Search

SummaryObjectiveElectronic health record (EHR) systems contain a wealth of clinical data stored as both codified data and free-text narrative notes, covering hundreds of thousands of clinical concepts available for research and clinical care. The complex, massive, heterogeneous, and noisy nature of EHR data imposes significant challenges for feature representation, information extraction, and uncertainty quantification. To address these challenges, we proposed an efficientAggregated naRrativeCodifiedHealth (ARCH) records analysis to generate a large-scale knowledge graph (KG) for a comprehensive set of EHR codified and narrative features.MethodsThe ARCH algorithm first derives embedding vectors from a co-occurrence matrix of all EHR concepts and then generates cosine similarities along with associatedp-values to measure the strength of relatedness between clinical features with statistical certainty quantification. In the final step, ARCH performs a sparse embedding regression to remove indirect linkage between entity pairs. We validated the clinical utility of the ARCH knowledge graph, generated from 12.5 million patients in the Veterans Affairs (VA) healthcare system, through downstream tasks including detecting known relationships between entity pairs, predicting drug side effects, disease phenotyping, as well as sub-typing Alzheimer’s disease patients.ResultsARCH produces high-quality clinical embeddings and KG for over 60, 000 EHR concepts, as visualized in the R-shiny powered web-API (https://celehs.hms.harvard.edu/ARCH/). The ARCH embeddings attained an average area under the ROC curve (AUC) of 0.926 and 0.861 for detecting pairs of similar EHR concepts when the concepts are mapped to codified data and to NLP data; and 0.810 (codified) and 0.843 (NLP) for detecting related pairs. Based on thep-values computed by ARCH, the sensitivity of detecting similar and related entity pairs are 0.906 and 0.888 under false discovery rate (FDR) control of 5%. For detecting drug side effects, the cosine similarity based on the ARCH semantic representations achieved an AUC of 0.723 while the AUC improved to 0.826 after few-shot training via minimizing the loss function on the training data set. Incorporating NLP data substantially improved the ability to detect side effects in the EHR. For example, based on unsupervised ARCH embeddings, the power of detecting drug-side effects pairs when using codified data only was 0.15, much lower than the power of 0.51 when using both codified and NLP concepts. Compared to existing large-scale representation learning methods including PubmedBERT, BioBERT and SAPBERT, ARCH attains the most robust performance and substantially higher accuracy in detecting these relationships. Incorporating ARCH selected features in weakly supervised phenotyping algorithms can improve the robustness of algorithm performance, especially for diseases that benefit from NLP features as supporting evidence. For example, the phenotyping algorithm for depression attained an AUC of 0.927 when using ARCH selected features but only 0.857 when using codified features selected via the KESER network[1]. In addition, embeddings and knowledge graphs generated from the ARCH network were able to cluster AD patients into two subgroups, where the fast progression subgroup had a much higher mortality rate.ConclusionsThe proposed ARCH algorithm generates large-scale high-quality semantic representations and knowledge graph for both codified and NLP EHR features, useful for a wide range of predictive modeling tasks.

Fluorescent probes based on molecular rotor are expected to be a powerful tool to uncover the relationship between intracellular viscosity and organelle functions as well as related diseases. Herein, a fluorophore MV-indol containing multiple rotors was constructed by linking triphenylamine group and benzindole unit. It was found that MV-indol can quantitatively detect viscosity with a high selectivity and a 75-fold enhancement in quantum yield (QY). Moreover, the viscosity-sensitive property could be ascribed to the inhibition of molecular rotation and restriction of twisted intramolecular charge transfer (TICT) in a high-viscous environment, which was confirmed by time-dependent density functional theory (TD-DFT). Importantly, we also demonstrated that MV-indol can precisely localize in mitochondria and monitor the change of mitochondrial viscosity at the cellular level.

Assortment optimization has received active explorations in the past few decades due to its practical importance. Despite the extensive literature dealing with optimization algorithms and latent score estimation, uncertainty quantification for the optimal assortment still needs to be explored and is of great practical significance. Instead of estimating and recovering the complete optimal offer set, decision-makers may only be interested in testing whether a given property holds true for the optimal assortment, such as whether they should include several products of interest in the optimal set, or how many categories of products the optimal set should include. This paper proposes a novel inferential framework for testing such properties. We consider the widely adopted multinomial logit (MNL) model, where we assume that each customer will purchase an item within the offered products with a probability proportional to the underlying preference score associated with the product. We reduce inferring a general optimal assortment property to quantifying the uncertainty associated with the sign change point detection of the marginal revenue gaps. We show the asymptotic normality of the marginal revenue gap estimator, and construct a maximum statistic via the gap estimators to detect the sign change point. By approximating the distribution of the maximum statistic with multiplier bootstrap techniques, we propose a valid testing procedure. We also conduct numerical experiments to assess the performance of our method.

