PET-CT

Question 1

What is PET-CT?

Answer 1

Synchronous functional and anatomical
imaging.
• PET – Functional Imaging
• CT – Anatomical Imaging

Question 2

Name 3 PET tracers.

Answer 2

• FDG - Glucose metabolism
• Choline, Methionine, Thymidine - proliferation
• Fluoride - bone metabolism
• Fluorothymidine (FLT) - Niche for tumour response
• Gallium 68 Dotatate/toc/noc - somatostatin receptor
• Fluorine-18 Florbetapir (Amyvid) - Alzheimers disease
• Gallium-68 DOTA ligands - Neuroendocrine tumours
• Rhubidium - myocardial uses
• Other: Eg Amyloid assessment (Florbetapir)

Question 3

What energy are the gamma rays given off in PET?

Answer 3

511 keV (pair emitted = better resolution than SPECT)

Question 4

What are the uses of PET-CT in radiotherapy?

Answer 4

• Staging of cancer
• Characterising lesions
• Finding source of infection/inflammation
• Assessing treatment response

Question 5

Name 2 types of cancer that are usually diagnosed using PET-CT.

Answer 5

• Lung
• Oesophageal and gastric cancer
• Lymphoma
• Head and Neck tumours
• Colorectal cancer
• Uterine and Cervical
• Sarcoma
• Melanoma
• Myeloma

Question 6

Why is FDG PET-CT used instead of CT when diagnosing lymphoma’s?

Answer 6

It is more sensitive and specific.

Question 7

What causes the increased uptake of FDG for activated inflammatory cells?

Answer 7

Increased expression of glucose transporters and Increased affinity of glucose transporters for deoxyglucose.

Question 8

Name a novel tracer in PET imaging.

Answer 8

Choline for prostate cancer.

Question 9

What is FDG used for in brain imaging?

Answer 9

Dementia or epilepsy diagnosis.

Question 10

In simple terms, explain the principles of PET.

Answer 10

A radioactive tracer is a positron emitting nuclide. The positron annihiliates with an electron which then emits 2 x 511 keV photons. The photons are emitted in opposing directions, and the lines of coincidence show where the annohilation occurred. The sum of the line of response is where the hot spot is.

Question 11

Define the Standard Uptake Volume in PET.

Answer 11

SUV = Activity Concentration / (Injected Dose / Body Weight)

where the injected dose is decay corrected, and body weight tends to be body mass, though lean body mass tends to be more accurate than body mass.

Question 12

Define SUVmax in PET and state an advantage and disadvantage.

Answer 12

SUVmax – simply uses the hottest pixel value within the lesion/region of interest.
This is the most common in routine use.
The plus side is that it is generally very robust between observers – so long as everyone’s ROI drawings include the hottest pixel it doesn’t matter how much they differ otherwise.
The down side is that in noisy images a single pixel value may be quite skewed by noise and also we are using a single voxel value to characterise much larger tissue volumes.

Question 13

Define SUVmean in PET and state an advantage and disadvantage.

Answer 13

SUVmean – this uses the mean value within the defined region/volume of interest (purple outline in this figure).
Deals very well with the problem of noise and should be a less biased estimate,
The region drawing is observer dependent and SUVmean values can be very variable as a result. Even when automated or semi-automated region defining algorithms are used the algorithm choice or any user inputs as starting points can make a significant difference to the result.

Question 14

Define SUVpeak in PET and state an advantage and disadvantage.

Answer 14

SUVpeak – proposed as a compromise solution and used in PERCIST criteria. This uses a small fixed volume and calculates the highest mean value of a volume this size within the ROI.
By using some averaging noise effects are suppressed, but the region is small enough that the overall region definition should have only a very limited effect. Note the PERCIST definition is for the highest uptake region, not just averaging around the max pixel, but software implementations are not always consistent with this and the actual definition used must be checked. SUVpeak can be calculated using different sizes for the small volume/region averaging – an appropriate choice will depend on the setting and needs to be standardised if results are to be comparable.

Question 15

What is the effect of an Incorrect cross-calibration between dose calibrator and scanner in SUV terms in PET imaging?

Answer 15

Systematic error in SUV equal to the relative cross-calibration

Question 16

What is the effect of unaccounted for residual activity in relation to SUV in PET imaging?

Answer 16

Lower net administration than has been used in the calculation, SUVs reduced.

Question 17

What is the effect of an incorrect decay correction in relation to SUV in PET imaging?

Answer 17

Incorrect SUV.

