Larry A. DeWerd, Ph.D., FAAPM UW ADCL & Dept. of Medical Physics University of Wisconsin NCCAAPM Meeting Oct 11,2013
Larry DeWerd has a partial interest in Standard Imaging
Talks from people who have been in the field for awhile tend to be historical or philosophical My attempt at philosophical and hopefully informational material.
I don t have enough time Why worry it will all work out. I don t have to think about it it is late. Didn t the manufacturer tell me it was okay.
Some Manufacturers harp on the time saved, etc but their instrument is not as precise and can result in errors. For example, some digital sampling electrometers (depending on their sampling algorithm) may miss some signal from pulses of the accelerator. The physicist should ask.
Technicians use numbers, Medical physicists use understanding. Do it right good enough is not good enough
Transfer of Knowledge is Important. ADCLs transfer latest knowledge Â\y \ {täx áxxç yâüà{xü à{tç Éà{xÜá? à á uç áàtçw Çz âñéç à{x á{éâäwxüá Éy z tçàáê f Ü \áttv axãàéç? DIJI f Ü \áttv
Just doing a measurement and applying ADCL calibration does not mean that it is right You need to understand the equipment and see through the measurement. Numbers no matter how precise cannot by themselves imply anything. Precision and accuracy of dose measurements and reporting of the measurement details should be sufficient to allow the work to be interpreted and repeated and to allow valid comparisons to be made, both in the same laboratory and by other laboratories.
Ion chambers require calibration for the most precise measurements ADCLs willing to discuss measurements and methodology ADCL discuss the operation of instrumentation.
1 0.9 primary standard measurement frequency of result 0.8 0.7 0.6 0.5 0.4 0.3 clinical measurement conventional true value 0.2 0.1 0-6 -4-2 0 2 4 6 a.u.
NBS petitioned AAPM to create Regional Calibration Laboratories in 1975- In 1983 called ADCLs. NBS acknowledges traceability to primary standards (Proficiency tests) Agreement for Proficiency tests and round robins < 0.5% UWADCL founded 1981 by LAD
International & National Standards Primary Calibration Secondary Standards ADCL Hospital or Clinical Standards Secondary Calibration Tertiary Calibration Operational Standards
AAPM in conjunction with NIST set up secondary laboratories for medical ionizing radiation calibrations - For therapy applications the precision is within 0.5%. Traceability of radiation dose is dependent on standards that are the same throughout the world at primary labs.
All calibrations (especially for medical applications) should have a traceability train to a standard This provides confidence in the dose given because it was measured and can be traced back to a national standard at NIST However, the SSDL network is not the equivalent to the ADCL network
The ADCLs have proven track records of providing precise calibrations of equipment for Therapy and Diagnostic applications Safety and treatment delivery or imaging is improved because of ADCL calibration and research.
0.994 0.995 0.996 0.997 0.998 0.999 1.001 1.002 1 1.003 1.004 1.005 N u m b e r 16 14 12 10 8 6 4 2 0 Mean of ADCLs/ADCL
These tests have been in place over 20 years NIST and ADCLs agree within 0.5% for Cobalt-60 beams NIST and ADCLs agree less than 2.0 % for x-ray beams between 20 kvp and 250 kvp
Maintenance of accuracy and precision is very important for Medical Applications Knowledge of characteristics of chambers is very important Medical Applications rely heavily upon precise traceable calibrations.
Calibration of chambers should be done across the energy range of use. Bracket your point of interest. Generally there is a reference energy in use that is the minimum energy point used.
Because of variation in response with energy for chambers a range of calibration is necessary. The calibration should be done under correct criteria e.g. TG 61 There is no protocol using an absorbed dose to water calibration for x-ray beams. An SSDL offers absorbed dose in water but no protocol.
Within + 1% for Mammography range: 24 kvp to 35 kvp, 0.15 to 1.0 mm Al Within + 2 % or 2.5% for entire diagnostic range, up to 150 kvp For superficial or orthovoltage (radiobiology) 1 % to 2 % is desirable
There can be electron contamination and full buildup is needed for calibration of low energy (< 100kV) clinical beams This requires an addition to the window to be put over a plane-parallel chamber The window of the chamber (for example 2.5 mg / cm 2) should be subtracted from the values in the following table for the thickness of the foil for full buildup.
