Abstract
534
Introduction: Absorbed dose calculation in clinical practice requires the determination of the location and number of radioactive sources in the patient, and the assessment of how the emitted energy propagates and is eventually absorbed. The way radiopharmaceutical dosimetry is implemented in clinical practice depends on the nuclear medicine procedure type: • For diagnostic applications, model-based dosimetry is most often performed, based on precomputed tables of S values, i.e. absorbed doses in a target per radioactive decay in the source (Gy.Bq-1.s-1). These are calculated from computing models (anatomical, then voxel-based or hybrid). • In therapy, patient-specific dosimetry requires the implementation of radiation transport algorithms that account for the anatomy of the patient. Depending on the relationship between source/target geometry and radiation range several algorithms can be used, from Local Energy Deposition to refined Monte Carlo approaches. Materials and Methods: The OpenDose collaboration was created in 2017 with the aim to support Nuclear Medicine dosimetry development by providing open access resources, such as data, software and educational material. The OpenDose collaboration currently gathers 18 research teams with experience in different domains of radiopharmaceutical dosimetry (data production, software development and clinical validation).
Results: For model-based dosimetry, the collaboration is generating specific absorbed fractions (SAF) and S values from reference models, using 6 of the most popular Monte Carlo radiation transport codes (EGSnrc/EGS++, FLUKA, GATE, Geant4, MCNP/MCNPX and PENELOPE). Generated data is provided with statistical uncertainties, verified, and stored in a relational database to insure traceability. Dosimetric data has been produced for all source/target combinations of the two reference ICRP 110 adult models (male and female). SAFs were compared with ICRP 133 results, and S values are in good agreement with OLINDA 2.0 and IDAC 2.1 software.In parallel, a patient-specific software, OpenDose 4D, was developed and is currently under validation. The patient-specific dosimetry package is based on 3D Slicer (www.slicer.org), which allows users to benefit from extensively validated solutions for image DICOM import and processing (multimodal image visualization, fusion, segmentation). The dosimetry-specific development focused on the implementation of the integration of time-indexed data (activity or absorbed dose rate 3D maps) and absorbed dose calculation algorithms (local energy deposition, convolution in homogeneous and heterogeneous media). An input/output module allows to perform external Monte Carlo absorbed dose calculation using GATE.A dedicated website (www.opendose.org) was developed to provide easy access to the model-based dosimetric database and to the patient-specific dosimetry software. A section of the website provides educational material such as lectures and a bibliography.
Conclusions: The OpenDose collaboration brings together the resources and expertise of many research teams involved in radiopharmaceutical dosimetry. The www.opendose.org website provides a global repository for internal dosimetry including data, software, and education material. The collaboration, which is open to new research teams, will extend data production to other dosimetric models and implement new free features such as online dosimetric tools.