Uniformity testing of gamma cameras at high count-rates NICK BATES UNIVERSITY OF SURREY ROYAL SURREY COUNTY HOSPITAL NATIONAL PHYSICAL LABORATORY UK
Uniformity testing of gamma cameras at high count-ratesNICK BATESUNIVERSITY OF SURREYROYAL SURREY COUNTY HOSPITALNATIONAL PHYSICAL LABORATORYUK
Background• Imaging patients after therapeutic
administration of radiopharmaceutical causes significant dead-time effects in gamma cameras
• Gamma cameras behave as paralysabledetectors so the count-rate can saturate at high activities
• To deal with this, manufacturers have implemented high count-rate modes. Some switch in automatically, others operate separately from “normal mode”
• GE SPECT/CT cameras have a “fast mode” setting that must be set-up for each radionuclide
Input Count Rate
Ou
tpu
t C
ou
nt R
ate
Gamma camera
Background• Prior to carrying out patient imaging, appropriate
Quality Assurance should be completed.• Flood-field uniformity is a standard test of overall
camera performance• Response of the detector to a uniform flux of radiation• Ideally equal image intensity over entire field of view• Non-uniformity indicates detector drift / inaccurate
corrections
• The stability of uniformity with time and activity in “fast mode” is unknown…
• As part of the set-up for the SELIMETRY study, RSCH has investigated the dependence of “fast mode” uniformity on source activity and the stability over time
Initial Set-up and Testing• GE engineers enabled “fast mode” after preliminary testing revealed
paralysis of detector in “normal mode” at ~1.5GBq source activity (~25kcps)
• A 60M count uniformity correction map was acquired using a 500MBq point source of I-131 positioned >5xUFOV distance from the uncollimated detector head.• This activity was chosen as it gave a count-rate of ~70kcps • For Tc-99m fast mode, GE recommend acquiring calibrations at 45-
100kcps, so we assumed the same would apply for I-131• Uniformity, energy peak position and FWHM was assessed by
acquiring 20M count flood-field with sources ranging from 50MBq (8kcps) to 700MBq (94kcps).
Initial Uniformity Results
y = -0.1675x + 13.844R² = 0.994
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
0 20 40 60 80 100
CF
OV
In
teg
ral
Un
ifo
rmit
y (
%)
Count rate (kcps)
y = 9E-05x2 + 0.0086x + 8.5638R² = 0.9988
8.4
8.6
8.8
9.0
9.2
9.4
9.6
9.8
10.0
10.2
10.4
360.9
361.0
361.1
361.2
361.3
361.4
361.5
361.6
361.7
361.8
0 20 40 60 80 100
Peak F
WH
M (
%)
Peak P
osit
ion
(keV
)Count Rate (kcps)
Initial Uniformity Results• Uniformity degrades linearly either side of the count-rate used for calibrations.• Uniformity becomes clinically unacceptable (>3%) within 10kcps of calibration
count-rate• Peak position does not change significantly with count-rate• Peak FWHM increased with count rate (8.6% at 8kcps to 10.2% at 94kcps)
Repeat testing• Suggestion from other hospitals (Oxford) to acquire uniformity map at lower
count-rate• ~140MBq source used ~20kcps• 60M count uniformity correction map as before• Prepared 7 sources (10, 20, 50, 100, 140, 400, 700MBq)
• Acquired individually and in combination to study dependence of uniformity on count-rate in more detail
Repeat Testing Results: Uniformity
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140
Un
ifo
rmit
y (%
)
Count Rate (kcps)
CFOV Int
CFOV Diff
UFOV Int
UFOV Diff
Repeat Testing Results: Uniformity
y = -0.0001x3 + 0.014x2 - 0.405x + 6.1714R² = 0.9797
0
2
4
6
8
10
12
0 10 20 30 40 50 60 70
Un
ifo
rmit
y (%
)
Count Rate (kcps)
CFOV Int
Polyn. (CFOV Int)
Repeat Testing Results: Peaking
360
360.5
361
361.5
362
362.5
363
363.5
0 50 100 150
Ph
oto
pe
ak p
osi
tio
n (
keV
)
Count Rate (kcps)
y = 0.0002x2 + 0.004x + 8.6525R² = 0.9936
8
8.5
9
9.5
10
10.5
11
11.5
12
12.5
0 50 100 150
FWH
M (
%)
Count Rate (kcps)
Observable Changes
Centres of PMTs have lower counts at higher count-rates
Increased count rate
20kcps (140MBq) 100kcps (890MBq)
Observable Changes
• Event pile-up towards centre of PMTs at very high rates
• Streaking between centres of PMTs – more prominent horizontally
Observable Changes20kcps (140MBq)
Decreased count rate
1.5kcps (10MBq)
Centres of PMTs have higher counts at lower count-rates
Possible causes• Events outside energy window contribute to
dead-time• Count-rate dependent spatial linearity
distortions?• Amplifier saturation?• GE factory response:
• Significant scatter downward shift in energy peak…?
• Tests have been conducted with increased event threshold; no effect
• Suggesting adding copper behind source to reduce scatter
• What about patient scatter?!
Stability with time
2
2.2
2.4
2.6
2.8
3
3.2
3.4
0 5 10 15 20 25 30CF
OV
In
teg
ral
Un
ifo
rmit
y (
%)
Days since calibration
Measured Predicted based on count-rate
• Possibly random variation versus predicted values?
• No signficant trend in uniformity seen over 1 month time-scale
• Longer-term stability still to be tested
• 1 month stipulation for SELIMETRY study seems sensible
Conclusion• Image uniformity degrades at count-rates above and below that used
for calibrations• Unusual artefacts in images at high count rates which pile up towards
centres of PMT, with streaks between the tubes• Multiple uniformity calibrations or list-mode data processing may be
required• Uniformity at calibrated count-rate is stable over ~1 month timescale• Experiments may need to be repeated with Cu behind source to reduce
scatter • Similar experiments are planned for “normal” mode, and with other
radionuclides• Has anyone else studied this?
• Thoughts from the audience would be greatly appreciated