| Sumario: | Accurate internal dosimetry is essential for the optimization and development of radionuclide therapies. Among activity-limiting organs, the red marrow (RM) is particularly challenging due to its heterogeneous microstructure and distributed nature. The microscopic geometry of the trabecular bone and marrow space influences energy deposition from both short- and long-range particles, making simplified local dose assumptions insufficient even for alpha emitters. Furthermore, interpatient variability in marrow cellularity can substantially affect absorbed dose estimates. To improve accessibility and reproducibility in marrow dosimetry, we developed a software tool, MyeloDose, that calculates S-value-based absorbed dose to the RM for different source tissues, radionuclides and cellularities. Emission, branching, and decay data for 1252 radionuclides were extracted from ICRP Publication 107, including alpha particles, beta particles, internal conversion and Auger electrons. For beta emitters, energy-dependent spectra were integrated over the full emission range. Specific absorption fractions (AFs) for electrons and alpha particles were obtained from the University of Florida series of computational phantoms, using datasets derived from Monte Carlo transport simulations. Energy-dependent AFs were interpolated using four different techniques: nearest neighbor, linear, log–log, and cubic spline. The lumbar vertebrae were chosen as a test case, with source regions defined as RM and trabecular bone surface (TBS). Cellularity factors ranging from 10% to 100% were evaluated in steps of 10%. Dose factors (mGy/MBq·s) were calculated for ⁹⁰Y, ¹⁶¹Tb, ¹⁷⁷Lu, ²²⁵Ac, ²²⁴Ra, ²²³Ra,²²⁷Th, and ²¹¹At. For radionuclides with decay chains, fractional branching and daughter contributions were automatically accounted for. Optional features include user-defined exclusion of long-lived daughters and the inclusion of alpha recoil energy deposition. A Python-based tool was developed to automate data handling, featuring both a simple graphical interface and batch processing functionality. Dose factors were calculated for the test selection of eight radionuclides, using a reference cellularity factor of 70 for the lumbar vertebrae, with source activity located either in (RM) or on the TBS. The variation introduced by using different interpolation techniques, expressed as the min–max difference normalized by the mean, remained within 1.05% For the beta-emitters ¹⁷⁷Lu, ¹⁶¹Tb, and ⁹⁰Y, RM dose factors ranged from 1.1×10⁻⁴ to 5.1×10⁻⁴ mGy/MBq, whereas TBS dose factors were approximately 30% lower for ¹⁷⁷Lu and ¹⁶¹Tb and only 5% lower for ⁹⁰Y. For alpha-emitters, RM dose factors were 5.5×10⁻³ mGy/MBq for ²¹¹At and 2.2–2.3×10⁻² mGy/MBq for ²²⁴Ra, ²²³Ra, and ²²⁵Ac; ²²⁷Th exhibited the dose factor (2.7×10⁻² mGy/MBq). When activity was located on TBS, the corresponding dose factors for alpha-emitters were reduced by approximately 70%. We developed a software for computing S-factors in skeletal sites using ICRP 107 emission data and University of Florida absorption fractions. The results demonstrate that interpolation technique has minimal impact on computed dose factors. This tool enables fast marrow dosimetry for both alpha- and beta-emitting radionuclides. This work was supported in part by the South-Eastern Norway Regional Health Authority (Grant 2025/018).
|
|---|