Radiotherapy Dose Calculation Accuracy: Monte Carlo vs Collapsed Cone at Air-Tissue Boundaries

This study establishes that the Photon Monte Carlo (pMC) method offers superior precision over Collapsed Cone Convolution (CCC) for surface dose calculations. Historically, achieving reliable radiotherapy dose calculation accuracy near air cavities—such as the throat in glottic cancer—has been a persistent technical hurdle. The abrupt transition from low-density air to high-density tissue creates a state of electronic disequilibrium, a physical phenomenon that frequently exposes the limitations of standard mathematical models used in treatment planning.

These results were observed under controlled laboratory conditions, so real-world performance may differ.

The investigation utilised a phantom model designed to simulate early-stage glottic cancer. Researchers introduced an air gap between water-equivalent blocks and delivered 6 MV X-rays at 200 MU. To establish a ground truth, the absorbed dose was measured physically using Gafchromic film. These readings were then benchmarked against the predictions of two widely used algorithms: Pinnacle3 (utilising CCC) and RayStation (utilising both CCC and pMC).

It is vital to distinguish the operational mechanics of the two algorithms compared here. Collapsed Cone Convolution (CCC) functions by approximating energy transport along discrete lines or 'cones' radiating from the source. It assumes a certain level of lateral equilibrium to speed up processing, which works well in solid tissue but falters when density drops precipitously, as in an air-filled trachea. In contrast, the Photon Monte Carlo (pMC) method adopts a statistical approach, simulating the random trajectories and interactions of millions of individual photons and electrons. Where CCC relies on analytical shortcuts to estimate scatter, pMC computationally tracks the probability of particle interaction, allowing it to better account for the erratic electron behaviour found in void spaces.

Implications for Radiotherapy Dose Calculation Accuracy

The results highlighted a substantial divergence in performance. The RayStation CCC algorithm deviated from the film measurements by up to 14.4%. This indicates a significant overestimation of the dose delivered to the tissue surface. Conversely, the RayStation pMC method maintained a difference of -3.8%. While pMC slightly underestimated the dose, its error margin was considerably tighter than the overestimation observed with CCC.

For clinical practice, these findings suggest that using CCC for plans involving the larynx or other air-filled cavities could lead to unintended under-dosing. If the system calculates a higher dose than is physically deposited, the tumour control probability may be compromised. The authors conclude that when the dose distribution around heterogeneous regions is clinically significant, the computational rigour of RayStation pMC is required to ensure patient safety.