Fluke 76-908 - Manual (Page 13)

Operation
Tests with 3D Resolution and Slice (3DRAS) Phantom
2
2-3
resonance frequency adjustment and some may be completely automated. Resonance frequency should
be recorded daily for trend analysis.
Values of resonance frequency should generally not deviate by more than 50 ppm between successive
daily measurements.
2.3.2 Signal-To-Noise Ratio
The signal is defined as the mean pixel value within the region-of-interest minus any pixel offset. Noise is
defined as the random variations in pixel intensity. Images with obvious artifacts are not suitable for
signal-to-noise ratio (SNR) determinations.
Factors contributing to variations in signal-to-noise include:
(I) General system calibration (resonance frequency, flip angles)
(2) Slice thickness
(3) Coil tuning
(4) RF shielding
(5) Coil loading
(6) Image processing
(7) Scan parameters (TR, TE)
(8) T-1 and T-2 solution values
When using large volume fluid-filled phantoms, it should be recognized that thermal and mechanically
induced motions can introduce artifacts. The unloaded coil allows the evaluation of system noise that is
the parameter of interest. In a clinical scan, it is recognized that the patient is the dominant source of
noise. In order to approximate the clinical situation, the coil must be electrically loaded by using an
appropriate filler material.
The signal is measured using a ROI that contains at least 100 pixels or 10% of the area of the signal
producing material, whichever is greater. The ROI should be positioned in the center of the image and
should not include any obvious artifacts. The signal is the mean value of the pixel intensity in the ROI
minus any offset. (An indication of the existence of an image intensity offset may be gained from an
examination of intensity values from ROl's taken over non-signal producing portions of a phantom.
Specific offset values should be obtained from the system manufacturer). The noise is the standard
deviation derived from the same ROI. The signal-to-noise ratio is then calculated.
2.3.3 High-Contrast Spatial Resolution
High contrast spatial resolution is a measure of the capacity of an imaging system to show separation of
objects when there is no significant noise contribution. High contrast spatial resolution for MRI systems is
typically limited by pixel size (field-of-view divided by the sampling in x or y). Traditionally, resolution has
been quantified by the point spread function (PSF), line spread function (LSF), or modulation transfer
function (MTF); however, these methods are not practical for routine measurements. Therefore, a visual
evaluation of test objects is used.
Factors contributing to high-contrast resolution include:
• Field-of-view (determined by gradient strength, acquisition matrix, sampling period), and image
reconstruction and display method.
• The image will be evaluated visually. Image analysis consists of viewing the image to determine the
smallest resolvable hole array (magnification may be used if desired). For an array to be resolved, all
holes and spaces must be displayed as separate and distinct then viewed with the narrowest window
width. The window level should be adjusted for optimum visualization.