An Integer Programming Approach to Conversion from Static to Continuous Delivery of Intensity Modulated Radiation Therapy

The aim of this thesis is to generate a starting guess for volumetric modulated arc therapy (VMAT) plan optimization by using data from an intensity modulated radiation therapy (IMRT) plan. We develop a mixed-binary linear programming model, based on choosing segments among a set of predefined segments. The objective is to deliver intensity modulation as similar as possible to the intensity modulation of the IMRT plan. The quality of the solutions is largely dependent on the quality of the predefined segments. However, the model achieves high similarity in intensity modulation when supplied with suitable segments. Unfortunately, high similarity in intensity modulation does not necessarily imply high similarity in dose distribution. In order for the model to generate VMAT plans with acceptable dose distributions the leaf travel between adjacent control points needs to be kept low. The model shows some promising features, but improvements, especially regarding implementation, need to be made in order for the model to be useful.