The application proposes to verify the applicability of a newly devised method of cancer therapy inverse treatment planning to clinical practice. Conformal radiation therapy planning is best carried out as the solution to an inverse problems: Given a set of required doses, data on the patient's anatomy and the characteristics of all the available beams, it calculates the optimum beam delivery parameters. The new method is based on a Dynamically Penalized Likelihood method that allows the therapist to specify tumor and sensitive tissue doses with the assurance that the resulting treatment plan will deliver a uniform dose to the tumor, without significant cold spots, and a much lower specified dose to the sensitive tissues. Specifically, this application proposes to carry out initial testing of the new methodology in a clinical environment, develop a faster version of the algorithm, compare it to other competing methods and, finally, carry out initial treatment planning tests with true patient data. The methodology to be tested can provide a strong impetus to precise treatment planning of difficult tumor/sensitive tissue configurations using multiple x-ray pencil beams by virtue of the robustness, flexibility and simplicity of adjustment of the methodology.
The successful completion of the proposed work would prepare the Principal Investigator for a Phase II SBIR application and commercialization of the DPL methodology in the form of computer programs to interface with current treatment planning software. As modulated x-ray pencil beam delivery methods become available from the large instrument manufacturers for treatment of difficult tumor configurations, successful, reliable programs for inverse treatment planning will become essential.