Tube potential (kv) Total wall thickness (mg cm -2 ) 40 3.0 50 4.0 60 5.5 70 7.3 80 9.1 90 11.2 100 13.4
Physicist should understand uncertainties as delivered by the ADCL For example, for cobalt calibrations
The expanded uncertainty (at k=2) is expected to include the conventional true value. The chance of the population mean (the true value) being within the measured value has a 95% certainty. Or we can be sure that our measurement indicates the correct value within 95%
ADCL Standard Chamber Parameter Type A Type B Charge 0.02 0.0 Timing 0.04 0.05 Air Density 0 0.1 Ionic Recombination 0 0 Distance from the source 0.1 0.07 Beam Uniformity 0 0.1 Air attenuation 0.1 0 Quadratic Sum 0.110 0.166 Type A and B Quadratic Sum 0.198 NIST Source Calibration 0.600 u NIST-ADCL Chamber Calibration 0.632 or at k=2 level, 1.26 %
Clinic Chamber Parameter Type A Type B Charge 0.02 0.1 Timing 0.04 0.05 Air Density 0 0.1 Ionic Recombination 0 0.1 Distance from the source 0.1 0.07 Beam Uniformity 0 0.1 Quadratic Sum 0.110 0.218 Type A and B Quadratic Sum 0.244 ADCL Electrometer Readout 0.057 ADCL Transfer Calibration 0.632 u ADCL-Clinic Chamber Calibration 0.680 or at k=2 level, 1.36 %
Standards changing Absorbed dose to water done with precision and uniformly. Calibration of chambers and sources essential for radiation therapy: Majority traceable through Cobalt and Cesium. Unfortunately, some manufacturers improvise dosimetry to market their product
Manufacturers need to pass some hoops to market product However, 510k from FDA does not mean that they have a calibration regimen for their machine. Also some manufacturers change without letting physicists know and assume everything still applies.
What happens when manufacturers improvise for Standards? Errors! NIST, ADCLs, AAPM should insist on a standard There is a need to have traceability delivered by the Accredited Dosimetry Calibration Laboratories Hospital Physicists generally have requests Users must insist on traceable standards from NIST through the ADCLs
Sr-90 applicators introduced in 1950. Plane or concave. 28 year half life Typical 8mm diameter with a beta shield Extremely high dose rate Calibrated by each mfr wrt Bragg-Gray theory Bizarre units like reps, beta-roentgens Hospitals asked UW ADCL accuracy of dose.
Wing-like growth in conjunctiva Requires surgery to remove Surgery fails 90% of time One dose with Sr-90 prevents recurrence in roughly 90% of patients treated
One Chicago hospital had an Amersham Model SIA-20 applicator calibrated at NBS: calibrations differed by 38%. Each lab stood by their calibration! NIST had to review their calibrations
Sr-90 sources had only been intercompared IN AIR at distances of 20 to 30cm Desired dose rate is IN CONTACT with Sr-90 on surface of silver matrix Chris Soares developed an extrapolation chamber for calibration.
NIST traceable determination of the absorbed dose to water rate in the central 4 mm of the applicator including color enhanced contour plots and two dimensional dose profiles. Routine ADCL calibrations from UW ADCL using radiochromic film began in 1996. NIST quotes uncertainty of 7 % Calibration disagreement (old versus new) average -27 %.
Example of a color enhanced contour plot of a uniform Sr-90 ophthalmic applicator. The dose weighted isocenter is equidistant from the hash marks. The outer circle represents the source physical diameter. 70 60 50 40 30 20 10 80 90
2-Dimensional dose profiles for a uniform source. Horizontal Profile Vertical Profile 0.50 0.50 Dose Rate (Gy/sec) 0.40 0.30 0.20 0.10 Dose Rate (Gy/sec) 0.40 0.30 0.20 0.10 0.00 0.00-6.0-4.0-2.0 0.0 2.0 4.0 6.0 Distance (mm) -6.0-4.0-2.0 0.0 2.0 4.0 6.0 Distance (mm)
Example of a color enhanced contour plot of a skewed-sr-90 ophthalmic applicator. Notice the offset or shift of the dose weighted isocenter from the physical source center, and the nonuniform dose distribution. 40 30 20 10 50 60 70 90 80 90
Horizontal Profile Vertical Profile 0.40 0.40 0.30 0.30 0.20 0.20 0.10 0.10 0.00-6.0-4.0-2.0 0.0 2.0 4.0 6.0 Distance (mm) 0.00-6.0-4.0-2.0 0.0 2.0 4.0 6.0 Distance (mm)
The average difference between the old calibration and the present calibration is -27%. The distribution of source activity can be very unique for the individual source.
NIST standard Pd-103 has very short half-life (17.0 d) so NIST traceable 109 Cd source (half-life of 463.2d) was used as a reference source for 12 years, then replaced in 1997 However, self-shielding of the source encapsulation was different between these two isotopes This resulted in a sudden 9% shift in calibration by letter to users in 1997
There can be significant errors If traceable to NIST, even if wrong, everyone is consistent if traceable standard- at least traceable through ADCL NIST, ADCLs and CIRMS need to insist upon standards. Manufacturers try hard but they need to rely on NIST and ADCLs for standards
Âg{Éáx ã{é wé ÇÉà ÜxÅxÅuxÜ à{x Ñtáà tüx véçwxåçxw àé ÜxÑxtà àaê George Santayana, Harvard Professor and poet
Thanks are due to Students and staff of the UW ADCL All those who send us calibration instruments that support the research program of the UW Medical Radiation Research Center.