Dosimetry consists in measuring (in case of external radiation) or estimating (in case of internal contamination) the committed effective dose, which depends on:
the source: whether sealed or unsealed, its activity or intensity, and the nature, energy and biological effectiveness of the radiation.
the mode of exposure: its duration, fractionation, and rate.
the target: the tissues or organs affected (up to the whole body level), and the age and radiosensitivity of the individual.
For irradiation, detection systems are calibrated and certified to be directly proportional to the dose received. An adjustment must be made to transpose operational quantities into dosimetric quantities depending on the heterogeneity of the irradiation, exposure time, rate and the nature of the emission.
As the level of internal contamination is not directly measurable, estimating the dosimetric impact is more complex. Indeed, evaluating the amount incorporated and then calculating the dosimetric detriment requires having an appropriate biokinetic model for the radionuclide(s) involved. These models and data are provided by the International Commission on Radiological Protection (ICRP). In practice, however, the use of these models is based on forming hypotheses regarding the circumstances of an incorporation event (either known or suspected), such as the chemical and physical characteristics of the radioactive contaminant to which the individual has been exposed, and the time interval between incorporation and the measurement. Repeated measurements over time make it possible to clarify the hypotheses and to refine the model, for a better dose evaluation.
To help address the health risks, the dose must be expressed in terms of the committed effective dose (50 years for workers) for all external and internal components, or as an equivalent dose to the most irradiated organ, or as an absorbed dose.
Indeed, the ICRP has defined radiological protection quantities, to determine the dose limits for occupational exposure. This ensures that deterministic effects can be avoided, and that the risk of developing stochastic health effects can be maintained at an acceptable level.