Please enable JavaScript.
Coggle requires JavaScript to display documents.
SPECT QC (Monthly (Jaszczak Phantom (Used for validating scanner geometry,…
SPECT QC
Monthly
Jaszczak Phantom
Used for validating scanner geometry, 3D contrast, uniformity, resolution, attenuation and scatter correction or alignment tasks
It consists of a cylinder made of acrylic plastic with several inserts and has 6 solids spheres as well as 6 sets of 'cold' rods
-
-
Spheres are used to measure the image contrast while the rods are used to investigate the image resolution in SPECT systems
-
-
Pixel Size
QC done by calculation - reconstruction filters requires the pixel size data to be within 0.5 pixels
- Knowing the FOV and applying specific matrix size allows you to quantify the actual size of the pixel
- To verity this value, the following procedure is recommmended
a. Take a large enough piece of cardboard, draw two 10cm lines at right angles to each
b. Place the drops of Tc - 99m at each end of both lines and a third one where the lines intersect forming the right angle
-
-
e. Count the amount of pixels between the peak of the two points for X and Y. These measurements define the amount of pixels in 10cm
Gantry Tilt Check
To determine accuracy of tilt indicators and to ensure that specified tilt can be accomplished under clinical conditions
- Tape film centred to side of acrylic plate, with film edge carefully aligned to edge of plate
- Lower table to body scan position
- For systems which tilt the gantry or table in vertical plane: Align plate with long axis of table, parallel to vertical plane. Fix in position with tape extending from table edge to table edge across top of plate
- For tables tilting in horizontal plane: Place plate flat on tabletop with plate edge aligned to table long axis, tape into position
- Move tabletop into gantry opening: centre plate to alignment light
- Expose the film at inner light location using narrowest slice setting at ~ 50 to 100 mAs, 120 to 140 kVp
- Tilt gantry to one extreme from operators console: record angle indicated on gantry and operator's console, repeat exposure.
- Measure gantry clearance from closest point on gantry to table midline
- Tilt to opposite extreme, record indicated angles, repeat exposure and clearance measurement
Weekly
High-Count Flood
To reduce small non-uniformities produced by the detector collimator which produce artifacts (cold or hot spots) in reconstructed images
- Correction for non-uniformity is normally done by the process of normalisation, a simple mathematical procedure
- The mean number of counts pen pixel is calculated for the high-count flood image. The edge of the flood image is not used in this calculation to avoid the possibility of dividing by a small number on zero
- A set of correction factors is generated by dividing the mean counts pen pixel by the number of counts in each individual pixel
- A renormalisation is accomplished by multiplying the raw data in each projection by the correction matrix
- High count images for flood correction can be acquired with Co-57 disk sources or Tc-99m liquid-filled flood phantoms
Spatial Resolution
- Use orthogonal hole phantom, parallel-line equal space pattern or a four quadrant test pattern which is designed to be compatible with the manufacturer's software for analysis of linearity and resolution.
- Depending on the type of test pattern and software supplied, the pattern may be imaged intrinsically or extrinsically
- A Co-57 sheet source or uniformly filled flood tank should be used from extrinsic testing and only with a low-energy collimator
- A point source located 6-t times the max detector diameter from detector should be used for intrinsic testing
- The count rate produced by either configuration should be about 30k cps or as recommended by the manufacturer
- Visually inspect and note the smallest visible structure in the phantom image
Centre of Rotation
To correctly reconstruct a 3D image, it must be acquired in 3D format. If the 360 degree acquisition does not collect data in the same dimensional space, then artifacts occur
- Distance from the cource has no effect on the AOR acquisition
- If multiple point sources are required then usually a FOV is applied
- Number of stops is not that important, however 32 projections is recommended
- Usually 1 milliCurie of Tc - 99m is considered an adequate amount of activity
- Apply 128 matrix and collect between 5 to 20k counts per projection. This usually only takes less than 5 seconds a stop
- From this analysis, the x-axis offset should be less than 0.5 pixel and requires no correction. However, if 1 or more pixels then a correction is applied. Any offsets of greater than 3 pixel requires a service call
-
-
-
-