Xiaowei He

Adaptive hp finite element method for fluorescence molecular tomography with simplified spherical harmonics approximation

Recently, the simplified spherical harmonics equations (SPN) model has attracted much attention in modeling the light propagation in small tissue geometries at visible and near-infrared wavelengths. In this paper, we report an efficient numerical method for fluorescence molecular tomography (FMT) that combines the advantage of SPN model and adaptive hp finite element method (hp-FEM). For purposes of comparison, hp-FEM and h-FEM are, respectively applied to the reconstruction process with diffusion approximation and SPN model. Simulation experiments on a 3D digital mouse atlas and physical experiments on a phantom are designed to evaluate the reconstruction methods in terms of the location and the reconstructed fluorescent yield. The experimental results demonstrate that hp-FEM with SPN model, yield more accurate results than h-FEM with diffusion approximation model does. The phantom experiments show the potential and feasibility of the proposed approach in FMT applications.

Sparse reconstruction for fluorescence molecular tomography via a fast iterative algorithm

Fluorescence molecular tomography (FMT) is a fast-developing optical imaging modality that has great potential in early diagnosis of disease and drugs development. However, reconstruction algorithms have to address a highly ill-posed problem to fulfill 3D reconstruction in FMT. In this contribution, we propose an efficient iterative algorithm to solve the large-scale reconstruction problem, in which the sparsity of fluorescent targets is taken as useful a priori information in designing the reconstruction algorithm. In the implementation, a fast sparse approximation scheme combined with a stage-wise learning strategy enable the algorithm to deal with the ill-posed inverse problem at reduced computational costs. We validate the proposed fast iterative method with numerical simulation on a digital mouse model. Experimental results demonstrate that our method is robust for different finite element meshes and different Poisson noise levels.

Effective and robust approach for fluorescence molecular tomography based on CoSaMP and SP3 model

Fluorescence molecular tomography (FMT) allows the detection and quantification of various biological processes in small animals in vivo, which expands the horizons of pre-clinical research and drug development. Efficient three-dimensional (3D) reconstruction algorithm is the key to accurate localization and quantification of fluorescent target in FMT. In this paper, 3D reconstruction of FMT is regarded as a sparse signal recovery problem and the compressive sampling matching pursuit (CoSaMP) algorithm is adopted to obtain greedy recovery of fluorescent signals. Moreover, to reduce the modeling error, the simplified spherical harmonics approximation to the radiative transfer equation (RTE), more specifically SP3, is utilized to describe light propagation in biological tissues. The performance of the proposed reconstruction method is thoroughly evaluated by simulations on a 3D digital mouse model by comparing it with three representative greedy methods including orthogonal matching pursuit (OMP), stagewise OMP(StOMP), and regularized OMP (ROMP). The CoSaMP combined with SP3 shows an improvement in reconstruction accuracy and exhibits distinct advantages over the comparative algorithms in multiple targets resolving. Stability analysis suggests that CoSaMP is robust to noise and performs stably with reduction of measurements. The feasibility and reconstruction accuracy of the proposed method are further validated by phantom experimental data.

Fast and Robust Reconstruction for Fluorescence Molecular Tomography via Regularization

Sparse reconstruction inspired by compressed sensing has attracted considerable attention in fluorescence molecular tomography (FMT). However, the columns of system matrix used for FMT reconstruction tend to be highly coherent, which means L 1 minimization may not produce the sparsest solution. In this paper, we propose a novel reconstruction method by minimization of the difference of L 1 and L 2 norms. To solve the nonconvex L 1-2 minimization problem, an iterative method based on the difference of convex algorithm (DCA) is presented. In each DCA iteration, the update of solution involves an L 1 minimization subproblem, which is solved by the alternating direction method of multipliers with an adaptive penalty. We investigated the performance of the proposed method with both simulated data and in vivo experimental data. The results demonstrate that the DCA for L 1-2 minimization outperforms the representative algorithms for L 1, L 2, L 1/2, and L 0 when the system matrix is highly coherent.

Performance evaluation of the simplified spherical harmonics approximation for cone-beam X-ray luminescence computed tomography imaging

As an emerging molecular imaging modality, cone-beam X-ray luminescence computed tomography (CB-XLCT) uses X-ray-excitable probes to produce near-infrared (NIR) luminescence and then reconstructs three-dimensional (3D) distribution of the probes from surface measurements. A proper photon-transportation model is critical to accuracy of XLCT. Here, we presented a systematic comparison between the common-used Monte Carlo model and simplified spherical harmonics (SPN). The performance of the two methods was evaluated over several main spectrums using a known XLCT material. We designed both a global measurement based on the cosine similarity and a locally-averaged relative error, to quantitatively assess these methods. The results show that the SP3 could reach a good balance between the modeling accuracy and computational efficiency for all of the tested emission spectrums. Besides, the SP1 (which is equivalent to the diffusion equation (DE)) can be a reasonable alternative model for emission wavelength over 6...

