We demonstrate a procedure in retrospective EPR dosimetry which allows for
virtually nondestructive sample evaluation in terms of sample irradiations.
For this procedure to work, it is shown that corrections must be made for
cavity response characteristics when using variable mass samples. Likewise,
methods are employed to correct for empty tube signals, sample anisotropy
and frequency drift while considering the effects of dose distribution opti
mization. A demonstration of the method's utility is given by comparing sam
ple portions evaluated using both the described methodology and standard fu
ll sample additive dose techniques. The samples used in this study are toot
h enamel from teeth removed during routine dental care. We show that by mak
ing all the recommended corrections, very small masses can be both accurate
ly measured and correlated with measurements of other samples. Some issues
relating to dose distribution optimization are also addressed. (C) 2000 Els
evier Science B.V. All rights reserved.