Development of portable image-guidance system (PIGS) for preclinical small animal experiments with the cone-beam irradiation of X-RAD

The purpose of this study is to develop a portable image-guidance system (PIGS) for preclinical experiments to identify anatomical structures of small animals and deliver beams accurately to their target locations more precisely. The PIGS consists of a gadolinium oxide (Gd 2 O 3 ) scintillator, a transparent acrylic, a thin mirror, opaque polymethyl methacrylate (PMMA), and a charge-coupled device (CCD) camera. In addition, the PIGS can reduce dose-measurement effects using tissue-equivalent as the principal. Using 320 X-RAD devices, the X-ray beam was irradiated with 130~35 kVp /12.5~45.0 mA of tube parameters. With 3D-printed grids and homemade mouse phantom, we evaluated the accuracy between the actual position and the location in the imaging coordinate. The image quality of the actual mouse's anatomical images was also examined by setting a 5.0-mm region of interest (ROI) on images, such as the actual mouse Lt. brain, Lt. lung and Lt. femur for each energy. Furthermore, the EBT3 film was used to compare the solid water phantom and PIGS with the dose measurements at a depth of 5.0 mm to confirm any differences. Overall, PIGS showed excellent results in evaluating the accuracy and image quality of image coordinates of small animals. The accuracy of the image coordinates showed error values below the moving average ± 0.2 mm. Furthermore, the mouse imaging results showed the best anatomical image results with high CNR values and low SNR values at 35 kVp/45 mA conditions. COV was the best at 0.0028 in non-filtered 35 kVp/40 mA conditions. Overall, NPS showed similar patterns without significant differences in all energy conditions but had less area noise with lower spatial frequency under lower energy conditions. In addition, the dose evaluation of PIGS and solid water phantom using EBT3 showed a Dmax error value of within 0.2% at a depth of 5 mm measured. This study showed that PIGS could accurately identify the main anatomical structures of mice to deliver radiation beams to the targets. In addition, we proved that the application of PIGS in 320 X-RAD causes no dosimetric influence.