Effective atomic number estimation using kV-MV dual-energy source in LINAC. (July 2017)

• Record Type: Journal Article
• Authors: Sakata, Dousatsu; Haga, Akihiro; Kida, Satoshi; Imae, Toshikazu; Takenaka, Shigeharu; Nakagawa, Keiichi
• Is Part Of: Physica medica. Volume 39(2017)
• Page Start: 9
• Page End: 15
• Publication Date: 2017-07
• Subjects: GEANT4 -- Effective atomic number -- Dual-energy CT -- Cone-beam CT -- kV-MV; Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2017.06.010
• Languages: English
• ISSNs: 1120-1797

Abstract

Highlights: A large energy gap dual-energy CT reconstruction was proposed by using a linear accelerator radiotherapy system. With X-ray energy spectrum simulated by GEANT4, the calibration for reconstructed images was performed. The reconstructed effective atomic number (EAN) was compared with that from an empirical formula. Agreement was within 11% and 3%, for calibration and validation phantoms, respectively. Abstract: Dual-energy computed tomography (DECT) imaging can measure the effective atomic number (EAN) as well as the electron density, and thus its adoption may improve dose calculations in brachytherapy and external photon/particle therapy. An expanded energy gap in dual-energy sources is expected to yield more accurate EAN estimations than conventional DECT systems, which typically span less than 100 kV. The aim of this paper is to assess a larger energy gap DECT by using a linear accelerator (LINAC) radiotherapy system with a kV X-ray imaging device, which are combined to provide X-rays in both the kV- and MV-energy ranges. Traditionally, the EAN is determined by parameterising the Hounsfield Unit; however, this is difficult in a kV-MV DECT due to different uncertainties in the reconstructed attenuation coefficient at each end of the energy spectrum. To overcome this problem, we included a new calibration step to produce the most likely linear attenuation coefficients, based upon the X-ray spectrum. To determine the X-ray spectrum, Monte Carlo calculations using GEANT4 were performed. Then the images were calibrated using information from eight inserts of known materials in a CIRS phantom (CIRS Inc., Norfolk, VA). Agreement between the estimated and empirical EANs in these inserts was within 11%. Validation was subsequently performed with the CatPhan500 phantom (The Phantom Laboratory, Salem). The estimated EAN for seven inserts agreed with the empirical values to within 3%. Accordingly, it can be concluded that, given properly reconstructed images based upon a well-determined X-ray spectrum, kV-MV DECT provides an excellent prediction for the EAN.