Measurement Uncertainty. Measurements Accuracy - hitting the center of the target Precision - tight pattern of hits Bias - all two inches high Uncertainty.

Measurement Uncertainty

Measurements

• Accuracy - hitting the center of the target

• Precision - tight pattern of hits

• Bias - all two inches high

• Uncertainty - circle containing 95% hits

Traceability

• Accuracy - close to system limitation

• Precision - range

• Bias - identified and corrected for

• Acceptable Uncertainty - 2 sigma

• Unbroken links to national standards

Calibration Traceability

• Unbroken linkage to national standards

• Uncertainty - elements identified and quantified

• Strong QC/QA program

• Periodic MQA testing

Measurement Traceability

• Starts with calibration uncertainty

• Identifies and quantifies all factors that affect the quality of a measurement

• Requires reproducibility

• Testing usually required to validate values

How this affects you

• Did I make the measurement right?

• Is it reproducible?

• What uncertainty is involved?

• Can I cite them for a violation?

2 cases, more than 2 results

• Package survey of radioactive material

• Mammography machine measurements

Package

• 31x31x27 cm

• Bicron RSO-50E

• Calibrated by secondary laboratory

• Maximum surface dose rate measured is 0.4 uSv/h (0.4 mR/h)

• Traveling by air

IATA (ST-1 Requirements)

• Accuracy - so that correctly labeled

• Precision/uncertainty - based on package surface to detector center distance

• Bias - as high as a factor of 4 low

ST-2 Measurement Guidance

a

D/P Distance Package Length CF

5 cm 20 - 40 cm

40 - 100 cm

>100 cm

2.3

1.6

1.0

Correction factors for package and detector sizes

Mammography Measurements

• Exposure at skin entrance

• Half value layer

• Non-invasive kVp measurements

• Mean glandular dose calculation

X-ray vs. Mammo Probes

Energy Response 10x5-6 vs 105-6M

0.8

0.85

0.9

0.95

1

1.05

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5

HVL (mm Al)

Ind

ica

ted

vs.

Tru

e

10x5 -6 probe

10x5 -6M

Phantom Variability

0.94

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

1.03

1.04

1.05

Normalization Factor

1202 2202 2203 1200 2790 8665 2791

Serial Number

Phantom Attenuation Variability

Temperature Effects

MDH 1515 Temperature Effect

1860

1880

1900

1920

1940

1960

1980

2000

2020

0 10 20 30 40 50 60 70

Minutes since brought indoors

mR/

min

HVL MeasurementsHVL Variation Due to Curve Fit

5.41

5.51

5.61

0.29 0.3 0.31 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.39 0.4 0.41

mm Al

ln o

f exp

osur

e

6 Points(FDA method - 0.3095 mm Al)

3 points (drop 0.2 mm Al - 0.3027 mmAl) (ACR method)

data points

MQSA method

ACR method

Non Invasive kVp Measurements

kVp Filter Dependence

-2

-1.5

-1

-0.5

0

0.5

1

22 24 26 28 30 32

Measured kVp

kV C

orr

ect

ion 0.015 mm Mo

0.02 mm Mo

0.025 mm Mo

0.03 mm Mo

0.035 mm Mo

Example of How Combined Uncertainties EffectMeasurement

0.03Machine Repro.

0.01Rate Dependence

0.01Energy Dependence

0.01Temp/Pressure

0.015Meter Calibration

Estimated Std. Dev.Component

(Cont.)

Example of How Combined Uncertainties EffectMeasurement

0.0070.02kVp (+/- 1 kVp)

0.020.06HVL(+/-0.02mmAl)

0.020.06Phantom Variability

0.020.06Probe Placement

Estimated Std. Dev.

Estimated Max. Dev.Component

Estimate Std Deviation as Max Deviation/3 (NCRP Report 112)

(Cont.)

Combined Uncertainty

• Total estimated uncertainty

• ( 1.52+ 12+ 12+12+ 32+ 22 + 22 + 22 + 0.72)1/2 = 5.2

• To estimate a confidence level of 95% use a coverage factor of 2 = 2 x 5.2 = 10.4 %