Patient motion causes the same image artifacts or ghosting patterns as system instabilities cause in Magnetic Resonance images (MRI). Misinterpreting a patient-motion- induced artifact as a system instability can cause unnecessary system downtime and expense while a field engineer searches for a non-existent system problem. Although the images may look the same, the original sampled data is different and can be used to determine the cause of the image artifacts. The process for segmentation is to detect the existing instabilities and bulk motion vectors within the sampled frequency-space complex data. Because the imaged object is real, the frequency-space sampled data is Hermitian and redundant information exists. The two halves of the data set are compared and statistics extracted. Vectors are created representing the instability types: magnitude, phase, and echo shift, in addition to frequency encoding and phase encoding bulk motion. Each element of the vector represents how the phase encoding view differed from the complex conjugate view. The vector is then used to correct the data to remove the instability or motion. The order used for segmentation is: magnitude, echo shift, frequency encoding bulk motion, phase, phase encoding bulk motion. Bulk motion in the frequency encoding direction can be segmented from all types of system instability. Bulk motion in the phase encoding direction can be segmented from magnitude and echo shift instabilities, but not from phase instabilities. Other types of motion like scaling and rotation are problematic since the quantized nature of the sampled data precludes their characterization. The methods developed for isolating bulk motion from system instabilities are demonstrated on synthetic data.
Library of Congress Subject Headings
Magnetic resource imaging--Mathematics; Magnetic resource imaging--Effect of motion on; Magnetic resonance imaging--Quality control
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Easton, Roger Jr
Prentice, Wayne, "Segmentation of patient motion from other MRI system instabilities" (1997). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus