Search

In-line X-ray phase contrast imaging offers considerable additional information beyond that acquired from conventional absorption contrast X-ray imaging, showing promising potentials in clinical diagnosis, materials characterization and so on. Given the physically intractable factors tangled inside, conventional phase retrieval methods typically suffer from limited feasibility. A deep-learning-augmented reconstruction strategy is proposed to improve the phase retrieval in spatial resolution and noise compression.The deep network is composed of a phase contrast refinement module and a phase retrieval module to stabilize and generalize the phase retrieval. The two modules are aggregated in a plug-and-play fashion with the final assembly finetuned using limited training data, essentially encouraging a semi-supervised training. Verification experiments were performed on simulated phase contrast images of histopathological images. The results were compared to those from conventional phase-attenuation duality method.The deep-learning-augmented reconstruction strategy increases structural similarity and peak signal-to-noise ratio of phase retrieval result by more than 8% and 30%, and reduces root mean squared error by 46% compared with conventional phase-attenuation duality method.The pilot study of deep learning deployment in in-line X-ray phase-contrast imaging exhibit advantages against conventional methods in terms of spatial resolution and noise robustness.

Diffuse fluorescence tomography (DFT) is a promising noninvasive imaging modality. To date, traditional physics-based approaches cannot substantially improve the imaging quality due to inherent light scattering property and severe ill-poseness of inverse problem. Therefore, three neural networks, namely, back-propagation neural network (BPNN), convolutional neural network (CNN) and U-net trained in an end-to-end fashion are explored to improve imaging quality. To evaluate their performance, a series of simulation, phantom and in vivo experiments are conducted for single and double-target reconstructions, and the results are compared with the traditional algebraic reconstruction technique (ART). The numerical results demonstrate that the neural network-based algorithms significantly enhance the image quality in terms of target shape, size, location, as well as noise robustness. The subsequent phantom and in vivo experimental results verify the effectiveness of the well-trained networks to estimate experimental data. Comparatively, BPNN is competitive in computational efficiency and generalization capacity.

Background Whether the characteristics and outcome of secondary acute promyelocytic leukemia (s-APL) are similar to de no APL (dn-APL) remains unknown. Patients and Methods Using the SEER database, we identified 3877 patients with APL diagnosed from 2000 to 2014, including 465 s-APL and 3412 dn-APL. Results Compared with dn-APL, s-APL werecharacterized by older median age, and a higher early mortality rate. Multivariate Cox model showed s-APL, older age, earlier year of diagnosis, and male gender were independently associated with worse survival. Notably, s-APL had a significantly inferior survival regardless of gender, race, marital status, and year of diagnosis. However, the difference between the 2 cohorts was only evident in younger patients (≤ 65 years) but was lost in older patients (> 65 years). Additionally, the majority of index cancer type was breast and prostate in female and male s-APL, respectively. Latency Conclusions Inferior survival of s-APL points to the need for treatment improvement.

Acute lung injury (ALI) is a common complication of severe acute pancreatitis (SAP) with a high mortality. Early prediction of patients at risk in initial stage can improve the long-term survival.A total of 91 patients with SAP out of 1647 acute pancreatitis patients from January 2015 to December 2020 were considered. A predictive model for SAP-associated ALI was constructed based on the valuable risk factors identified from routine clinical characteristics and plasma biomarkers. The value of the model was evaluated and compared with Lung Injury Prediction Score (LIPS). A nomogram was built to visualize the model.Diabetes, oxygen supplementation, neutrophil count and D-dimer were found to be associated with ALI in SAP. The predictive model based on these factors had an area under the receiver operating characteristic curve [AUC: 0.88, 95% confidence interval (CI): 0.81-0.95], which was superior to LIPS (AUC: 0.71, 95% CI: 0.60-0.83), also with the higher sensitivity (65%) and specificity (96%) than LIPS (62%, 74%, respectively). Decision curve analysis of the model showed a higher net benefit than LIPS. Visualization by a nomogram facilitated the application of the model.Diabetes, oxygen supplementation, neutrophil count and D-dimer were risk factors for SAP-associated ALI. The combination of these routine clinical data and the model visualization by a nomogram provided a simple and effective way in predicting ALI in the early phase of SAP.

