Dose Calculation Algorithm Comparison

Modern treatment planning systems use different dose calculation algorithms with varying levels of accuracy and computational complexity. Understanding their differences is crucial for optimal treatment planning, especially in heterogeneous media.

Dose Distribution Comparison

0 cm Depth: 10 cm 30 cm

Lateral Profile Comparison

2x2 cm² Field: 10x10 cm² 20x20 cm²

Finite Pencil Beam (FPB) Algorithm

Physical Principle

FPB decomposes the broad beam into multiple narrow pencil beams. Each pencil beam's dose contribution is calculated using pre-computed kernels derived from Monte Carlo simulations in water. The algorithm applies density scaling for heterogeneity corrections.

D(x,y,z) = ∑∑ K(x-x',y-y',z) × Φ(x',y') × CF(ρ)

Where K is the pencil beam kernel, Φ is the fluence, and CF is the density correction factor.

Clinical Scenario Accuracy Deviation from Monte Carlo
Homogeneous water/tissue Excellent < 2%
Lung (low density) Poor 5-15%
Bone interfaces Moderate 3-5%
Small fields (<3x3 cm²) Poor 5-10%
Air cavities Poor 5-20%

Key Limitations

  • Assumes rectilinear particle transport (no lateral scatter modeling)
  • Simple density scaling inadequate for low-density media
  • Cannot account for electron transport at interfaces
  • Inaccurate for small fields due to lack of lateral equilibrium
  • No modeling of neutron dose for high-energy beams (>10 MV)

Clinical Applications

Best suited for:

  • Conventional 3D-CRT in homogeneous regions
  • Breast treatments (relatively homogeneous)
  • Brain treatments (minimal heterogeneity)
  • Large field sizes (>5x5 cm²)

Avoid for:

  • Lung SBRT/SABR
  • Head and neck with air cavities
  • Small field dosimetry
  • Treatments near metal implants

Anisotropic Analytical Algorithm (AAA)

Physical Principle

AAA uses separate modeling of primary photons, scattered photons, and contamination electrons. It employs Monte Carlo-derived energy deposition kernels that are scaled anisotropically based on radiological path length in different directions.

D(r) = ∫∫∫ Ψ(r') × K(r-r',ρ) × A(θ,φ) dV'

Where Ψ is the energy fluence, K is the anisotropic kernel, and A accounts for angular dependencies.

Clinical Scenario Accuracy Deviation from Monte Carlo
Homogeneous water/tissue Excellent < 2%
Lung (low density) Good 2-5%
Bone interfaces Good 2-3%
Small fields (<3x3 cm²) Moderate 3-5%
Air cavities Moderate 3-7%

Key Advantages over FPB

  • Anisotropic scaling accounts for lateral heterogeneities
  • Better modeling of secondary electron transport
  • Improved accuracy in lung and air cavities
  • Separate modeling of scatter components
  • More accurate for IMRT/VMAT optimization

Clinical Applications

Best suited for:

  • IMRT/VMAT planning
  • Lung treatments (non-SBRT)
  • Head and neck treatments
  • Prostate and pelvis

Limitations:

  • Still uses kernel-based approximations
  • Limited accuracy near high-Z materials
  • May overestimate dose beyond low-density regions
  • Calculation time increases with heterogeneity

Acuros XB (External Beam)

Physical Principle

Acuros XB solves the Linear Boltzmann Transport Equation (LBTE) deterministically using the grid-based Boltzmann solver method. It explicitly models the transport of photons and electrons through matter without using kernel approximations.

Ω·∇ψ + σ_t ψ = ∫∫ σ_s(E'→E,Ω'→Ω) ψ dE'dΩ' + S

Where ψ is the angular flux, σ_t is total cross-section, σ_s is scattering cross-section, and S is the source.

Clinical Scenario Accuracy Deviation from Monte Carlo
Homogeneous water/tissue Excellent < 1%
Lung (low density) Excellent < 2%
Bone interfaces Excellent < 2%
Small fields (<3x3 cm²) Excellent < 2%
Air cavities Excellent < 2%
Metal implants Very Good 2-3%

Key Advantages

  • Direct solution of radiation transport equation
  • Explicit modeling of all particle interactions
  • Accurate in all heterogeneous conditions
  • Models electron transport explicitly
  • Can report dose to medium or dose to water
  • Comparable accuracy to Monte Carlo with faster computation

Clinical Applications

Recommended for:

  • Lung SBRT/SABR treatments
  • Small field dosimetry
  • Treatments near metal implants
  • Head and neck with air cavities
  • Any treatment requiring highest accuracy

Considerations:

  • Longer calculation time than AAA (but still clinically practical)
  • Choice between dose-to-medium vs dose-to-water reporting
  • May require different MU for same prescription vs AAA
  • Excellent for plan QA and verification