Unsupervised PET Reconstruction from a Bayesian Perspective

• Published in: 2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI) pp. 1 - 5
• Main Authors: Shen, Chenyu, Xia, Wenjun, Ye, Hongwei, Hou, Mingzheng, Chen, Hu, Liu, Yan, Zhou, Jiliu, Zhang, Yi
• Format: Conference Proceeding
• Language: English
• Published: IEEE 28-03-2022
• Subjects: Bayes methods; Bayesian inference; deep image prior; Gaussian noise; Monte Carlo methods; Noise reduction; PET image reconstruction; regularization by denoising; Representation learning; Training; unsupervised learning; Visual effects
• ISSN: 1945-8452
• DOI: 10.1109/ISBI52829.2022.9761540

Positron emission tomography (PET) reconstruction becomes an ill-posed inverse problem due to the low-count projection data (sinogram). In this paper, we leverage DeepRED from a Bayesian perspective to reconstruct PET image from a single corrupted sinogram without any supervised or auxiliary in-formation. DeepRED is a typical representation learning that combines deep image prior (DIP) and regularization by de-noising (RED) to mitigate the overfitting of network training. Instead of the conventional denoisers usually used in RED, DnCNN-like denoiser, which can constrain the DIP adaptively and facilitate the derivation, is employed. Moreover, stochastic gradient Langevin dynamics (SGLD) is utilized to approximate the Markov chain Monte Carlo (MCMC) sampler. Specifically, Gaussian noise is injected into the gradient updates to further relieve the overfitting. Experimental studies on whole-body dataset demonstrate that our proposed method can achieve better performance compared to several classic and state-of-the-art methods in both qualitative and quantitative aspects.