SU-E-J-114: Automated Processing of Sensitometry Data for Radiation Therapy Treatment Planner and Quality Assurance

Abstract

Automated Processing of Sensitometry Data for Radiation Therapy Treatment Planning and Quality Assurance Purpose: To investigate the efficacy of the calibration of the Hounsfield unit (HU) number to electron density (ED) for a range of traditional and experimental sensitometry targets and optimize the calibration for different sets of targets using automated image‐processing software. The possibility of calibration including virtualization of calcium‐based targets was studied. Methods: Ninety‐four scans of the Catphan 700 phantom were made on three different scanners using different filters, slice thicknesses, and energies. The phantom has ten sensitometry targets ranging in CT number from about ‐ 1000 to +1000 HU, and includes two calcium based bone analogs. For each scan, contrast scale and effective energy calculations were performed using the Image Owl Catphan QA automated image processing system. Separately, ten different calcium based targets of varying density and composition were manufactured and purchased. Theoretical calculations of linear attenuation coefficients of calcium and non‐calcium based targets were performed. Measured and theoretical CT#s were compared for all materials and HU vs. ED calibrations were performed. Results: Measured CT numbers can vary significantly (+/− 20 HU or more for some materials), depending on the scanner, filtration, slice thickness, and energy of the scan, but this variation in HU is not significant with respect to establishing HU vs. ED calibration within acceptable tolerances for planning purposes. The greatest variation in HU is seen in high density and low density materials. Contrast scale and HU vs. ED calibrations including a wide range of calcium‐based bone analogs proved to be robust. Conclusion: Automated image processing of well‐quantified sensitometry targets provides a potentially powerful tool for both accurate and consistent HU to ED calibration for radiation therapy treatment planning as well as statistical evaluation of scanner performance. This research was supported by Image Owl, Inc., and The Phantom Laboratory, Inc.