Image restoration based on projection onto convex sets algorithm for beam modulation CT acquisition

摘要

Recently, the use of a low-dose computed tomography (CT) system to reduce the radiation dose in medical imaging has been actively investigated. Region-of-interest (ROI) imaging is considered an effective method to reduce the exposure dose. Of the various studies of ROI imaging conducted in the medical field, most have not focused on improving the image quality of the entire anatomy. We propose ROI-based beam modulation acquisition to restore the information outside of the ROI. The CT system was simulated using the MAMAS R2017b program. The 3D voxelized phantom was made up of an abdominal DICOM image from a clinical CT system. A total of 360 projection data points were obtained, from which CT reconstructed images were obtained using a filtered back projection (FBP) reconstruction algorithm. The ROI CT images were reconstructed using the 360 truncated projection data points. Beam modulation CT images were reconstructed using 288 truncated and 72 full-size projection data points. An interpolation method and our proposed method based on a projection onto convex sets (POCS) algorithm corrected the truncated projection data. The image quality of three ROIs was evaluated using the structural similarity index measure (SSIM) value. Using beam modulation acquisition, it is possible to restore the external information of a reconstructed image. All ROI-based methods yield high SSIM values, close to 1, within the ROI. The reconstructed image obtained by beam modulation acquisition had a much higher SSIM value for the external information than that obtained by the ROI scan. According to the SSIM value, the proposed method based on a POCS algorithm provides the best image quality in beam modulation acquisition. In conclusion, we have verified the possibility of restoring the ROI external information using beam modulation acquisition. Additionally, we suggested a new restoration method based on a POCS algorithm to obtain better image quality in beam modulation acquisition.