The microstructural and crystallographic features of the β phase in TNM alloy during the hot extrusion at 1290 °C and the subsequent cooling process were investigated. Except the original grain boundary β phase, two categories of β particles, along grain boundaries of primarily recrystallized α grains (intergranular β) and in grain interiors of abnormally grown α grains (intragranular β), were found in the extruded microstructure. The former obeyed the Burgers orientation relationship with, at least, one of the neighboring α grains, whereas the latter did not show any specific orientation relationship with the α host. Analysis revealed that the thermally-induced α→β phase transformation interweaving with the abnormal α grain growth happened during the transition period from end-extrusion to water-quenching. The intergranular β particles were derived from the thermally-induced α→β phase transformation, and the intragranular β particles were resulted from the intergranular β precipitates after being swallowed by the abnormally grown α grains. The present work provides new information of thermally-induced phase transformation by adiabatic heating and the annexation of the second phase by abnormal grain growth during high-speed hot deformation.

Purpose We developed a novel dose verification method using a camera-based radioluminescence imaging system (CRIS) combined with a deep learning-based signal processing technique. Methods The CRIS consists of a cylindrical chamber coated with scintillator material on the inner surface of the cylinder, coupled with a hemispherical mirror and a digital camera at the two ends. After training, the deep learning model is used for image-to-dose conversion to provide absolute dose prediction at multiple depths of a specific water phantom from a single CRIS image under the assumption of a good consistency between the TPS setting and actual beam energy. The model was trained using a set of captured radioluminescence images and the corresponding dose maps from the clinical treatment planning system (TPS) for the sake of acceptable data collection. To overcome the latent error and inconsistency that exists between the TPS calculation and the corresponding measurement, the model was trained in an unsupervised manner. Validation experiments were performed on five square fields (ranging from 2 × 2 cm2 to 10 × 10 cm2 ), and three clinical IMRT cases. The results were compared to the TPS calculations in terms of gamma index at 1.5 cm, 5 cm and 10 cm depths. Results The mean 2% / 2mm gamma pass rates were 100% for square fields and 97.2% (range from 95.5% to 99.5%) for the IMRT fields. Further validations were performed by comparing the CRIS results with measurements on various regular fields. The results show a mean gamma pass rate of 91% (1% / 1mm) for cross-profiles and a mean percentage deviation of 1.15% for percentage depth doses (PDDs). Conclusions The system is capable of converting the irradiated radioluminescence image to corresponding water-based dose maps at multiple depths with a spatial resolution comparable to the TPS calculations. This article is protected by copyright. All rights reserved.

Pharmacokinetic parameter estimation with the support of dynamic diffuse fluorescence tomography (DFT) can provide helpful diagnostic information for tumor differentiation and monitoring. Adaptive extended Kalman filtering (AEKF) as a nonlinear filter method has the merits of high quantitativeness, noise robustness, and initialization independence. In this paper, indirect and direct AEKF schemes combining with a commonly used two-compartment model were studied to estimate the pharmacokinetic parameters based on our self-designed dynamic DFT system. To comprehensively compare the performances of both schemes, the selection of optimal noise covariance matrices affecting estimation results was first studied, then a series of numerical simulations with the metabolic time ranged from 4.16 min to 38 min was carried out and quantitatively evaluated. The comparison results show that the direct AEKF outperforms the indirect EKF in estimation accuracy at different metabolic velocity and demonstrates stronger stability at the large metabolic velocity. Furtherly, the in vivo experiment was conducted to achieve the indocyanine green pharmacokinetic-rate images in the mouse liver. The experimental results confirmed the capability of both schemes to estimate the pharmacokinetic-rate images and were in agreement with the theory predictions and the numerical simulation results.

Time-domain diffuse optical tomography can efficiently reconstruct both absorption and reduced scattering coefficients but is heavily limited by the ill-posedness in its inverse problem and low spatial resolution. To deal with these adversities, the truncated singular value decomposition (TSVD)-based whole-weighting-matrix inversion scheme can be a particularly suitable implementation. Unfortunately, TSVD is subject to a storage challenge for three-dimensional imaging of a bulk region, such as breast. In this paper, a multi-scale mesh strategy based on computed tomography (CT) anatomical geometry is adopted to solve the storage challenge, where a fine mesh is used in forward calculation to ensure accuracy, and a coarse mesh in the inversion process to enable TSVD-based inversion of the whole-weighting matrix. We validate the proposed strategy using simulated data for a single lesion model from clinical positron emission tomography images of a breast cancer patient, and further, for a complex model that is constructed by setting dual lesions at different separations in the CT breast geometry.

