Treatment plan evaluation for interstitial photodynamic therapy in a mouse model by Monte Carlo simulation with FullMonte

Abstract

Monte Carlo (MC) simulation is recognized as the "gold standard" for biophotonic simulation, capturing all relevant physics and material properties at the perceived cost of high computing demands. Tetrahedral-mesh-based MC simulations particularly are attractive due to the ability to refine the mesh at will to conform to complicated geometries or user-defined resolution requirements. Since the chosen MC method does not inherently require approximations on the materials, it is applicable to the broadest set of biophotonic simulation problems. MC methods also have other implementation features including inherent parallelism, and permit a continuously-variable quality-runtime tradeoff. We demonstrate here a complete MC-based prospective fluence dose evaluation system for interstitial PDT to generate dose-volume histograms on a tetrahedral mesh geometry description. To our knowledge, this is the first such system for general interstitial photodynamic therapy employing MC methods and is therefore applicable to a very broad cross-section of anatomy and material properties. We demonstrate that evaluation of dose-volume histograms is an effective variance-reduction scheme in its own right which greatly reduces the number of packets required and hence runtime required to achieve acceptable result confidence. We conclude that MC methods are feasible for general PDT treatment evaluation and planning, and considerably less costly than widely believed.

Interstitial Photodynamic Therapy—A Focused Review

“…Several groups developed treatment planning and dosimetry systems to administer I-PDT [ 9 , 14 , 15 , 17 , 18 , 39 , 40 , 46 , 47 , 48 , 49 ]. These plans simulate light propagation in 3-D geometries mimicking the anatomy of target tumor and adjacent structures.…”

Photodynamic therapy for glioblastoma: A preliminary approach for practical application of light propagation models

“…The improvement in parallel GPU computing that drastically decreases the computing time ensured strong enthusiasm for developing the Monte‐Carlo method for PDT planning . With the assistance of automatic tissue segmentation methods , Monte‐Carlo can model the fluence rate according to optical coefficients for each voxel of the MRI volume used for planning.…”

Summary of numerical analyses for therapeutic uses of laser-activated gold nanoparticles

“…Utilising Hadoop, an open-source MapReduce framework was reported in [33] on Amazon's elastic compute cloud. The MC code was also imported to Intel's many-integrated core (MIC) architecture [34], or to field-programmable gate arrays (FPGA) [35]. Doronin et al developed a fast peerto-peer MC code that allowed multiple users to simultaneously access the online MC simulation [36].…”