TU‐F‐CAMPUS‐I‐03: Enhancement of 4D CBCT Image Quality Using An Adaptive Prior Image Constrained Compressed Sensing

Abstract

Purpose: To develop an iterative reconstruction algorithm using a compressed sensing with adaptive prior image constraints to solve 4D CBCT reconstruction problem. Methods: The images reconstructed by the FDK algorithm with a full set of unsorted projections are served as prior images for partial projections in each phase group and are utilized as an initial guess. Additionally, the prior images are clustered into several regions by applying intensity‐based thresholding, which is referred to as the segmented prior images. The segmented prior images are employed to detect any possible mismatched areas compared with the target images generated by partial projection data. With these two prior images, our algorithm alternately performs the simultaneous algebraic reconstruction technique and anisotropic total variation regularization while adjusting a weighted relaxation map during the iterative reconstruction process. The weighted relaxation map depends on binary images created by the voxel‐dependent comparison between the segmented prior and segmented target images. For the segmented target images, the k‐means clustering with a geometric weighting is applied on the reconstruction images generated in each iteration step. The inverse values of the distance map converted from binary images are assigned to be the values of the relaxation map. Evaluations using Catphan504 phantom with a motion platform were carried out. Results: Qualitative and quantitative analyses showed that the method provides high‐quality CBCT reconstruction images when compared with those generated by the FDK, CS, and PICCS algorithms, with higher contrast‐to‐noise ratio and faster convergence caused by minimizing data fidelity. Especially, the proposed method was superior to PICCS in the aspect of updating locally‐mismatched region. Conclusion: The proposed method not only improves the image quality of 4D CBCT by adaptive updates during the reconstruction process, but also leads to a lower imaging dose and faster acquisition time by using a regular 3D CBCT scan.

Sparse-View Computed Tomography Reconstruction Based on a Novel Improved Prior Image Constrained Compressed Sensing Algorithm

“…Szczykutowicz et al designed a novel denoising method based on PICCS for dual energy CT [29]. Lee et al presented an adaptive PICCS method to enhance 4D CBCT reconstruction image quality [30]. Essam A. Rashed et al proposed a powerful statistical image reconstruction algorithm for CT for which prior information obtained from a probabilistic atlas is modeled for the CT image reconstruction [31].…”