High-Resolution Reconstruction of FMT Based on Elastic Net Optimized by Relaxed ADMM
文献类型:期刊论文
作者 | Yan, Daxiang; An, Yu; Li, Guanghui; Li, Jiaqian; Du, Yang; Tian, Jie |
刊名 | IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING |
出版日期 | 2022-07-12 |
页码 | / |
DOI | 10.1109/TBME.2022.3190049 |
英文摘要 | Fluorescence Molecular Tomography (FMT), providing the three-dimensional fluorescent distribution information of specific molecular probes in tumors, is widely applied to detect in vivo tumors. However, the ill-posedness of reconstruction greatly affects the resolution of FMT. Traditional methods have introduced different regularization terms to solve this problem, but there are still challenges for the high-resolution reconstruction of small tumors under complex conditions. In this paper, we proposed an elastic net method optimized by the relaxed Alternating Direction Method of Multipliers (EN-RADMM) to improve the reconstruction resolution for small tumors. The objective function consisted of the Least-Square term and elastic net regularization. Relaxation, equivalent deformation directing at ill-posed equations, and LU decomposition were applied to accelerate algorithm convergence and improve solution accuracy. Thereby, the light from small tumors can be precisely reconstructed. We designed a series of digital tumor models with different distances, sizes, and shapes to verify the performance of EN-RADMM, and utilized the real glioma-bearing mouse models to further verify its feasibility and accuracy. The simulation results demonstrated that EN-RADMM can achieve significantly higher resolution and reconstruction accuracy of morphology and position with less time compared with other advanced methods. Furthermore, in vivo experiments proved the broad prospect of EN-RADMM in pre-clinical application of FMT reconstruction. |
源URL | [http://ir.ia.ac.cn/handle/173211/50908] |
专题 | 自动化研究所_中国科学院分子影像重点实验室 |
通讯作者 | An, Yu; Du, Yang; Tian, Jie |
推荐引用方式 GB/T 7714 | Yan, Daxiang,An, Yu,Li, Guanghui,et al. High-Resolution Reconstruction of FMT Based on Elastic Net Optimized by Relaxed ADMM[J]. IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING,2022:/. |
APA | Yan, Daxiang,An, Yu,Li, Guanghui,Li, Jiaqian,Du, Yang,&Tian, Jie.(2022).High-Resolution Reconstruction of FMT Based on Elastic Net Optimized by Relaxed ADMM.IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING,/. |
MLA | Yan, Daxiang,et al."High-Resolution Reconstruction of FMT Based on Elastic Net Optimized by Relaxed ADMM".IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING (2022):/. |
入库方式: OAI收割
来源:自动化研究所
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。