.png){kind=icon}
.png){kind=icon}
Return
Low-complexity reconstruction of low-dose spectral CT via double low-rank tensor factorization with adaptive transforms
.png){kind=icon}
PRE
AI
C
T
H
D
B
J
DOI：10.1016/j.media.2026.104119.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• The research focuses on a low-complexity method to reduce the radiation dose and noise of spectral CT images while ensuring their imaging quality. • A spatial factor denoising method under the framework of double low-rank tensor factorization is proposed for sparse-view spectral CT image reconstruction. • A proximal alternating minimization (PAM) algorithm is developed to efficiently solve the proposed NDLRTF model, and its global convergence to the critical point is theoretically proved. • Numerical experiments on simulations and clinical patient datasets show that the proposed NDLRTF method outperforms existing popular algorithms.
Keywords:
spectral CT
low-dose imaging
double low-rank tensor factorization
denoising
proximal alternating minimization
Journal
IF:
11.8
Papers:
3.7K
Citations:
2.4W
Organization
Cited Papers