Advancing Internal Dosimetry in Personalized Nuclear Medicine: Toward Optimized Radiopharmaceutical Use

MCNP6, GATE and GAMOS now produce convergent organ-absorbed dose estimates, reducing cross-platform variability and enabling actionable, patient-specific radiopharmaceutical dosing.

Comparative testing across diagnostic and therapeutic scenarios shows consistent organ-level dose estimates between engines. Studies included 99mTc myocardial distributions, 18F voxelized phantom cases, and therapeutic models using 177Lu and 225Ac in spherical tumors within anthropomorphic phantoms (n=5) and cohort-style synthetic datasets (n=20). Median absolute organ-dose differences were roughly 5–12% (lower for diagnostic cases), and concordance statistics frequently exceeded 0.9—results that strengthen confidence in applying Monte Carlo outputs for clinical dosimetry validation and reproducibility.

Dose patterns remain tracer- and biodistribution-dependent. In the tested agents (including 177Lu and selected diagnostic tracers such as 99mTc), the highest absorbed doses were most often observed in the kidneys and, depending on distribution, in the pancreas or gallbladder.

Each engine brings distinct operational strengths: GATE offers high anatomical fidelity with detailed photon–tissue modeling; GAMOS supports rapid prototyping with efficient runtimes; and MCNP6 serves as a robust benchmark for validation. Match the platform to the clinical question—voxel-level anatomical accuracy favors GATE, iterative protocol development favors GAMOS, and stringent verification favors MCNP6.

The synthesis suggests multidisciplinary teams will integrate Monte Carlo dosimetry into staged clinical workflows to support safer, more personalized radiopharmaceutical dosing.

Adoption should proceed via phased institutional validation with physicists, nuclear medicine physicians, and dosimetrists. Practical next steps include standardized validation cases, workflow integration for routine plan generation, and cross-disciplinary checkpoints to ensure simulation outputs inform therapeutic decisions reliably.