Full-field fan-beam x-ray fluorescence computed tomography with a conventional x-ray tube and photon-counting detectors for fast nanoparticle bioimaging

Li, Liang; Zhang, Siyuan; Li, Ruizhe; Chen, Zhiqiang

doi:10.1117/1.oe.56.4.043106

Cited by 35 publications

(22 citation statements)

References 30 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It uses a conventional Hamamatsu L12161-07 x-ray tube and a linear photon-counting detector array (eV3500, eV PRODUCTS, Saxonburg, Pennsylvania). 46,48 The detector array has 256 elements and 5 energy-bins of each one. The energy-bins were set at [26,33], [33,40], [40,50], [50,60], and [60, 80] keV.…”

Section: Physical Experiments Resultsmentioning

confidence: 99%

“…It works well even the linear attenuation coefficients of the basis materials are close. Due to the robustness of CMD, it is also expected to be used in other incomplete-data or sparse-data CT cases in the future, e.g., limited-angle CT, few-view CT, and low-dose CT. [46][47][48][49][50][51][52] Because the forward-propagation process in the CNN can be configured on GPU, the CMD algorithm can complete the material decompositions in a few seconds once it has been well trained. However, CMD also has some disadvantages.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Chen

2019

Opt. Eng.

Self Cite

View full text Add to dashboard Cite

Spectral computed tomography (CT) can reconstruct scanned objects at different energy-bins and thus solve the multimaterial decomposition (MMD) problem. Because the linear attenuation coefficients of different basis materials may be extremely close, the decomposition problem is often ill-conditioned. Meanwhile, traditional material decompositions with image-domain algorithms are usually voxelwise based. Therefore, these algorithms rely heavily on image quality. Ring artifacts often exist in the reconstructed images of spectral CT due to the inconsistency feature of energy-resolved detectors and beam-hardening effect. Considering the enlargement of the receptive field and taking advantage of the modeling ability of convolutional neural networks in deep learning, we proposed a convolutional material decomposition algorithm to solve the MMD problem through a basis of patches instead of pixels of the spectral CT images. Simulations and physical experiments were performed to validate the proposed algorithm, and its quality was compared with a traditional MMD algorithm in the image domain. Results show that the proposed method achieves good accuracy, reduces mean squared errors by one to two orders, and exhibits robustness in the MMD of spectral CT images even in the case that obvious ring artifacts is presented.

show abstract

Section: Physical Experiments Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Chen

2019

Opt. Eng.

Self Cite

View full text Add to dashboard Cite

show abstract

“…During the last decade the method has been extended to laboratory sources combined with NPs with the goal to provide molecular or functional bioimaging with new contrast and improved resolution. Data acquisition can be based either on a scanning pencil beam geometry [1,[3][4][5][6], cone-beam geometry with a collimated detector [7][8][9] or parallel approaches with single or multiple pinholes on the detector side [10,11]. The pencil-beam approach typically yields higher spatial resolution while the parallel methods have potential for faster data acquisition.…”

Section: Introductionmentioning

confidence: 99%

Quantitative predictions in small-animal X-ray fluorescence tomography

Shaker¹,

Larsson²,

Hertz³

2019

Biomed. Opt. Express

View full text Add to dashboard Cite

X-ray fluorescence (XRF) tomography from nanoparticles (NPs) shows promise for high-spatial-resolution molecular imaging in small-animals. Quantitative reconstruction algorithms aim to reconstruct the true distribution of NPs inside the small-animal, but so far there has been no feasible way to predict signal levels or evaluate the accuracy of reconstructions in realistic scenarios. Here we present a GPU-based computational model for small-animal XRF tomography. The unique combination of a highly accelerated Monte Carlo tool combined with an accurate small-animal phantom allows unprecedented realistic fullbody simulations. We use this model to simulate our experimental system to evaluate the quantitative performance and accuracy of our reconstruction algorithms on large-scale organs as well as mm-sized tumors. Furthermore, we predict the detection limits for sub-mm tumors at realistic NP concentrations. The computational model will be a valuable tool for optimizing next-generation experimental arrangements and reconstruction algorithms.

show abstract

“…The main issue of current XFCT imaging is the low emission and detection efficiency of the XRF photons, resulting in long data collection time and high statistic noise. This issue could be resolved by using high‐flux x‐ray sources, high‐sensitivity photon‐counting detector, and suitable probes . Due to high‐cost and bulky size of such sort of x‐ray sources and photon detectors, it would be more practical to develop an imaging probe that could generate more XRF photons under the same x‐ray excitation condition.…”

