Faculty Collaboration Database

Development of an inverse optimization package to plan nonuniform dose distributions based on spatially inhomogeneous radiosensitivity extracted from biological images. Med Phys 2007 Apr;34(4):1198-205

Date

05/16/2007

Pubmed ID

17500451

DOI

10.1118/1.2710948

Scopus ID

2-s2.0-34147140787 (requires institutional sign-in at Scopus site)   22 Citations

Abstract

An inverse optimization package which is capable of generating nonuniform dose distribution with subregional dose escalation is developed to achieve maximum equivalent uniform dose (EUD) for target while keeping the critical structure doses as low as possible. Relative cerebral blood volume (rCBV) maps obtained with a dynamic susceptibility contrast-enhanced MRI technique were used to delineate spatial radiosensitivity distributions. The voxel rCBV was converted to voxel radiosensitivity parameters (e.g., alpha and alpha/beta) based on previously reported correlations between rCBV, tumor grade, and radiosensitivity. A software package, DOSEPAINT, developed using MATLAB, optimizes the beamlet weights to achieve maximum EUD for target while limiting doses to critical structures. Using DOSEPAINT, we have generated nonuniform 3D-dose distributions for selected patient cases. Depending on the variation of the pixel radiosensitivity, the subregional dose escalation can be as high as 35% of the uniform dose as planned conventionally. The target dose escalation comes from both the inhomogeneous radiosensitivities and the elimination of integral target dose constraint. The target EUDs are found to be higher than those for the uniform dose planned ignoring the spatial inhomogeneous radiosensitivity. The EUDs for organs at risk are found to be approximately equal to or lower than those for the uniform dose plans. In conclusion, we have developed a package that is capable of generating nonuniform dose distributions optimized for spatially inhomogeneous radiosensitivity. Subregional dose escalation may lead to increased treatment effectiveness as indicated by higher EUDs. The current development will impact biological image guided radiotherapy.

Author List

Chen GP, Ahunbay E, Schultz C, Li XA

Authors

Ergun Ahunbay PhD Professor in the Radiation Oncology department at Medical College of Wisconsin
Guang-Pei Chen PhD Associate Professor in the Radiation Oncology department at Medical College of Wisconsin
Christopher J. Schultz MD Chair, Professor in the Radiation Oncology department at Medical College of Wisconsin

MESH terms used to index this publication - Major topics in bold

Algorithms
Anisotropy
Brain Neoplasms
Humans
Image Interpretation, Computer-Assisted
Magnetic Resonance Imaging
Organ Specificity
Radiation Tolerance
Radiotherapy Dosage
Radiotherapy Planning, Computer-Assisted
Radiotherapy, Computer-Assisted
Radiotherapy, Conformal
Software