Image-Guided Treatment Planning for Photodynamic Therapy

Monte Carlo calculation of the optical fluence in the vicinity of a single, cylindrical diffusing optical fiber embedded in a tissue-simulating medium (left), and the fluence distribtion resulting from 4 fibers embedded in a region of interest derived from computed tomography images of a head and neck cancer patient (right).

Photodynamic therapy is an emerging approach to the treatment of many cancers and local microbial infections. In order to deliver light to tumors deep within the body, optical fibers are advanced into the tissue under imaging guidance. The Foster lab is developing computational treatment plans for these tumors, in which light is propagated from the fibers into the surrounding tissue using Monte Carlo simulation.

The treatment field is derived from magnetic resonance or computed tomography images. The clinical image data are imported into the simulation space, where the fibers are embedded at user-specified locations and orientations. The resulting optical fluence maps may be compared with treatment plans developed for ionizing radiation therapy. We are also working on methods for obtaining estimates of the tissue optical properties from interstitial measurements. The use of custom Monte Carlo codes written for execution on graphical processor units reduces computation time significantly and will allow for these treatment plans to be developed and revised quickly enough to be useful clinically.