Our study investigated the role of oxygen in mediating the FLASH effect. This effect, which was first reported in vitro in the 1950s and in vivo in the 1970s, recently gained prominence with a number of publications showing differential sparing between normal tissues and tumors. Oxygen depletion (and subsequent induction of transient hypoxia) is the oldest and most prominent hypothesis to explain this effect. To better understand how the oxygen depletion hypothesis and oxygen enhancement ratio (OER) are relevant for interpreting FLASH benefits, an analytical model was proposed to estimate the sparing factor. The model incorporated factors such as OER, oxygen partial pressure (pO2), loco-regional oxygen diffusion/metabolism, total dose and dose rate. The sparing factor, was used to quantify the sparing of normal tissue (initially physoxic). The radiosensitivity parameters of two cell types (V79 Chinese hamster cells and T1 human kidney cells) were selected. Furthermore, the transient behavior of OER during finite time intervals was modeled, for both without and with the presence of oxygen transport using a diffusion model. For tissues with an oxygen consumption rate of 20 mmHg/s and a distance of 60 μm away from blood vessels, the sparing factor demonstrates an increase from 1.03/1.06 (V79/T1) at 2.5 Gy/s up to 1.28/1.72 (V79/T1) at 100 Gy/s (total dose: 10 Gy). For normal tissues of initial pO2 between 1.5 and 8 mmHg, the benefit from pushing the dose rate above 100 Gy/s is found to be marginal. Preliminary animal experiments have been conducted for validation. Overall, our study predicts that the dose rate associated with maximum normal tissue protection is between 50 Gy/s and 100 Gy/s. Other than the postulation of the hypoxic stem cell niches in normal tissues, we believe that a framework based upon the oxygen depletion hypothesis and OER is not able to efficiently interpret differential responses between normal and tumor tissue under FLASH irradiation.

Single-pixel imaging (SPI) enables the use of advanced detector technologies to provide a potentially low-cost solution for sensing beyond the visible spectrum and has received increasing attentions recently. However, when it comes to sub-Nyquist sampling, the spectrum truncation and spectrum discretization effects significantly challenge the traditional SPI pipeline due to the lack of sufficient sparsity. In this work, a deep compressive sensing (CS) framework is built to conduct image reconstructions in classical SPIs, where a novel compression network is proposed to enable collaborative sparsity in discretized feature space while remaining excellent coherence with the sensing basis as per CS conditions. To alleviate the underlying limitations in an end-to-end supervised training, for example, the network typically needs to be re-trained as the basis patterns, sampling ratios and so on. change, the network is trained in an unsupervised fashion with no sensing physics involved. Validation experiments are performed both numerically and physically by comparing with traditional and cutting-edge SPI reconstruction methods. Particularly, fluorescence imaging is pioneered to preliminarily examine the in vivo biodistributions. Results show that the proposed method maintains comparable image fidelity to a sCMOS camera even at a sampling ratio down to 4%, while remaining the advantages inherent in SPI. The proposed technique maintains the unsupervised and self-contained properties that highly facilitate the downstream applications in the field of compressive imaging.

Cherenkov imaging is a unique verification tool that could provide both dosimetric and tissue functional information during radiation therapy. However, the number of interrogated Cherenkov photons in tissue is always limited and tangled with stray radiation photons, severely frustrating the measurement the signal-to-noise ratio (SNR). As such, here, a noise-robust photon-limited imaging technique is proposed by comprehensively exploiting the physical rationale of low-flux Cherenkov measurements together with the spatial correlations of the objects. Validation experiments confirmed that the Cherenkov signal could be promisingly recovered with high SNR by irradiating at as few as one x ray pulse from a linear accelerator (10 mGy dose), and the Cherenkov excited luminescence imaging depth can be extended by >100% on average, for most concentrations of phosphorescent probe. This approach demonstrates that improved applications in radiation oncology could be seen when signal amplitude, noise robustness, and temporal resolution are comprehensively considered in the image recovery process.

