Related Experiment Videos
Decoupled automated rotational and translational registration for functional MRI time series data: the DART
L C Maas1, B D Frederick, P F Renshaw
1Brain Imaging Center, McLean Hospital, Belmont, Massachusetts 02178, USA.
Magnetic Resonance in Medicine
|January 1, 1997
Summary
This study introduces a fast, automated algorithm for correcting motion in medical images. The novel method accurately registers rotational and translational movements, significantly reducing artifacts in functional imaging data.
Area of Science:
- Medical Imaging
- Image Processing
- Computational Neuroscience
Background:
- Motion artifacts are a significant challenge in functional neuroimaging, particularly in echo-planar imaging (EPI).
- Previous image registration methods often involve iterative processes and can introduce image degradation.
- Accurate motion correction is crucial for reliable analysis of brain activity and reduction of spurious findings.
Purpose of the Study:
- To develop and validate a rapid, automated, in-plane image registration algorithm for motion correction in echo-planar imaging (EPI).
- To improve computational efficiency and reduce image degradation compared to existing methods.
- To demonstrate the effectiveness of the algorithm in correcting rotational and translational motion in phantom and human subject data.
Main Methods:
- A one-pass, non-iterative algorithm that decouples rotation and translation estimation using cross-correlation and cross-spectrum techniques.
- Utilizes k-space regridding and modulation for image correction, avoiding linear interpolation.
- Validated using simulated data, phantom imaging, and human functional imaging datasets.
Main Results:
- Achieved processing times of 7.5 s for 128x128 and 1.7 s for 64x64 images.
- Demonstrated high accuracy in phantom studies with minimal rotational (mean error -0.09°, SD 0.17°) and translational (mean error -0.035 pixels, SD 0.054) errors.
- Significantly reduced motion artifacts, including linear trends and stimulus-correlated motion artifacts, in human functional imaging data.
Conclusions:
- The developed algorithm offers a computationally efficient and accurate solution for in-plane motion correction in EPI.
- Its non-iterative, one-pass nature and reduced image degradation make it a valuable tool for neuroimaging analysis.
- The method effectively minimizes motion-related artifacts, enhancing the reliability of functional imaging results.