SU‐E‐J‐190: Characterization of Radiation Induced CT Number Changes in Tumor and Normal Lung During Radiation Therapy for Lung Cancer

Abstract

Purpose:To measure CT number (CTN) changes in tumor and normal lung as a function of radiation therapy (RT) dose during the course of RT delivery for lung cancer using daily IGRT CT images and single respiration phase CT images.Methods:4D CT acquired during planning simulation and daily 3D CT acquired during daily IGRT for 10 lung cancer cases randomly selected in terms of age, caner type and stage, were analyzed using an in‐house developed software tool. All patients were treated in 2 Gy fractions to primary tumors and involved nodal regions. Regions enclosed by a series of isodose surfaces in normal lung were delineated. The obtained contours along with target contours (GTVs) were populated to each singlephase planning CT and daily CT. CTN in term of Hounsfield Unit (HU) of each voxel in these delineated regions were collectively analyzed using histogram, mean, mode and linear correlation.Results:Respiration induced normal lung CTN change, as analyzed from single‐phase planning CTs, ranged from 9 to 23 (±2) HU for the patients studied. Normal lung CTN change was as large as 50 (±12) HU over the entire treatment course, was dose and patient dependent and was measurable with dose changes as low as 1.5 Gy. For patients with obvious tumor volume regression, CTN within the GTV drops monotonically as much as 10 (±1) HU during the early fractions with a total dose of 20 Gy delivered. The GTV and CTN reductions are significantly correlated with correlation coefficient >0.95.Conclusion:Significant RT dose induced CTN changes in lung tissue and tumor region can be observed during even the early phase of RT delivery, and may potentially be used for early prediction of radiation response. Single respiration phase CT images have dramatically reduced statistical noise in ROIs, making daily dose response evaluation possible.