Diffuse optical tomography (DOT) is a non-invasive imaging modality that uses near-infrared light to probe the optical properties of tissue. In conventionally used deterministic methods for DOT inversion, the measurement errors were not taken into account, resulting in unsatisfactory noise robustness and, consequently, affecting the DOT image reconstruction quality. In order to overcome this defect, an extended Kalman filter (EKF)-based DOT reconstruction algorithm was introduced first, which improved the reconstruction results by incorporating a priori information and measurement errors to the model. Further, to mitigate the instability caused by the ill-condition of the observation matrix in the tomographic imaging problem, a new, to the best of our knowledge, estimation algorithm was derived by incorporating Tikhonov regularization to the EKF method. To verify the effectiveness of the EKF algorithm and Tikhonov regularization-based EKF algorithm for DOT imaging, a series of numerical simulations and phantom experiments were conducted, and the experimental results were quantitatively evaluated and compared with two conventionally used deterministic methods involving the algebraic reconstruction technique and Levenberg–Marquardt algorithm. The results show that the two EKF-based algorithms can accurately estimate the location and size of the target, and the imaging accuracy and noise robustness are obviously improved. Furthermore, the Tikhonov regularization-based EKF obtained optimal parameter estimations, especially under the circumstance of low absorption contrast (1.2) and high noise level (10%).

In-line X-ray phase contrast imaging (IL-PCI) is a promising technology for clinical diagnosis because of its great advantage in distinguishing low contrast tissues and simple structure to implement. In order to recover the phase projections from the phase contrast measurements, conventional phase retrieval methods were developed based on assumptions such as homogeneous material, weak attenuation, and thus suffered from limited generalizability, practicability and feasibility. Deep learning-based methods have been proposed for phase retrieval and great success has been achieved. While the practical physical model of phase contrast imaging hasn’t been fully considered including the non-ideal effects of finite size of the x-ray micro focal spot, finite pixel size of the detector and the system noise. In this paper, a convolutional network based on generative adversarial network is proposed to retrieve the phase projections with fully considering the non-ideal effects in IL-PCI. The network composed of a generating network from which the phase projections were retrieved and a discriminating network from which the difference between the output of generation network and the reference phase projection is processed and backpropagated to the input of the network. Phase contrast measurements of the microspheres phantom were simulated and retrieved by the conventional methods and the proposed network. Results show the superiority of the proposed network in spatial resolution and noise suppression compared with the conventional method.

PURPOSE: To develop a deep unsupervised learning method with control volume (CV) mapping from patient positioning daily CT (dCT) to planning computed tomography (pCT) for precise patient positioning. METHODS: We propose an unsupervised learning framework, which maps CVs from dCT to pCT to automatically generate the couch shifts, including translation and rotation dimensions. The network inputs are dCT, pCT and CV positions in the pCT. The output is the transformation parameter of the dCT used to setup the head and neck cancer (HNC) patients. The network is trained to maximize image similarity between the CV in the pCT and the CV in the dCT. A total of 554 CT scans from 158 HNC patients were used for the evaluation of the proposed model. At different points in time, each patient had many CT scans. Couch shifts are calculated for the testing by averaging the translation and rotation from the CVs. The ground-truth of the shifts come from bone landmarks determined by an experienced radiation oncologist. RESULTS: The system positioning errors of translation and rotation are less than 0.47 mm and 0.17 degree, respectively. The random positioning errors of translation and rotation are less than 1.13 mm and 0.29 degree, respectively. The proposed method enhanced the proportion of cases registered within a preset tolerance (2.0 mm / 1.0 degree) from 66.67 percent to 90.91 percent as compared to standard registrations. CONCLUSIONS: We proposed a deep unsupervised learning architecture for patient positioning with inclusion of CVs mapping, which weights the CVs regions differently to mitigate any potential adverse influence of image artifacts on the registration. Our experimental results show that the proposed method achieved efficient and effective HNC patient positioning.

