Quality assurance in clinical dosimetry.

Introduction: Clinical dosimetry in nuclear medicine is developing fast, with an increasing number of procedures performed for a variety of therapeutic applications. In that context, the advent of CE-marked commercial clinical dosimetry software is a positive signal, as they should, in principle, optimize the workflow and increase robustness. However, they pose the problem of the evaluation of their performances, in terms of accuracy but also ease of use (user-friendliness).

Aim: The aim of this presentation is to discuss the various steps required for the evaluation of clinical dosimetry procedures in general, and dosimetry software in particular.

Materials and methods: The clinical dosimetry workflow (CDW) is the suite of steps that lead from calibration procedures to the final reporting of the clinical dosimetry procedure. The study of the CDW implemented in various software shows a age variability in the implementation of the steps that constitute the CDW, and the order of their implementation. This can be accepted, however it raises the issue of comparing software that, basically, do not do the same thing, or do things in a different order.

Results and discussion: The various steps that compose the CDW have to be studied (benchmarked) using specific tools: If assessing calibrations/reconstructions can be made using phantoms filled with radioactive sources, rigid objects are not adapted to the evaluation of registration procedures. Computing anthropomorphic models can be used to verify absorbed dose calculation algorithms (for example using Monte Carlo radiation transport modelling as the gold standard). As can be seen, a range of tools of different type (test objects, models, patient data) must be used - and sometimes developed - to evaluate each step of the CDW. Finally, the end-to-end process must be benchmarked on "real" clinical data, but the price to pay is that the ground truth is not known, thereby limiting these approaches to precision rather than accuracy.

Conclusion: Nuclear medicine dosimetry quality assurance (QA) is in its infancy. However, procedures already applied in external beam radiotherapy may be easily transposed to nuclear medicine, and it will not take decades until nuclear medicine benefits from sound, reproducible procedures that will increase the robustness of clinical dosimetry procedures.