PET QA Flashcards
What testing is required after installation and before clinical use?
- Critical Exam is completed first which is the duty of the installer but the testing may be delegated to Medical Physics
- Acceptance testing of the CT component
- Acceptance testing of the PET and scanner as a whole
- Commissioning and testing for specialist applications
What are the different PET Acceptance -NEMA tests?
- Spatial Resolution
- Sensitivity
- Scatter Fraction
- Count Losses and Random Coincidences
- Count Losses and Random Coincidence corrections
- Overall Image Quality
- Coincidence timing resolution in ToF scanners
- Co-registration accuracy in PET/CT scanners
How is the Spatial Resolution assessed for a PET system?
- Measure the FWHM FWTM of reconstructed point sources
- Radial and tangential resolution in transverse slice and axial resolution measured
- Three positions within the transverse plane tested
- Measurements performed in centre of detector bore (axial position=0 and at ¼ of axial field of view and the averaged result is reported
- FBP recon without smoothing
- Not intended to reflect clinical performance, but rather a’best case’ comparison between scanners
What sources are used for spatial resolution testing?
- F-18 point source of less
than 1mm x 1mm x 1mm ~200 MBq/mL - Droplets collected via capillary tube to form point source and sealed with plastacine to stop any leaks
- 3 sources required, @ 1 cm above isocentre, @ 10 cm above isocentre and @ 10 cm to the side of the isocentre
How is the Sensitivity assessed for a PET system?
- Repeat measurements of a line source with known absorbers, from which sensitivity in zero attenuation case may be extrapolated
- 5 metal tubes surrounding a plastic tube containing 700 ± 5mm of activity are used as the known absorber
- Positioning of the source must be consistent for each acquisition if extrapolation of results is to work
- Sensitivity is measured at both the centre and at 10 cm from the centre
How are the sensitivity results analysed?
- Decay correct each of the acquisitions and plot the total count rate versus the number of tubes along the x-axis
- Fit to R = R0 e(-µ_m x 2X_j) to find R0
- µ_m is the linear attenuation coefficinet of the material ut we dont include the known value because that is assuming an unscattered scenario, looking for a broad beam attenuation coefficient
- X_j is for the accumulating thickness ofattenuating material
- Plot axial sensitivity profile versus the slice number as not a point source
- Report both axial profile and R0
What test object configuration is used to assess Scatter Fraction, Count Loss and Random Coincidences & Corrections?
- Activity source is a line source of 700mm length positioned 45mm from the phantom centre
~900MBq at scan start time - Centre of phantom to align with centre of scanner bore
What data acquisition is used to assess Scatter Fraction, Count Loss and Random Coincidences & Corrections?
- Ideally repeat scanning more frequent than radionuclide half life with
minimum 500,000 prompt counts until true events losses <1% and
randoms rate <1% of trues rate* This isnt possible in LSO and LYSO detector systems due to self irradiation. - Therefore actual frequency of repeat and activities to be recommended by
manufacturer to assess peak count rate with adequate precision - Analysis for Scatter Fraction, Count Loss and Random Coincidences proceeds from an set of single slice rebinned sinograms
How is the corrections accuracy assessed?
Uses the same aquisition as for the count losses and random coincidences
Analysis based on reconstructed data with corrections applied
Correct count rates are extrapolated from those seen at lowest activity levels where correction are deemed negligible
Want to find out if the corrected count rates provide what was expected at higher activity levels..
How is image quality assessed?
- All spheres are ‘hot’ with 4:1
ratio of sphere content to
background - Phantom to be positioned so as
to centre the central lung insert
as closely as possible, without
raising the phantom from the
patient table surface
How is PET-CT Co-registratio accuracy assessed?
- A PET/CT alignment scan is performed to analyze the registration between the PET and the CT image.
- The VQC phantom consists of 5 point sources of 68Ge which are visible on both PET and CT images and are embedded in a moulded polyurethane foam.
- The coordinates of every point source on both PET and CT images is determined after reconstruction.
- The difference between the PET and CT coordinates along the 3 axes as well as the total distance between the PET and CT positions are calculated for each point source
Where can you find advice for routine PET QC?
- Manufacturer guidance
- IPEM Report 108
- EANM
- IAEA
What are the recommended daily checks?
- CT warm up and calibration
- PET uniformity scan with uniform phantom
- Look at block performance
- Create normalisation values
- Adjust PMT gains
