A linear optimization model for high dose rate brachytherapy using a novel distance metric.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• High dose rate brachytherapy (HDR-BT) is a crucial cancer treatment modality.
• Current optimization models for HDR-BT often lack anatomical precision.
• There is a need for advanced models to improve treatment plan quality.

Purpose of the Study:

• To introduce a linear network-based optimization model (LNBM) for HDR-BT.
• To utilize a novel distance metric for dose discrepancy in HDR-BT.
• To leverage voxel adjacency structures for improved treatment planning.

Main Methods:

• Application of LNBM to 7 cervical cancer HDR-BT cases.
• Comparison with inverse planning by simulated annealing (IPSA).
• Evaluation using tumor coverage, OAR dosimetric indices, and novel homogeneity metrics (hot-spot volumes/diameters).

Main Results:

• LNBM demonstrated superior tumor coverage compared to IPSA.
• Improved isodose contour plots and dosimetric indices for organs at risk (bladder, rectum).
• LNBM generated more homogeneous treatment plans, confirmed by new metrics.

Conclusions:

• The proposed LNBM efficiently generates high-quality, homogeneous HDR-BT treatment plans.
• LNBM's network-based approach effectively utilizes anatomical structures for optimized planning.
• This model offers a significant advancement over traditional penalty-based methods in HDR-BT.