Yoon Suk Whan, Lin Hui, Mihailidis Dimitris, Kennedy Christopher, Li Taoran
Department of Radiation Oncology, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
Department of Radiation Oncology, University of California San Francisco, San Francisco, California, USA.
Med Phys. 2022 May;49(5):3489-3496. doi: 10.1002/mp.15571. Epub 2022 Mar 4.
To identify sources of systemic errors and estimate their effects, especially the vendor-provided sensitivity S , on total body irradiation (TBI) and total skin electron therapy (TSET) in vivo OSLD measurements.
Calibration nanoDot OSLDs were irradiated 50-300cGy under reference conditions. Raw OSLD readings M were corrected by S to obtain corrected readings M . A quadratic calibration curve relating M to delivered dose D was established and commissioned for clinical use. For clinical measurements, directly adjacent pairs of nanoDot OSLDs were placed on TBI and TSET patients with a medical tape with or without 1.5 cm of bolus respectively before treatment. Used OSLDs were bleached between each use until cumulative dose of 15 Gy. Relative difference in corrected counts (∆M = pair-difference/mean) was fitted multi-linearly versus possible sources of systemic errors (S , bleaching history, cumulative dose, and age differences). Total of 101 TBI and 110 TSET measurement pairs from calibrated batches were analyzed.
S introduced a residual systemic error to corrected readings M (-0.98% per +0.01, p = 4e-12). Given S distribution is σ = ±0.025, measured dose 1-σ error is ±2.5%, compared to ±2.8% uncertainty reported in the literature which may include this systemic error. Bleaching or cumulative dose did not affect M significantly after adjusting for S . Adjusting for the systemic error in S decreased two-sample mean D median absolute error from ±2.6% to ±1.9% and 95-percentile absolute error from ±7.1% to ±5.5%. Variability between paired clinical OSLDs was larger for TBI versus TSET at σ = ±4.7% and ±6.3%, respectively, despite similar predictor distributions.
Our findings suggest that M correction via vendor-provided sensitivity results in a small but significant systemic error. Dosimeters with outlier sensitivities should be excluded during batch calibration to minimize error. Bleaching and cumulative dose likely minimally affect measurements if cumulative dose is controlled below 15 Gy. Random errors were higher for TSET than TBI.
识别系统误差来源并估计其影响，特别是供应商提供的灵敏度S对体内光激发光剂量计（OSLD）测量全身照射（TBI）和全身皮肤电子治疗（TSET）的影响。
校准纳米点OSLD在参考条件下接受50 - 300cGy的照射。原始OSLD读数M通过S进行校正以获得校正读数M。建立了将M与给予剂量D相关的二次校准曲线并投入临床使用。对于临床测量，在治疗前，将直接相邻的纳米点OSLD对分别用有或没有1.5 cm bolus的医用胶带放置在TBI和TSET患者身上。使用过的OSLD在每次使用之间进行漂白，直至累积剂量达到15 Gy。校正计数的相对差异（∆M = 配对差异/均值）与系统误差的可能来源（S、漂白历史、累积剂量和年龄差异）进行多线性拟合。分析了来自校准批次的总共101对TBI和110对TSET测量数据。
S给校正读数M引入了残余系统误差（每 +0.01为 -0.98%，p = 4e - 12）。鉴于S的分布为σ = ±0.025，测量剂量的1 - σ误差为±2.5%，相比文献中报道的±2.8%的不确定度，后者可能包括此系统误差。在对S进行校正后，漂白或累积剂量对M没有显著影响。对S中的系统误差进行校正后，两样本均值D的中位数绝对误差从±2.6%降至±1.9%，95百分位数绝对误差从±7.1%降至±5.5%。尽管预测因子分布相似，但TBI时配对临床OSLD之间的变异性在σ = ±4.7%时大于TSET时的±6.3%。
我们的研究结果表明，通过供应商提供的灵敏度对M进行校正会导致一个小但显著的系统误差。在批次校准期间应排除具有异常灵敏度的剂量计，以尽量减少误差。如果累积剂量控制在15 Gy以下，漂白和累积剂量可能对测量影响最小。TSET的随机误差高于TBI。