Section: Introductionmentioning

confidence: 99%

“…This issue could be resolved by using high-flux x-ray sources, high-sensitivity photon-counting detector, and suitable probes. [10][11][12] Due to high-cost and bulky size of such sort of x-ray sources and photon detectors, it would be more practical to develop an imaging probe that could generate more XRF photons under the same x-ray excitation condition.…”

Section: Introductionmentioning

confidence: 99%

The feasibility of NaGdF₄ nanoparticles as an x‐ray fluorescence computed tomography imaging probe for the liver and lungs

Zhang

Gao

et al. 2019

Medical Physics

Self Cite

View full text Add to dashboard Cite

Purpose As a novel imaging modality, x‐ray fluorescence computed tomography (XFCT) can provide distribution and concentration information of contrast agents containing high atomic number elements, such as iodine, gadolinium, barium, gold, and platinum. Since XFCT has a better sensitivity and detection limit of high‐Z elements compared with traditional and spectral CT, it becomes a powerful quantitative imaging tool for biological studies. The main problem of current XFCT imaging is its low emission and detection efficiency of x‐ray fluorescence (XRF) photons. Increasing XRF photons generation by choosing a high atomic element as a contrast agent is essential to improve the imaging quality of XFCT. Gadolinium emits at least a few times more of XRF photons than gold under the same x‐ray excitation condition, leading to a detection limit at a level of sub‐mg/mL as the next generation of clinical imaging modality. However, most current XFCT studies have utilized gadolinium salt as the contrast agent, which could not accumulate in organs or tumors efficiently, making in vivo XFCT imaging quite difficult. In this study, we present NaGdF4 nanoparticles with ultra‐small size as nanoprobes to test the feasibility for in vivo XFCT application for the first time. Methods NaGdF4 nanoparticles with different sizes (3–10 nm) were successfully synthesized via a coprecipitation process by controlling the reaction time at temperature of 290 °C. The morphology, crystal phase, chemical composition, and size of such NPs were further characterized with HR‐TEM, XRD, and EDX. The abilities of XRF photons from different sizes of NPs were quantified by our customized XFCT imaging system. To access the in vivo application of as‐synthesized NPs, such hydrophobic NPs capped OA molecules were further modified with AEP via a ligand‐exchange process and characterized with FT‐IR. For in vivo XFCT imaging, 0.1 mL of 30 mg/mL NPs were injected into nude mice via the tail vein. The Varian G‐297 x‐ray tube was set to 150 kV and 0.5 mA. The XRF photons were captured by a Kromek eV‐3500 photon counting detector at each 8° for 10 s. Result The successfully synthesized NaGdF4 nanoparticles (3–10 nm) were monodisperse, highly uniform spherical morphology and hexagonal crystalline phases. No significant influence on XRF photons yields or XFCT imaging quality were found by varying the nanoparticle size. The XRF photons were 2.5 times more emitted from NaGdF4 nanoparticles (NPs) compared to gold nanoparticles, thereby leading to a better image quality. With the AEP surface modification, such NPs were readily adapted for use in in vivo XFCT applications with monodispersity in aqueous solution and negligible cytotoxicity. With the tail‐vein injection, the liver, spleen, and lungs could be clearly imaged with XFCT at a sub‐mg/mL level. Conclusions Such NaGdF4 NPs, which were synthesized with coprecipitation process, were modified with AEP for in vivo XFCT applications. With both the phantom and in vivo experiments, such NPs were proved to be appropriate probe...

show abstract

Full-field fan-beam x-ray fluorescence computed tomography with a conventional x-ray tube and photon-counting detectors for fast nanoparticle bioimaging

Cited by 35 publications

References 30 publications

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Quantitative predictions in small-animal X-ray fluorescence tomography

The feasibility of NaGdF₄ nanoparticles as an x‐ray fluorescence computed tomography imaging probe for the liver and lungs

Contact Info

Product

Resources

About

Full-field fan-beam x-ray fluorescence computed tomography with a conventional x-ray tube and photon-counting detectors for fast nanoparticle bioimaging

Cited by 35 publications

References 30 publications

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Robust multimaterial decomposition of spectral CT using convolutional neural networks

Quantitative predictions in small-animal X-ray fluorescence tomography

The feasibility of NaGdF4 nanoparticles as an x‐ray fluorescence computed tomography imaging probe for the liver and lungs

Contact Info

Product

Resources

About

The feasibility of NaGdF₄ nanoparticles as an x‐ray fluorescence computed tomography imaging probe for the liver and lungs