Use of Split Bregman denoising for iterative reconstruction in fluorescence diffuse optical tomography

Abstract

Abstract. Fluorescence diffuse optical tomography (fDOT) is a noninvasive imaging technique that makes it possible to quantify the spatial distribution of fluorescent tracers in small animals. fDOT image reconstruction is commonly performed by means of iterative methods such as the algebraic reconstruction technique (ART). The useful results yielded by more advanced l 1 -regularized techniques for signal recovery and image reconstruction, together with the recent publication of Split Bregman (SB) procedure, led us to propose a new approach to the fDOT inverse problem, namely, ART-SB. This method alternates a cost-efficient reconstruction step (ART iteration) with a denoising filtering step based on minimization of total variation of the image using the SB method, which can be solved efficiently and quickly. We applied this method to simulated and experimental fDOT data and found that ART-SB provides substantial benefits over conventional ART.

High-Performance Fluorescence Molecular Tomography through Shape-Based Reconstruction Using Spherical Harmonics Parameterization

“…Hence, it has the advantage of requiring fewer assumptions on the shape geometry and the weakness of not reducing the unknown dimensions. In recent years, TV has seen advances for FMT and demonstrated its superiority over traditional L 2 regularization in background-free cases [10], [11], [12]; future progress may demonstrate its effectiveness for low fluorescence contrast conditions. Since TV problem is highly nonlinear, its performance depends on the developed solution algorithm and selected regularization parameters.…”

“…Information on the sparse distribution is utilized through different compressed sensing techniques [7], [8]. Edge enhancement priors are utilized by penalizing the fluorescence intensity gradient as a regularized term, such as in the total variation method [9], [10], [11], [12]. The development of multimodality FMT systems [13], [14] has boosted the fusion of information derived from anatomical structures [15], [16].…”

Effects of refractive index mismatch in optical CT imaging of polymer gel dosimeters

“…The images reconstructed by ART and SIRT algorithms, without any denoising, tend to be noisy when there is noise in the projection data. 22 The performance of the proposed method is analyzed by reconstructing images using noisy simulated measurement data. Noisy measurements are generated by adding varying levels (1% and 5%) of Poisson noise to the sinogram of the simulated phantom.…”

“…For denoising, the split Bregman anisotropic total variation (SB-ATV) penalty on the ART reconstructed µ image is minimized. [22][23][24] This solution is termed as SB-ART reconstruction. 22 The regularization parameter α depends on the image and the level of noise in the image.…”

Improved sparse reconstruction for fluorescence molecular tomography with L_1/2 regularization

“…Although L 0 -norm is an ideal sparsity regularizer, reconstruction based on L 0 -norm regularization involves a problem of combinatory optimization, which makes it inefficient or unfeasible in practical applications. Inspired by the theories of compressive sensing and sparse signal recovery, researchers resort to L 1 -norm based reconstruction methods and produce better results over L 2 -norm based methods in scenarios of recovering small localized targets [18][19][20][21][22][23]. However, solutions of L 1 -norm methods are still less sparse and inefficient when heavytail distribution errors exist in the measurements [24].…”