Suspicion of radiological accident in Georgia: the role of the IRSN
P. Voisin, L. Lebaron-Jacobs, J.-F. Bottollier-Depois, P. Gourmelon
IRPA-10: 10. international congress of the International Radiation Protection Association Hiroshima (Japan) 14-19 May 2000, P11-211
When localised accidental irradiation occurs, it is necessary to determine the extent to which the tissue and vital organs have been damaged, mainly in the zone located near the source. At present, biological markets cannot be used to estimate the heterogeneousness of the dose, useful for devising the therapeutic strategy. An alternative would be to calculate the dose due to exposure in order to map the absorbed dose in the different regions of the organism. Our technique, combining Computer-Assisted Design (CAD) and Monte-Carlo calculation, can assess the dose distribution in a numerical anthropomorphic phantom fitted on the specific morphology of the victim. The CAD is used to simulate the accident, taking into account the specific morphology of the irradiated person, as well as the source and the environment of the accident. The Monte-Carlo code runs to transport photons and neutrons in this three-dimensional space, following a probabilistic method: each particle generated at the source and its interactions or energy losses along its path are calculated by reproducing faithfully the random nature of the interactions between particles and matter. The result of the calculation provides the relative absorbed dose distribution per source particle, so it has to be normalised in order to give the absolute value of the absorbed dose. The first way to get this data is straightforward if one knows both the activity of the source and the exposure time. However, in practice, the exposure time is very badly estimated, and even unknown. We propose a new approach to resolve this uncertainty. The technique consists in fitting the calculation results at one point where the absolute value of the absorbed dose is known or assessed. Different methods are proposed: the results are matched to the clinical signs of the lesion, particularly around the rim of the radiological necrosis, where the dose can be estimated to roughly 25 Gy. the whole-body dose assessed by the Monte-Carlo code is matched on the biological dose based on the scoring of dicentrics in peripheral blood lymphocytes. any other means giving the absorbed dose at a given point, i.e. ESR measurements on tooth enamel for instance. This technique was applied after several accident on patients who presented very severe lesions due to acute localised irradiation: Lilo (Georgia) in 1996-1997, in South Africa on a building site in 1998 and in Peru on a hydroelectric power plant in 1999.