance imaging
- Motion artifact correction
- Rotation at constant angular velocity
Fingerprint
Dive into the research topics of 'Correction of motion artifacts in MRI caused by rotations at constant angular velocity'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver