ERROR · A pEdiatRic dosimetRy personalized platfORm based on computational anthropomorphic phantoms

A pEdiatRic dosimetRy personalized platfORm based on computational anthropomorphic phantoms

Patient-specific internal dosimetry is of great interest, especially in childhood and puberty where the relative risk from the exposure to ionizing radiation compared to that for adults is widely debated. Pediatric dosimetry has been extensively discussed over the last two decades, in terms of administered activity reduction while retaining acceptable image quality of standalone acquisitions (CT,SPECT,PET) or hybrid examinations (SPECT/CT, PET/CT, PET/MR). The radiation absorbed dose delivered to pediatric patients in diagnostic and therapeutic procedures is still a matter of concern. Monte Carlo simulations is the gold standard method for accurate dosimetry assessment. However, due to complexity and large computational time needed, they are not applied in clinical routine. Instead, for couple of decades, there have been dosimetry factors calculated through MC techniques and are still applied for the estimation of the absorbed dose. Modern medicine requires alternative and updated techniques to approach personalized dosimetry, considering patient's individual anatomical characteristics. The main goal of the ERROR is the creation of a framework, freely available to the scientific community for the dose assessment on specific organs of interest in pediatric diagnostic and therapeutic applications. A novel methodology is applied to calculate dose factors for a pre-defined children population, acquiring realistic simulations, in terms of physics modeling, human anatomy, activity distribution/irradiation, low statistical uncertainty. The dosimetry framework will provide the definition of absorbed dose per organ according to the age, gender, weight, administered activity, duration time and the examination. In collaboration with the 3 academic partners, there is a novel combination of several methodologies and tools for achieving personalized pediatric dosimetry. Machine learning techniques, GPU based MC simulations, detailed anthropomorphic models, image processing algorithms, clinical pediatric data and feedback from clinicians will be used to facilitate the dosimetry calculation for each pediatric patient, based on pre-calculated data.

Data: CORDIS, © European Union