Image quality of four-dimensional cone-beam computed tomography obtained at various gantry rotation speeds for liver stereotactic body radiation therapy with fiducial markers. (January 2018)

• Record Type: Journal Article
• Authors: Shimohigashi, Yoshinobu; Araki, Fujio; Maruyama, Masato; Yonemura, Keisuke; Nakaguchi, Yuji; Kai, Yudai; Toya, Ryo
• Is Part Of: Physica medica. Volume 45(2018)
• Page Start: 19
• Page End: 24
• Publication Date: 2018-01
• Subjects: Liver tumor -- Four-dimensional cone-beam computed tomography -- Gantry rotation speed -- Image quality -- Image quality analysis -- Fiducial markers; 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.2017.11.020
• Languages: English
• ISSNs: 1120-1797

Abstract

Highlights: The effect of gantry rotation speed on 4D-CBCT images for liver SBRT was studied. The signal- and contrast-to-noise ratios and structural similarity were measured. The fiducial marker positions were compared with setup and reference values. The gantry rotation speed was found to affect image quality significantly. Gantry rotation speeds of 67 and 85° min −1 were optimal. Abstract: In this study, qualities of 4D cone-beam CT (CBCT) images obtained using various gantry rotation speeds (GRSs) for liver stereotactic body radiation therapy (SBRT) with fiducial markers were quantitatively evaluated. Abdominal phantom containing a fiducial marker was moved along a sinusoidal waveform, and 4D-CBCT images were acquired with GRSs of 50–200° min −1 . We obtained the 4D-CBCT projection data from six patients who underwent liver SBRT and generated 4D-CBCT images at GRSs of 67–200° min −1, by varying the number of projection data points. The image quality was evaluated based on the signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and structural similarity index (SSIM). The fiducial marker positions with different GRSs were compared with the setup values and a reference position in the phantom and clinical studies, respectively. The root mean square errors (RMSEs) were calculated relative to the reference positions. In the phantom study, the mean SNR, CNR, and SSIM decreased from 37.6 to 10.1, from 39.8 to 10.1, and from 0.9 to 0.7, respectively, as the GRS increased from 50 to 200° min −1 . The fiducial marker positions were within 2.0 mm at all GRSs. Similarly, in the clinical study, the mean SNR, CNR, and SSIM decreased from 50.4 to 13.7, from 24.2 to 6.0, and from 0.92 to 0.73, respectively. The mean RMSEs were 2.0, 2.1, and 3.6 mm for the GRSs of 67, 100, and 200° min −1, respectively. We conclude that GRSs of 67 and 85° min −1 yield images of acceptable quality for 4D-CBCT in liver SBRT with fiducial markers.