An anatomically realistic lung model for Monte Carlo-based dose calculations.

Area of Science:

• Medical Physics
• Radiotherapy
• Computational Biology

Background:

• Accurate geometric description is crucial for radiotherapy planning, particularly in the lung due to its complex, heterogeneous structure.
• Current treatment planning often uses simplified lung models, such as the mean density model, which may not fully capture anatomical variations.

Purpose of the Study:

• To introduce a novel heterogeneous random lung model based on human lung morphology.
• To compare dose calculations between this new model and the conventional mean density model.
• To assess the impact of lung heterogeneity on radiotherapy dose distributions.

Main Methods:

• Development of a heterogeneous random lung model incorporating morphological data.
• Utilized Monte Carlo methods for accurate dose calculations.
• Compared dose distributions from the random lung model against the mean density model using various beam energies and field sizes.
• Preliminary comparison with a voxelized CT-based model.

Main Results:

• Significant dose differences (up to 34% of D(max)) were observed between the random and mean density lung models, particularly with small field sizes (1x1 cm²) and high energy beams (18 MV).
• A distal 'shadow' region with dose reduction (up to 7% of D(max)) was identified.
• Lung heterogeneity altered isodose line shapes, affecting penumbra size.
• Mean lung doses were sensitive to the relative positions of structures within the heterogeneous model for small fields.

Conclusions:

• The mean density lung model is inadequate for accurate dose calculations in radiotherapy, especially when using small field sizes.
• A more sophisticated CT-based lung model is necessary for precise treatment planning.
• The developed random lung model provides a valuable tool for assessing CT-based treatment plan accuracy.