Applying the concept of linear energy transfer (LET) to modelling of biological effects of charged particles usually involves calculation of the average LET. To calculate this, the energy distribution of particles is characterized by either the source spectrum or fluence spectrum. Also, the average can be frequency- or dose-weighted. This makes four methods of calculating the average LET, each producing a different number. The purpose of this note is to describe which of these four methods is best suited for radiobiological modelling. We focused on data for photons (x-rays and gamma radiation) because in this case differences in the four averaging methods are most pronounced. However, our conclusions are equally applicable to photons and hadrons. We based our arguments on recently emerged Monte Carlo data that fully account for transport of electrons down to very low energies comparable to the ionization potential of water. We concluded that the frequency average LET calculated using the fluence spectrum has better predictive power than does that calculated using any of the other three options. This optimal method is not new but is different from those currently dominating research in this area.

将线性能量转移 (LET) 的概念应用于带电粒子的生物效应建模通常涉及平均 LET 的计算。为了计算这一点，粒子的能量分布由源光谱或注量光谱来表征。此外，平均值可以是频率或剂量加权的。这产生了四种计算平均 LET 的方法，每种方法产生不同的数字。本说明的目的是描述这四种方法中哪一种最适合放射生物学建模。我们专注于光子数据（X 射线和伽马辐射），因为在这种情况下，四种平均方法的差异最为明显。然而，我们的结论同样适用于光子和强子。我们的论点基于最近出现的蒙特卡洛数据，这些数据完全解释了电子传输到与水的电离势相当的非常低的能量。我们得出的结论是，使用注量谱计算的频率平均 LET 比使用其他三个选项中的任何一个计算的具有更好的预测能力。这种最佳方法并不新鲜，但与目前在该领域占主导地位的研究不同。