Beyond TG‑43: A PRISMA-based systematic review on model-based dose-calculation algorithms in brachytherapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• The AAPM TG-43 formalism is the standard for brachytherapy dose calculation but assumes a homogeneous medium, neglecting tissue heterogeneities and scatter.
• Model-based dose-calculation algorithms (MBDCAs), including Monte Carlo (MC) simulations, address these limitations by considering real tissue composition and scatter.

Purpose of the Study:

• To systematically review and evaluate TG-43, MBDCAs, and MC methods for brachytherapy dose calculation.
• Assess dosimetric accuracy, validation strategies, computational feasibility, and clinical implementation barriers.

Main Methods:

• A PRISMA-guided literature search of the Scopus database identified 42 relevant studies.
• Eligible studies compared TG-43, MBDCAs, and/or MCs in pelvic, breast, or head-and-neck brachytherapy.
• Data extracted included dosimetric discrepancies, validation methods, computational performance, and workflow integration.

Main Results:

• TG-43 showed variable bias depending on tissue composition and scatter; it often overestimated doses in homogeneous tissues but underestimated in low-density regions.
• MBDCAs and MCs agreed within ~3% with experimental data, offering more reliable radiobiological metrics.
• GPU acceleration and deep learning reduced computation times; standardization, commissioning, and QA remain challenges.

Conclusions:

• A transition to TG-186-compliant, heterogeneity-aware dose calculation frameworks is supported by evidence.
• MBDCAs/MC algorithms offer superior accuracy and are compatible with adaptive and biologically guided planning.
• Established QA standards and benchmarking datasets facilitate implementation, linking precision to patient outcomes.