Metal-dielectric heterostructures have shown great application potentials in physics, chemistry and material science. In this work, we have designed and manufactured ordered metal-dielectric multiple heterostructures with tunable optical properties, which can be as large as the order of square centimeters in size. We experimentally realized that the surface-enhanced Raman scattering signal of the periodic multiple heterostructures increased 50 times compared with the silicon nanodisk-gold film arrays, which is attributed to the large-scale hotspots and high efficient coupling between the optical cavities and surface plasmon resonance modes. More importantly, the substrate also features a good uniformity and an excellent reproducible fabrication, which is very promising for practical applications.

Single-molecule imaging is a powerful method for unveiling precise molecular mechanisms. Particularly, single-molecule analysis with total internal reflection fluorescence (TIRF ) microscopy has been successfully applied to the characterization of molecular mechanisms in ncRNA studies. Tracing interactions at the single-molecule level have elucidated the intermediate states of the reaction, which are hidden by ensemble averaging in combinational biochemical approaches, and clarified the key steps of the interaction. However, applying a single-molecule technique to ncRNA analysis still remains a challenge, requiring laborious trial and error to identify a suitable glass surface passivation method. In this chapter, we revisit the major glass surface passivation methods using polyethylene glycol (PEG) treatment and summarize a detailed protocol for single-molecule analysis of the dicing process of Dcr-2, which may apply piRNA studies in the future.

With growing interest in resource recovery and/or reuse, waste materials have been considered a promising alternative for phosphorus (P) adsorption because they are low-cost and easily accessible. Crushed autoclaved aerated concrete (CAAC), as representative construction waste, has been extensively studied for P removal in ecological technologies such as treatment wetlands. However, most of the previous studies focused on the adsorption of orthophosphate, namely reactive phosphorus, and lacked attention to non-reactive phosphorus (NRP) which is widely present in sewage. This study presents the first investigation on the potential and mechanism of CAAC removing four model NRP compounds. Adsorption isotherm and kinetics of NRP onto CAAC indicate that the removal of NRP was a chemisorption process and also involved a two-step pore diffusion process. The desorption experiment shows that different NRP species showed varying degrees of desorption. Most NRP was irreversibly adsorbed on CAAC. Among the model compounds considered in this study, the adsorption capacity and hydrolysis rate of organophosphorus were much less than that of inorganic phosphorus. Moreover, the adsorption of different NRP species by CAAC in the mesocosm study was different from the results of laboratory adsorption experiments, and the possible biodegradation was essential for the conversion and removal of NRP. The findings confirmed the validity of CAAC for NRP removal and the potential advantages of CAAC in terms of costs and environmental impact. This study will contribute to a better understanding of NRP conversion and environmental fate and that can be the basis for a refined risk assessment.

Excess phosphorus (P) in surface runoff has significant deleterious impacts on water quality through eutrophication. Commonly, P is transported via non-point pollution and the proportion of easily plant-available reactive P (RP) among other P forms may vary significantly. Non-reactive P (NRP) can potentially contribute to the eutrophication of waterbodies, however the cleavage into bio-available P forms and eventually their biological uptake remains uncertain. This holds also true for floating treatment wetlands (FTWs) which became established as nutrient mitigation measures for surface waters in recent years. However, little information is available about the conversion and removal of NRP in FTWs. In this study, the conversion and removal of different forms of P in FTWs were investigated. Experiments were operated in batch mode and treatments consisted of (1) two concentration levels: a high P concentration of 3.0 mg/L and a low P concentration of 1.0 mg/L, and (2) four mesocosm treatments: (a) artificial roots only, (b) substrates only, (c) plants only, (d) plants and substrates. The results showed that RP removal mainly depended on sedimentation, substrate sorption, and biological assimilation. The removal of NRP mainly depended on hydrolysis, microbial-mediated conversion, and biological absorption. The combination of plant and substrate provided stable and efficient phosphorus removal performance in high P conditions, while plants were important for P removal in low P conditions. Living plants were indispensable and greatly affected the performance of FTWs. The specific enrichment and culling of microorganisms by plants resulted in the formation of specific rhizosphere microbial communities and promoted the removal of NRP. Pseudomonas, Enterobacter, Acidovorax might be responsible for P mineralization in the FTWs. Comprehensive analysis indicated that the conversion and removal pathways of P in the FTWs were not mutually independent, and the plant-microbe-substrate interactions cannot be underestimated.

Recovery of phosphorus (P) from wastewater can help establish a new P cycle. However, there are many P forms in wastewater, not always in reactive forms, which are the most suitable for direct recovery. The enhanced biological phosphorus removal process with sidestream phosphorus recovery (EBPR-SPR) is an effective way to remove and recover P resources in wastewater, but there is a lack of research on the transformation and fate of non-reactive phosphorus (NRP) in it. This study selected four model NRP to investigate their transformation and fate in an EBPR-SPR process. The transformation of NRP in pure water and activated sludge under anaerobic and aerobic conditions were compared. The effects of Ca/P ratio and pH on NRP recovery were studied, and the recovery products of NRP were characterized. It was found that NRP containing phosphoanhydride and phosphoester bonds were more easily hydrolyzed to reactive P (RP) than that containing PC bonds. NRP will be adsorbed and accumulated by activated sludge, and activated sludge will accelerate the conversion of NRP to RP. Tripolyphosphate can form complex precipitation with Ca

