Abstract
Background: Quantifying absorbed doses from radiopharmaceuticals within human organs necessitates advanced computational modeling, as direct in vivo measurement remains impractical. Methods: In this study, three Monte Carlo-based simulation codes, Monte Carlo N-Particle version 6 (MCNP6), GEANT4 Application for Tomographic Emission (GATE), and GEANT4-based Architecture for Medicine-Oriented Simulations (GAMOS), were employed to evaluate internal dosimetry following the Medical Internal Radiation Dose (MIRD) formalism. As an illustrative case, simulations were first performed for (99m)Tc-MIBI uptake in the myocardium using the anthropomorphic phantom, with the heart modeled as the source organ to assess energy deposition in key target organs. Dose assessments were conducted at two time points: immediately post-injection and at 60 min post-injection (representing the cardiac rest phase), allowing comparison against established clinical reference data. Results: Across all codes, organ-specific dose distributions exhibited strong consistency. The pancreas absorbed the highest dose (GATE: 21%, GAMOS: 20%, MCNP6: 22%), followed by the gallbladder (GATE: 18%, GAMOS: 17%, MCNP6: 18%) and kidneys (GATE: 16%, GAMOS: 15%, MCNP6: 16%). These findings established a consistent organ dose ranking: pancreas > gallbladder > kidneys > spleen > heart/liver, corroborating previously published empirical data. To demonstrate the versatility of the framework, additional simulations were performed with (18)F in an anthropomorphic phantom and with spherical tumor models using therapeutic radionuclides ((177)Lu and (225)Ac). This broader application underscores the adaptability of the tri-code approach for both diagnostic and therapeutic scenarios. Conclusions: This comparative analysis highlights the complementary advantages of each Monte Carlo platform. GATE is well-suited for high-fidelity clinical applications where anatomical and physical accuracy are critical. GAMOS proves advantageous for rapid prototyping and iterative modeling workflows. MCNP6 remains a reliable benchmark tool, particularly effective in scenarios requiring robust radiation transport validation. Together, these Monte Carlo frameworks form a validated and adaptable toolkit for advancing internal dosimetry in personalized nuclear medicine, supporting both clinical decision-making and the development of safer, more effective radiopharmaceutical therapies.