Errors introduced by dose scaling for relative dosimetry

Abstract

Some dosimeters require a relationship between detector signal and delivered dose. The relationship (characteristic curve or calibration equation) usually depends on the environment under which the dosimeters are manufactured or stored. To compensate for the difference in radiation response among different batches of dosimeters, the measured dose can be scaled by normalizing the measured dose to a specific dose. Such a procedure, often called “relative dosimetry”, allows us to skip the time‐consuming production of a calibration curve for each irradiation. In this study, the magnitudes of errors due to the dose scaling procedure were evaluated by using the characteristic curves of BANG3 polymer gel dosimeter, radiographic EDR2 films, and GAFCHROMIC EBT2 films. Several sets of calibration data were obtained for each type of dosimeters, and a calibration equation of one set of data was used to estimate doses of the other dosimeters from different batches. The scaled doses were then compared with expected doses, which were obtained by using the true calibration equation specific to each batch. In general, the magnitude of errors increased with increasing deviation of the dose scaling factor from unity. Also, the errors strongly depended on the difference in the shape of the true and reference calibration curves. For example, for the BANG3 polymer gel, of which the characteristic curve can be approximated with a linear equation, the error for a batch requiring a dose scaling factor of 0.87 was larger than the errors for other batches requiring smaller magnitudes of dose scaling, or scaling factors of 0.93 or 1.02. The characteristic curves of EDR2 and EBT2 films required nonlinear equations. With those dosimeters, errors larger than 5% were commonly observed in the dose ranges of below 50% and above 150% of the normalization dose. In conclusion, the dose scaling for relative dosimetry introduces large errors in the measured doses when a large dose scaling is applied, and this procedure should be applied with special care.PACS numbers: 87.56.Da, 06.20.Dk, 06.20.fb

Dosimetric verification for intensity-modulated arc therapy plans by use of 2D diode array, radiochromic film and radiosensitive polymer gel

“…Hence, the rescaled measured dose tends to deviate from the calculated doses for the doses far from the normalization dose. The error is particularly large for doses much lower than the normalization dose [ 17 ]. These observations partially explain the poor agreement between the measured and calculated doses in regions with doses < 50% of the maximum dose.…”

3D evaluation of 3DVH program using BANG3 polymer gel dosimeter

“…In the previous study 30 we demonstrated that a dose scaling procedure used for a relative dosimetry introduces an error in the measured dose values. The magnitude of the error increases as the magnitude of dose scaling increases.…”

“…It is known that this procedure causes a large error, in particular, at the dose level lower than 50% of the normalization dose. 30 Therefore, the agreement between Pinnacle 3 and BANG3 data is expected to be not good near the phantom surface and in the region where the expected dose is low. Figure 4 shows the gamma passing rates as the function of the dose threshold.…”

“…The error analysis in Ref. 30 showed that the rescaled dose (i.e., scaling factor = 0.93) is as much as 2.5% lower for the doses below the normalization dose. The maximum dose error occurs at a dose about 50% of the normalization dose.…”

Improving the calibration of radiochromic films by the use of uncertainties in optical density and dose

“…Usually this table is fitted to a parametric model, where the dose is the dependent variable. [5][6][7][8][9][10] The nature of the model is phenomenological, because the physics behind the darkness of the film is only known qualitatively. The fit is necessary to filter out the noise and to interpolate the unknown doses.…”