Evaluation of optimal kilovoltage-cone beam technique on image quality, registration accuracy, time of imaging and relative dose for head radiotherapy: A phantom study

Abstract

Background: Image-guided radiation therapy (IGRT) has improved geometry accuracy of patient positioning in radiotherapy. Nowadays, a kilo-voltage cone-beam computed tomography (kV-CBCT) is widely applied in IGRT to confirm daily patient positioning. However, one problem with IGRT is the increased dose to normal tissue outside the target area. Reduction in the dose of kV-CBCT verification imaging leads to deterioration in image quality and registration results. Objectives: The objective of this study was to evaluate an optimal kV cone-beam technique on image quality, registration accuracy, relative dose, and imaging time using different combinations of angular range, angular separation, and mA (in total eight protocols) in the head phantom. Materials and methods: Catphan® 503 phantom was used to evaluate image quality and a PIXY Anthropomorphic Training/Teaching Phantom was used to verify image registration accuracy. The kV-CBCT imaging was performed using an X-ray volumetric imaging (XVI) system mounted on an Elekta Versa HD linear accelerator. The absorbed dose of kV-CBCT imaging was determined using the head phantom with an ionization chamber. The eight protocols were analyzed for image quality, image registration, relative dose change, and imaging delivery time. Results: Image quality parameter results showed that maximum contrast to noise ratio (CNR) increased by approximately 65% at 100 kV, 20 mA, θ=360°, and Δθ=0.54°, while uniformity compared with 100 kV, 10 mA, θ=200°, Δθ=0.54° (default protocol) varied within 7%. The spatial resolution showed no change (0.167 cm), with geometric distortions of less than 0.2 mm. Image registration errors were within 0.2 cm, with the highest magnitude of error seen in the vertical direction. Imaging dose can be reduced using 100 kV, 10 mA, θ=200°, Δθ=1.09° about 50% and 100 kV, 10 mA, θ=360°, Δθ=1.09° about 15% with similar CNR of default protocol. Imaging time decreased by approximately 2-folds, while Δθ increased by 2-folds. Conclusion: The suggested protocols suitable for optimal kV-CBCT image quality, accuracy of image registration, decreased imaging time, and reduced image dose in the head region were 1) 10 mA, θ=200°, Δθ=1.09° 2) 10 mA, θ=360°, Δθ=1.09°, and 3) 20 mA, θ=360°, Δθ=1.09°. These protocols decreased imaging dose about 50%, 15%, and 2%, respectively. The developing kV-CBCT technique should be counterbalanced by careful consideration of imaging dose, image quality, imaging time, and accuracy of image registration.