Reconstruction for Limited-Projection Fluorescence Molecular Tomography Based on a Double-Mesh Strategy

Limited-projection fluorescence molecular tomography (FMT) has short data acquisition time that allows fast resolving of the three-dimensional visualization of fluorophore within small animal in vivo. However, limited-projection FMT reconstruction suffers from severe ill-posedness because only limited projections are used for reconstruction. To alleviate the ill-posedness, a feasible region extraction strategy based on a double mesh is presented for limited-projection FMT. First, an initial result is rapidly recovered using a coarse discretization mesh. Then, the reconstructed fluorophore area in the initial result is selected as a feasible region to guide the reconstruction using a fine discretization mesh. Simulation experiments on a digital mouse and small animal experiment in vivo are performed to validate the proposed strategy. It demonstrates that the presented strategy provides a good distribution of fluorophore with limited projections of fluorescence measurements. Hence, it is suitable for reconstruction of limited-projection FMT.

A monocentric centerline extraction method for ring-like blood vessels

Centerline is generally used to measure topological and morphological parameters of blood vessels, which is pivotal for the quantitative analysis of vascular diseases. However, previous centerline extraction methods have two drawbacks on complex blood vessels, represented as the failure on ring-like structures and the existing of multi-voxel width. In this paper, we propose a monocentric centerline extraction method for ring-like blood vessels, which consists of three components. First, multiple centerlines are generated from the seed points that are chosen by randomly sprinkling points on blood vessel data. Second, multi-centerline fusion is used to repair the notches of centerlines on ring-like vessels, and the local maximum of distance from oundary is employed to remedy the missing centerline points. Finally, monocentric processing is devised to keep the vascular centerline with single voxel width. We compared the proposed method with Wan et al.’s method and topological thinning on five groups of data including synthesized vascular datasets and MR brain images. The result showed the proposed method performed better than the two contrast methods both by visual inspection and by quantitative assessment, which demonstrated the performance of the proposed method on ring-like blood vessels as well as the elimination of multi-voxel width points.

Laplacian manifold regularization method for fluorescence molecular tomography

Abstract. Sparse regularization methods have been widely used in fluorescence molecular tomography (FMT) for stable three-dimensional reconstruction. Generally, ℓ1-regularization-based methods allow for utilizing the sparsity nature of the target distribution. However, in addition to sparsity, the spatial structure information should be exploited as well. A joint ℓ1 and Laplacian manifold regularization model is proposed to improve the reconstruction performance, and two algorithms (with and without Barzilai–Borwein strategy) are presented to solve the regularization model. Numerical studies and in vivo experiment demonstrate that the proposed Gradient projection-resolved Laplacian manifold regularization method for the joint model performed better than the comparative algorithm for ℓ1 minimization method in both spatial aggregation and location accuracy.

An automatic multi-class coronary atherosclerosis plaque detection and classification framework

AbstractDetection of different classes of atherosclerotic plaques is important for early intervention of coronary artery diseases. However, previous methods focused either on the detection of a specific class of coronary plaques or on the distinction between plaques and normal arteries, neglecting the classification of different classes of plaques. Therefore, we proposed an automatic multi-class coronary atherosclerosis plaque detection and classification framework. Firstly, we retrieved the transverse cross sections along centerlines from the computed tomography angiography. Secondly, we extracted the region of interests based on coarse segmentation. Thirdly, we extracted a random radius symmetry (RRS) feature vector, which incorporates multiple descriptions into a random strategy and greatly augments the training data. Finally, we fed the RRS feature vector into the multi-class coronary plaque classifier. In experiments, we compared our proposed framework with other methods on the cross sections of Rotterdam Coronary Datasets, including 729 non-calcified plaques, 511 calcified plaques, and 546 mixed plaques. Our RRS with support vector machine outperforms the intensity feature vector and the random forest classifier, with the average precision of 92.6u2009±u20091.9% and average recall of 94.3u2009±u20092.1%. The proposed framework provides a computer-aided diagnostic method for multi-class plaque detection and classification.n Graphical abstractDiagram of the proposed automatic multi-class coronary atherosclerosis plaque detection and classification framework.ᅟ

Sparse non-convex Lp regularization for cone-beam X-ray luminescence computed tomography

ABSTRACT Cone-beam X-ray luminescence computed tomography (CB-XLCT) is an attractive hybrid imaging modality, and it has the potential of monitoring the metabolic processes of nanophosphors-based drugs in vivo. However, the XLCT imaging suffers from a severe ill-posed problem. In this work, a sparse nonconvex Lp (0u2009<u2009pu2009<u20091) regularization was utilized for the efficient reconstruction for early detection of small tumour in CB-XLCT imaging. Specifically, we transformed the non-convex optimization problem into an iteratively reweighted scheme based on the L1 regularization. Further, an iteratively reweighted split augmented lagrangian shrinkage algorithm (IRW_SALSA-Lp) was proposed to efficiently solve the non-convex Lp (0u2009<u2009pu2009<u20091) model. We studied eight different non-convex p-values (1/16, 1/8, 1/4, 3/8, 1/2, 5/8, 3/4, 7/8) in both 3D digital mouse experiments and in vivo experiments. The results demonstrate that the proposed non-convex methods outperform L2 and L1 regularization in accurately recovering sparse targets in CB-XLCT. And among all the non-convex p-values, our Lp(1/4u2009<u2009pu2009<u20091/2) methods give the best performance.