Precise Determination of HDR Timer Accuracy with Novel Multiple-Regressions Method
K Wang*, K Padgett, N Dogan, University of Miami, Miami, FLSU-E-T-140 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose:HDR remote after-loader timer accuracy QA mainly includes linearity QA and timer uncertainty QA, which has always been a challenge due to its millisecond scale. This study illustrates the procedures to precisely measure timer uncertainty with our novel multi-regression method, as part of the endeavor to fulfill a thorough HDR Timer accuracy QA.
Methods:The intercept of linearity QA curve is mainly due to exposures from source travelling inside transmission tube, altered by timer uncertainty for HDR remote after-loader. Comparing the reading from a secondary timer, as sometime suggested, essentially tells nothing unless either timer is broken. We first normalized the position responses for the Well chamber using source strength calibration dataset (for seeking the Sweet Spot), because the readings are position-dependent in Well chamber. We determined the Zero position of transmission tube inside Well chamber by seeking the sudden increase of readings from the same calibration dataset. We measured the exposures at 1, 4, 7 cm from the tip end, respectively. At each given position the dwelling time was 100 seconds, but with 0, 1, 2 and 3 interruptions in middle, so the source travelled 1, 2, 3 and 4 round trips, respectively. Thus at each position we acquired a straight line and a corresponding linear regression, in which the slope tells the travelling response at given position. The 3 slopes were collected and drawn against the travelling distances from the zero position, forming a perfect straight line with the intercept to be the timer uncertainty.
Results:The final linear regression showed that the exposure is linear with the tube length (travelling time). The timer uncertainty is a millisecond constant.
Conclusion:Multi-regressions method described can fully check HDR Timer functions, including linearity, travelling time dependence and timer uncertainty, which could be part of the HDR annual QA.