Optimizing Fluence Maps to Optimize Target Dose

Optimizing Fluence Maps to Optimize Target Dose

Moitvation

My research addressed optimization models for radiation therapy treatment planning and patient scheduling. In intensity modulated radiation therapy (IMRT) treatment planning problems, I use direct aperture optimization (DAO) that explicitly formulates the °uence map optimization (FMO) problem as a convex optimization problem in terms of all multileaf collimator (MLC) deliverable apertures and their associated intensities and solve it using column generation method. In addition, the interfraction motion has been incorporated to the stochastic-programming based FMO and DAO models. Optimization models for patient scheduling problems in proton therapy delivery have also been studied in this research (Trofimov et.al, 2005).

Background of Radiation relevant to your device

Patients receive radiotherapy via a linear accelerator th at rotates on a gantry around the patient, emitting beams of X-rays from various angles. Relatively low-energy beams are “shot” from the different angles to produce a superposition of dose absorp­ tion at the beam intersection. This eliminates the need to shoot the patient with one immensely powerful beam to kill the tum or and thus prevents m ajor damage to the skin and lowers dose to other healthy tissue as well. Note that one can have many targets (tumors) and many organs-at-risk (critical organs) in the region of interest. Moreover, a beam with rectangular cross-section yields less than adequate results.

Consider a spherical target. We would prefer a beam with a circular field shape so that we can hit the target shape precisely, without letting any more radiation bleed over to the critical organs than necessary. Similarly, if the tum or is next to a critical organ we want to protect, it might behoove us to have a beam whose cross-section was shaped so as to avoid the organ-at-risk, while matching the shape of the tumor. This consideration led to the development of the multi-leaf collimator (MLC) device. Its computer-controlled leaves act as a filter, blocking or allowing radiation through in order to tailor the beam field to the shape of the tum or and minimize exposure of the critical organs. Purpose of your device/process (Sodertrom, & Brahme, 1993).

Theory

The IMRT fluence map optimization problem has been extensively studied for a number of years. Several classes of optimization models have been proposed for fluence optimization. Linear models have long been used in this problem. They offered very satisfactory theoretical results at the cost of a large number of constraints and variables. The fluence map optimization problem also can be viewed as mixed integer programming (MIP). The challenge of using MIP modeling for IMRT is that the resulting instances are very large-scale, and since general MIP is NP-hard, specialized algorithms designed to solve IMRT instances are required. If an objective function is associated with each anatomical structure, then this problem can be naturally viewed as a multi-objective optimization problem. In multi-objective optimization, many different objective functions have been used for the fluence map problem in IMRT. In fact, many of these seemingly different approaches ...