Fair evaluation of different sparse-view CT reconstruction models(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2025年第6期
- Page:
- 796-800
- Research Field:
- 医学影像物理
- Publishing date:
Info
- Title:
- Fair evaluation of different sparse-view CT reconstruction models
- Author(s):
- CAO Ximing; WEN Menghuang; MA Jianhua; BIAN Zhaoying
- School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China
- Keywords:
- computed tomography; sparse-view reconstruction; deep learning; model evaluation
- PACS:
- R318
- DOI:
- DOI:10.3969/j.issn.1005-202X.2025.06.013
- Abstract:
- Objective To evaluate the performance of reconstruction networks with different sparse views under the conditionof keeping the same number of model parameters. Methods The number of network channels and network layers wereadjusted to make the parameter quantity of each network similar when keeping the structure of each image-domain networkand dual-domain network unchanged. The reconstruction performance of each network at different sparsity levels wascompared. The AAPM Low-Dose CT Grand Challenge datasets were used in the experiment, including 10 976 images fortraining, 979 images for validation, and 4 256 images for testing. The performance of each model was evaluated visually incombination with objective metrics such as peak signal-to-noise ratio, structural similarity and root mean square error.Results Before adjusting the model parameters, the hybrid domain network Tensor-Net obtained the best visual evaluationand objective evaluation metrics. After parament adjustment, with a similar number of parameters, Tensor-Net outperformedthe other models at various projection angles in image anatomical detail recovery, but its structural similarity was slightlylower than that of RED-CNN. The parameters of the hybrid domain model Dual-FBPConvNet were all worse than those ofFBPConvNet. Conclusion The hybrid domain model is advantageous in sparse-view CT reconstruction, but it faces moreserious overfitting problems. Using a larger image domain model can achieve results similar to those of hybrid domain model.
Last Update: 2025-07-01