Strategies for reducing intra-fraction motion induced dosimetric effects in proton therapy

Li Zhao, Purdue University

Abstract

Intra-fraction respiration motion during radiation delivery presents a major challenge to radiation therapy. There has been a growing effort to characterize and manage internal organ motion in radiation therapy, however very few studies focus on tackling this issue in proton therapy. Current practice for treating lung tumors in proton therapy is still to apply population-based margins to account for internal tumor motion, which can lead to target underdosage and normal tissue overdosage. This thesis explores the intra-fraction motion induced dosimetric effects from both computational treatment planning and experimental studies. Four-dimensional CT scans are used to analyze the patient-specific tumor motion characteristics. A feasible method to design the range compensator by using the maximum intensity projection (MIP) images is proposed. Results demonstrate that this MIP approach ensures adequate tumor coverage throughout the entire respiratory cycle whilst maintaining normal tissue dose under clinical constraints. Based on 4D-CT scans, dose convolution is used for assessing the accuracy of Gaussian probability density function for modeling the patient-specific respiratory motion on dose distribution. Non-negligible dose discrepancy is observed in comparisons of convolved dose distributions, and patient-specific respiration PDF is advocated. In addition, an experimental phantom study primarily focusing on the interplay effect between target motion and the scanning beam motion is implemented in two proton beam delivery systems: double scattering and uniform scanning. Measurement results suggest that dose blurring effect is dominant, and interplay effect is trivial in the uniform scanning system due to dose repainting.

Degree

Ph.D.

Advisors

Stewart, Purdue University.

Subject Area

Medical imaging|Nuclear physics|Medicine

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS