Abstract
Traditional model-based image reconstruction (MBIR) methods combine forward and noise models with simple object priors. In this work, we propose a hybrid supervised-unsupervised (SUPER) learning framework for X-ray computed tomography (CT) image reconstruction. The proposed learning formulation leverages both sparsity or unsupervised learning-based priors and neural network reconstructors to simulate a fixed-point iteration process. Each proposed trained block consists of a deterministic MBIR solver and a neural network. The information flows in parallel through these two reconstructors and is then optimally combined, and multiple such blocks are cascaded to form a reconstruction pipeline.
Fig. 1: Overall structure of the proposed Parallel SUPER framework.
We demonstrate the efficacy of this learned hybrid model for low-dose CT image reconstruction with limited training data, where we use the NIH AAPM Mayo Clinic Low Dose CT Grand Challenge dataset for training and testing. In our experiments, we study combinations of supervised deep network reconstructors and sparse representations-based (unsupervised) learned or analytical priors. Our results demonstrate the promising performance of the proposed framework compared to recent reconstruction methods.
Results
Fig. 2: Reconstruction of slice 50 from patient L067 and reconstruction of slice 150 from patient L310 using various methods. The display window is [800, 1200] HU.
References
Chen, Ling, Zhishen Huang, Yong Long, and Saiprasad Ravishankar. "Combining deep learning and adaptive sparse modeling for low-dose CT reconstruction." In 7th International Conference on Image Formation in X-Ray Computed Tomography, vol. 12304, pp. 390-395. SPIE, 2022.
Ye, Siqi, Zhipeng Li, Michael T. McCann, Yong Long, and Saiprasad Ravishankar. "Unified supervised-unsupervised (super) learning for x-ray ct image reconstruction." IEEE Transactions on Medical Imaging 40, no. 11 (2021): 2986-3001.