A procedure of separating the primary- and scatter-dose components in
therapeutic photon beams is examined. It is based on the observation t
hat the scatter-dose component is proportional to the variable z=rd/(r
+d), where r and d are beam radius and depth in phantom, respectively.
It is, therefore, possible to express an absorbed dose in the form of
a linear equation D(r)=P+Nz, where at a fixed depth d, both primary d
ose P and coefficient N are constant. A method of linear extrapolation
of an absorbed dose D(r) to ''zero-field'' size, i.e., r=0, is utiliz
ed. Since Monte Carlo technique is capable of scoring separately the p
rimary- and scatter-dose components, it is used to evaluate the accura
cy of the linear extrapolation method within the range of Co-60 - 15-M
V nominal photon energies. The results demonstrate that this method is
sufficiently accurate to obtain the primary dose component in photon
beams. For Co-60 gamma radiation in water, tabulated sets of measured
depth-dose data are analyzed by the linear extrapolation method to rev
iew ''zero-field'' dose values [percentage depth dose (PDD) and tissue
-air ratio (TAR) tables of the British Journal of Radiology, Suppl. 17
]. The ''zero-field'' PDD data are found to be accurate within limits
of experimental uncertainties. Inconsistencies in the TAR table are il
lustrated and discussed. Co-60 tables of relative doses, D(r,d)/P(d(ma
x)), including ''zero-field'' values for both fixed SSD and isocentric
geometries, are generated. Dose calculation in irregular fields is co
nsidered. The linear extrapolation method is recommended as a standard
procedure for separating primary dose from depth-dose data in high-en
ergy photon beams.