Purpose The study aims at a novel dosimetry methodology to reconstruct a 3D dose distribution as imparted to a virtual cylindrical phantom using an electronic portal imaging device (EPID). Methods A deep learning-based signal processing strategy, referred to as 3DosiNet, is utilized to learn a mapping from an EPID image to planar dose distributions at given depths. The network was trained with the volumetric dose exported from the clinical treatment planning system (TPS). Given the latent inconsistency between measurements and corresponding TPS calculations, unsupervised learning is formulated in 3DosiNet to capture abstractive image features that are less sensitive to the potential variations. Results Validation experiments were performed using five regular fields and three clinical intensity-modulated radiation therapy (IMRT) cases. The measured dose profiles and percentage depth dose (PDD) curves were compared with those measured using standard tools in terms of the 1D gamma index. The mean gamma pass rates (2%/2 mm) over the regular fields are 100% and 97.3% for the dose profile and PDD measurements, respectively. The measured volumetric dose was compared to the corresponding TPS calculation in terms of the 3D gamma index. The mean 2%/2 mm gamma pass rates are 97.9% for square fields and 94.9% for the IMRT fields. Conclusions The system promises to be a practical 3D dosimetric tool for pre-treatment patient-specific quality assurance and further developed for in-treatment patient dose monitoring.

Shenxian pigs are the only local black pig of Hebei Province, and were listed in the Genetics of Livestock and Poultry Resources of China in 2016. This breed of pig is considered to be a valuable local pig germplasm genetic resource in China. When compared with other introduced pig breeds, the Shenxian pig breed is characterized with early sexual maturity, short oestrus intervals, large litter sizes, and good meat quality, which are all of good research significance. However, the Shenxian pig variety was previously declared extinct in 2004 due to the introduction of a large number of foreign pig breeds. In order to preserve and study the Shenxian pig breed, the Hebei Zhengnong Animal Husbandry Co., Ltd. was established in Hebei Province for the purpose of preserving the purified Shenxian pig strain. In the present study, in order to understand the genetic variations of Shenxian pigs, identify selected regions related to superior traits, and accelerate the breeding processes of Shenxian pigs, the whole genome of the Shenxian pigs was resequenced and compared with that of large white pigs. The goal was to explore the germplasm characteristics of Shenxian pigs.The results obtained in this research investigation revealed that the genetic relationships of the Shenxian pig breed were complex, and that sub-populations could be identified within the general population. A total of 23M SNP sites were obtained by whole genome resequencing, and 1,509 selected sites were obtained via bioinformatics analyses. It was determined after annotation that a total of 19 genes were enriched in three items of bioengineering, molecular function, and cell composition.During this research investigation, the aforementioned 19 genes were subjected to GO and KEGG analyses. Subsequently, the candidate genes related to cell proliferation were obtained (DMTF1 and WDR5), which were considered to possibly be related to the slow growth and development of Shenxian pigs. In addition, the candidate genes related to lactation were obtained (CSN2 and CSN3). However, no genes related to meat quality traits were successfully screened.

The United Nations' "Sustainable Development Goals" (SDGs) express attention to climate action. Even though a considerable number of papers have targeted this issue, the literature on the top five, "China, India, Japan, Russia, and the United States" economies is uncommon. Therefore, this paper is targeted to examine the influence of renewable energy (RE), environmental technologies (ETs), and economic policy uncertainty (EPU) on carbon emissions. By using the most recent data available from 1992 to 2020, results are estimated with robust econometric techniques, i. e. "cross-sectionally augmented autoregressive distributed lag (CS-ARDL) and augmented mean group (AMG)". Findings reflect the harmful role of EPU. However, RE and ETs have a supportive role in the transition towards a sustainable environment. The findings are also strong in terms of policy implications for the top five polluters.