Question 18

What is the effect of a tissued injection in relation to SUV in PET imaging?

Answer 18

Incorrect SUV.

Question 19

What does SUV stand for in PET?

Answer 19

Standard uptake volume.

Question 20

What are the SUV technical considerations associated with scan acquisition parameters in PET?

Answer 20

Affects SNR. Upwards bias seen when SNR is low.

Question 21

What are the SUV technical considerations associated with image reconstruction parameters in PET?

Answer 21

Affects convergence. Partial volume effects are worse when convergence is insufficient.
If a smoother tumour is seen, the SUV is lower.

Question 22

What are the SUV technical considerations associated with ROI strategy parameters in PET?

Answer 22

Type and size of ROI can change SUVs in either direction.

Question 23

What are the SUV technical considerations associated with the normalisation factor in SUV equation & use of blood glucose correction factors in PET?

Answer 23

BSA, BM and LBM normalisations are not comparable. SUVs are reduced in the presence of high blood glucose unless corrected.

Question 24

What are the SUV technical considerations associated with contrast agents in PET?

Answer 24

SUV is reduced as some 511keV xrays are attenuated.

Question 25

What is the effect associated with a blood glucose level error in PET?

Answer 25

Lower uptake with increasing blood glucose.

NB: Blood glucose is an FDG specific error.

Question 26

What is the effect associated with uptake period error in PET?

Answer 26

Different point on uptake curve imaged = different SUV.

Question 27

What is the error associated with patient comfort in PET imaging?

Answer 27

FDG uptake is higher in muscle and fat when patient moves. This may reduce SUV elsewhere or cause spillover of counts in adjacent lesions.

Question 28

What is the error associated with inflammation in PET?

Answer 28

False positive increased SUV if in ROI.

Question 29

What is the effect associated with patient breathing/motion error?

Answer 29

Mismatched attenuation correct CT can cause artefacts and motion effects blur out counts potentially lowering SUV estimates.

Question 30

Explain PERCIST in PET imaging.

Answer 30

PERCIST – Positron Emission Tomography Response Criteria In Solid Tumours.
Attempt at systematic response assessment with FDG PET SUVs
Metabolic response precedes anatomical response to treatment
Uses SUV peak definition of region and lean body mass a normalisation – termed SUL
Baseline and follow-up lesions are the hottest by this SUL metric – not necessarily the same lesion (though usually) - looks for trends between scans.

Question 31

Name and explain the corrections required for calculating activity concentration for PET imaging.

Answer 31

Randoms Correction - two uncorrelated single events happen with such proximity to appear as a single true event. These can be estimated from the singles count rates.

Normalisation - compensates for variable sensitivities of each line of response (detector) in the scanner. a correction factor is given for each detector. Method: positron source to illuminate all possible LORs - Ge-68 used. Use a long period of time to acquire enough counts to get a good normalisation matrix. BUT dependent on good uniformity and source is not comparable to patient.

Dead Time Correction - PET system is made up of many sub systems which require a minimum time between each to register events properly. The effect is greater at higher activities - the fractional dead time is the ratio of measured to expected count rate. Method: use a decaying source, and extrapolate from very low count data (where dead time is assumed to be negligible). Use this data to create a look-up table.

Scatter Correction - Most likely interaction is Compton scatter; a photon can scatter up to 45 degrees and only lose 115keV to the recoil electron. PET scanners have wide energy windows (350-650keV) = poor energy resolution = scatter not excluded. Scatter is a significant correction in PET - thus it is required. There are many ways to measure: Method: Fitting scatter tails, Direct Measurement (only applicable to scanners with retractable septa), Multiple energy window techniques, Convolution and Deconvolution approaches, Analytical Simulation, Monte Carlo Simulation.

Attenuation Correction - a coincidence event requires simultaneous detector of both annihilation photons - the probability depends on the combined path. Wherever along a given LOR the source lies the number of detected coincidences is the same. Method: HU’s describing CT are used to estimate attenuation coefficients for 511keV photons. Implants and artefacts can cause issues.

Partial volume effect - Activity concentrations are underestimated in small volumes due to the finite spatial resolution and limited spatial sampling frequency. The total activity appears spread over a larger volume. This underestimates activity when quantifying and using a reduced contrast.

PET Scanner Calibration - For absolute activity concentrations, the calibration is usually done as a cross-calibration to ensure consistency of activity measurements for SUV calculation. If multiple calibrators are used, all must agree to within a tight tolerance or a one-to-one relationship maintained between scanner and calibrator.