Low fluence.

The question of the appropriate extrapolation to low dose has long been a subject of controversy. A linear no-threshold model is favored by regulatory bodies as the basis of RBE assignments and estimates of radiation hazards to the general population. This model is largely supported by extensive application of the linear-quadratic survival formula "fitted" statistically to a wide variety of experimental data obtained at doses typically exceeding 1 Gy, and then extrapolated to mGy for practical applications, and even to the prediction of hazards from single electrons. Such extrapolations are questionable at best, and may even prove hazardous for risk evaluations. Fluence and geometry rather than dose based data are proposed as a basis for a limiting "threshold" for a "low dose" extrapolation. The proposed threshold is one where the fluence of particles is one per square micron, where on average only 2/3 of the 1 micrometers2 pixels covering an irradiated area are traversed by one or more particles. The corresponding dose threshold is determined by the LET of the bombarding radiation. For relativistic electrons this dose is about 0.032 Gy.