Fluke 76-908 - Manual (Page 21)

Operation
Tests with Uniformity and Linearity (UAL) Phantom
2
2-11
(2) Non-linear magnetic field gradients
Variability is best observed over the largest field-of-view. The phantom should occupy at least 60% of the
largest field-of-view. Figure 1-1 provides an illustration of a pattern that is used to evaluate spatial
linearity.
Consideration should be given to determining the spatial linearity for a typical multi-slice acquisition with
the largest available image matrix to maximize spatial resolution. Since NMR imaging is inherently a
volumetric imaging technique, the evaluation should be performed for each orthogonal plane to define the
useful imaging volume. Spatial linearity is not expected to depend significantly on image timing
parameters such as TE, TR, and the number of signal acquisitions.
Percent distortion = (true dimension -observed dimension) / true) x 100%
Distortion measurement may be performed between any two points within the field-of-view, provided that
pixel-resolution is not a significant source of error. It is recommended that the true dimension be greater
than 10 pixels. Spatial linearity measurements performed directly on the image-processing unit will
provide information only about the MR imaging system. Measurements can also be performed upon
filmed images and will provide combined performance information about the MR imager, as well as the
video and filming systems.
Percent distortions in the linearity are generally acceptable if they are less than 5%.
2.4.3 Image Artifacts
Phase related errors are defined in terms of inappropriate (either increased or decreased) image signal at
specified spatial locations. Generally, these artifacts are characterized by increased signal intensity in
areas that are known to contain no signal producing material. Commonly called "ghosts," the application
of phase-encoding gradients for imaging and errors in both RF transmit and receive quadrature phase,
result in unique ghost artifacts. A "DC-offset" error is defined here as high-intensity or low-intensity pixels
at the center of the image matrix due to improper scaling of low-frequency components (typically DC) in
the Fourier transformation of the NMR time domain signal.
(1) Phase encoding gradient instability
(2) Quadrature phase maladjustment in the synthesis of slice selective RF pulses (transmit error)
(3) Improper quadrature phase decoding on receive
Any typical multi-slice sequence may be used. Separate scans must be made to assess both transmit and
receive errors if a phantom similar to the phantom in Figure 1-5 is used. More complex volume phantoms
may be designed which both transmit and receive errors and may be assessed with a single scan
sequence. The scan for assessing receive quadrature errors is made with the phantom placed at the
magnet isocenter with the central slice of the multi-slice sequence passing through the phantom. The
same scan may be used to assess both DC-offset and phase encoding errors. The scan for assessing
transmit quadrature errors is made with the phantom placed at a convenient offset slice position (typically
either + or - S cm from the isocenter slice) with the center slice passing through the magnet isocenter and
an offset slice passing through the phantom.
Phase Encoding Errors: Phase-encoding ghosts will appear as multiple images (possibly smeared into a
column) originating at the true object position but displaced along the phase-encoding axis of the image
(perpendicular to the frequency encoding direction). The presence of these characteristic ghost images
will generally identify the two axes; however, the orientations should be verified by the manufacturer or
the operator’s manual. Regions-of-interests values are taken from both the true image and the brightest
ghost image. The magnitude of the error (E) is quantified by expressing the ghost ROI value (G) as a
percent of the true ROI (T):
E = ((T-G)/T) x 100%