Significance: Deep-tissue penetration by x-rays to induce optical responses of specific molecular reporters is a new way to sense and image features of tissue function in vivo. Advances in this field are emerging, as biocompatible probes are invented along with innovations in how to optimally utilize x-ray sources. Aim: A comprehensive review is provided of the many tools and techniques developed for x-ray-induced optical molecular sensing, covering topics ranging from foundations of x-ray fluorescence imaging and x-ray tomography to the adaptation of these methods for sensing and imaging in vivo. Approach: The ways in which x-rays can interact with molecules and lead to their optical luminescence are reviewed, including temporal methods based on gated acquisition and multipoint scanning for improved lateral or axial resolution. Results: While some known probes can generate light upon x-ray scintillation, there has been an emergent recognition that excitation of molecular probes by x-ray-induced Cherenkov light is also possible. Emission of Cherenkov radiation requires a threshold energy of x-rays in the high kV or MV range, but has the advantage of being able to excite a broad range of optical molecular probes. In comparison, most scintillating agents are more readily activated by lower keV x-ray energies but are composed of crystalline inorganic constituents, although some organic biocompatible agents have been designed as well. Methods to create high-resolution structured x-ray-optical images are now available, based upon unique scanning approaches and/or a priori knowledge of the scanned x-ray beam geometry. Further improvements in spatial resolution can be achieved by careful system design and algorithm optimization. Current applications of these hybrid x-ray-optical approaches include imaging of tissue oxygenation and pH as well as of certain fluorescent proteins. Conclusions: Discovery of x-ray-excited reporters combined with optimized x-ray scan sequences can improve imaging resolution and sensitivity.

The automatic diagnosis of various retinal diseases from fundus images is important to support clinical decision-making. However, developing such automatic solutions is challenging due to the requirement of a large amount of human-annotated data. Recently, unsupervised/self-supervised feature learning techniques receive a lot of attention, as they do not need massive annotations. Most of the current self-supervised methods are analyzed with single imaging modality and there is no method currently utilize multi-modal images for better results. Considering that the diagnostics of various vitreoretinal diseases can greatly benefit from another imaging modality, e.g., FFA, this paper presents a novel self-supervised feature learning method by effectively exploiting multi-modal data for retinal disease diagnosis. To achieve this, we first synthesize the corresponding FFA modality and then formulate a patient feature-based softmax embedding objective. Our objective learns both modality-invariant features and patient-similarity features. Through this mechanism, the neural network captures the semantically shared information across different modalities and the apparent visual similarity between patients. We evaluate our method on two public benchmark datasets for retinal disease diagnosis. The experimental results demonstrate that our method clearly outperforms other self-supervised feature learning methods and is comparable to the supervised baseline., Comment: IEEE Transactions on Medical Imaging, code is at https://github.com/xmengli999/self_supervised

Cone-beam x-ray luminescence computed tomography (CB-XLCT) has been proposed as a promising imaging tool, which enables three-dimensional imaging of the distribution of nanophosphors (NPs) in small animals. However, the reconstruction performance is usually unsatisfactory in terms of spatial resolution due to the ill-posedness of the CB-XLCT inverse problem. To alleviate this problem and to achieve high spatial resolution, a reconstruction method consisting of inner and outer iterations based on a restarted strategy is proposed. In this method, the primal-dual Newton conjugate gradient method (pdNCG) is adopted in the inner iterations to get fast reconstruction, which is used for resetting the permission region and increasing the convergence speed of the outer iteration. To assess the performance of the method, both numerical simulation and physical phantom experiments were conducted with a CB-XLCT system. The results demonstrate that compared with conventional reconstruction methods, the proposed re-pdNCG method can accurately and efficiently resolve the adjacent NPs with the least relative error.

Program slicing, as a technique of program decomposition, is widely used in the field of program testing, model checking, software verification, symbolic execution, and other fields. However, the traditional approaches of program slicing tend to produce too large slices and the static program analyses are hard to be precise enough. Scenario-oriented program slicing, which considers the actual usage of software, gives a more precise perspective of program slicing. In this paper, we propose an approach of scenario-oriented program slicing, which combines constraint logic programming and program transformation. According to the observation that the output of a program transformation is a semantically equivalent program where the properties of interest are preserved, we can apply a sequence of transformations, more powerful than those needed for program specialization, refining the slicing to the desired degree of precision. And constraint logic programming has been shown to be a powerful, flexible formalism to reason about the correctness of programs. The novel contributions of this paper are as follows: 1) converting the problem of program slicing into program transformation and retrieval; 2) presenting a set of constraint handling rules for scenario-oriented program slicing in constraint logical programming programs; and 3) deriving a scenario-oriented program slicing algorithm. The method of scenario-oriented program slicing has been implemented and we have demonstrated its effectiveness and efficiency on three open source software projects in GitHub.