A Novel Stacked Denoising Autoencoder-Based Reconstruction Framework for Cerenkov Luminescence Tomography

Abstract

Cerenkov luminescence tomography (CLT) is a promising imaging modality in the field of optical molecular imaging (OMI), which successfully bridges the OMI and tradition nuclear medical imaging and provides the location and quantitative analysis of the distribution of radionuclide probes inside the biological objects. As the CLT is an inherent highly ill-posed inverse problem, it is still a challenge to obtain an accurate reconstruction result. Here, we proposed a novel reconstruction framework based on stacking denoising autoencoders (SDAE), which serve as one famous structure of the artificial neural network (ANN). In our framework, the initial permission region is the whole domain and then a traditional reconstruction algorithm is used to reconstruct each node's energy. Then these nodes are clustered into two regions: permission region and non-permission region, and the permission region is used to start a new reconstruction loop where a new result can be obtained. The procedures above are repeated before the result meets the stop conditions. The numerical simulation experiments, physical phantom experiments and in vivo experiments are all carried out to validate the feasibility and potential of our framework. Results demonstrate that the proposed framework can indeed achieve a good performance in CLT reconstruction. INDEX TERMS Cerenkov luminescence tomography, optical molecular imaging, reconstruction algorithm, stacking denoising autoencoders.

ABPO-TVSCAD: alternating Bregman proximity operators approach based on TVSCAD regularization for bioluminescence tomography

“…Because the deep learning methods train on the large dataset and establish the mapping from input to output, it can avoid the inaccuracy of the forward photon propagation modeling and ill-posed inverse problem (Li et al 2020). However, its generalization ability is weak, so the trained artificial neural network can only be applied to specific imaging objects (Cao et al 2019).…”

A Multilevel Probabilistic Cerenkov Luminescence Tomography Reconstruction Framework Based on Energy Distribution Density Region Scaling

“…The first radiographic imaging using CR was made in 2009 and is known as Cerenkov luminescence imaging (CLI) (2). Since its inception, CLI has been widely used in surgical guidance, drug development, endoscopic imaging, tumor detection, and other fields (3)(4)(5)(6)(7). The most significant advantage of CLI over other optical imaging methods is that it can use many approved radioactive sources for clinical imaging (8)(9)(10).…”

A robust elastic net-ℓ 1 ℓ 2 reconstruction method for x-ray luminescence computed tomography

“…In order to investigate the potential and feasibility of the Elastic net-ℓ 1 ℓ 2 algorithm, three representative regularizers, including ℓ 1 , ℓ 2 and ℓ 1−2 , are used for comparison, which are respectively incomplete variables truncated conjugate (IVTCG-ℓ 1 ) (He et al 2010), tikhonov regularization algorithm (Tikhonov-ℓ 2 )(Cao et al 2013) and dierence of convex algorithm(DCA-ℓ 1−2 ). By quantitative evaluation of various algorithms on image quality, the energy location error (LE)(He et al 2015), normalized root-mean-square error (NRMSE)(Guo et al 2015), fluorescent yield relative error (FYRE)(Yi et al 2018), contrast-to-noise ratio (CNR)(Cao et al 2019) and Dice are described as quantitative evaluation indexes.…”