Metal nanoparticles (MNPs) possess the intriguing property of enhancing and limiting the energy of light field to subwavelength scale, which have attracted extensive attention in physics, chemistry, and life sciences. Meanwhile, squeezing is a general concept in quantum optics and can be used to engineer matter interaction scenarios. Here we consider an artificial hybrid molecule consisting of an MNP interacting with a semiconductor quantum dot (QD), combining with a squeezed reservoir. Using experimentally realistic parameters for the MNP-QD architecture and fully quantized approach, we theoretically explore the optical properties of the hybrid molecule driven by an external applied laser field via probing the resonance fluorescence in the steady state. We show that in this way it is possible to engineer and control the peak-value magnitudes, widths, and shapes of the resonance fluorescence spectra under the influence of the squeezed vacuum field without the need for the strong-coupling condition between the MNP and QD. In particular, some interesting phenomena in rich spectral responses are attainable, including Fano-type resonance fluorescence, fluorescence quenching, fluorescence narrowing, and fluorescence enhancement. Our method can also be extended to other nanoscopic structures, such as a plasmonic nanoantenna coupled to an emitter. These unique line shapes obtained here may have potential applications in developing quantum plasmonic platforms and sensitive on-chip devices such as optical switches and sensors.

Plasmonic structures such as metal nanoparticles (MNPs) provide a promising way to explore and control light in a solid-state environment. Here, we study a hybrid system composed of a localized surface plasmon mode of a MNP interacting with a semiconductor quantum dot (QD) subject to a Faraday magnetic field. The MNP-QD system is driven by an externally applied probe laser field and then its scattering photons are detected in the far field. To this end, employing a full quantum optical model, we theoretically calculate and demonstrate the scattering spectrum of the system via a combination of analytical and numerical techniques with realistic parameters. It is found that the applied magnetic field can be utilized to efficiently regulate and control the scattering line shapes. In particular, a class of optical transparency is induced in the scattering spectrum of the plasmon-excition (plexcitonic) system that we refer to as magnetically induced plexcitonic transparency. At the same time, a sharp double-Fano resonance profile is generated in the presence of the magnetic field. The predicted features of the spectra have potential applications in designing sensitive on-chip devices.

SARS-CoV-2 vaccines are powerful tools to combat the COVID-19 pandemic, but vaccine hesitancy threatens these vaccines’ effectiveness. To address COVID-19 vaccine hesitancy and ensure equitable distribution, understanding the extent of and factors associated with vaccine acceptance and uptake is critical. We report the results of a large nationwide study conducted December 2020-May 2021 of 34,470 users from COVID-19-focused smartphone-based app How We Feel on their willingness to receive a COVID-19 vaccine. Nineteen percent of respondents expressed vaccine hesitancy, the majority being undecided. Of those who were undecided or unlikely to get a COVID-19 vaccine, 86% reported they ultimately did receive a COVID-19 vaccine. We identified sociodemographic and behavioral factors that were associated with COVID-19 vaccine hesitancy and uptake, and we found several vulnerable groups at increased risk of COVID-19 burden, morbidity, and mortality were more likely to be vaccine hesitant and had lower rates of vaccination. Our findings highlight specific populations in which targeted efforts to develop education and outreach programs are needed to overcome vaccine hesitancy and improve equitable access, diversity, and inclusion in the national response to COVID-19.

Recovery of phosphorus from sewage can help establish a new phosphorus cycle and hydroxyapatite (HAP) crystallization is a promising way. HAP crystallization is an amorphous calcium phosphate (ACP) mediated process, and its induction time reflects the rate of HAP nucleation, and seriously affects the efficiency of phosphorus recovery. In this study, the effects of different types of dissolved organic matter (DOM) on the induction time and phosphorus recovery performance of ACP-mediated HAP phosphorus recovery were studied, and the mechanism was analyzed by X-Ray Diffraction, Fourier transform infrared spectroscopy, and scanning electron micrograph with energy dispersive spectrometry. The results show that DOM greatly prolongs the induction time of ACP-mediated HAP crystallization and leads to an increase in the yield of microcrystals, thus leading to a decrease in phosphorus recovery efficiency. DOM inhibits ACP-mediated HAP crystallization by complexing lattice ions and occupying active growth sites on the crystal surface. Pre-removal of DOM can not only improve the speed and efficiency of phosphorus recovery by the HAP crystallization process but also improve product quality.

The depletion of phosphorus resources and the excess discharge of phosphorus into waste streams are contrasting problems. The key to solving both problems is to recover phosphorus from the waste streams. Current phosphorus recovery technologies require high phosphorus concentrations and lack the ability to separate toxic substances from recovered phosphorus products. Membrane separation processes such as nanofiltration, forward osmosis, and electrodialysis are examples of effective methods for solving some of these issues. In this paper, the mechanisms, performance, and influential factors affect phosphorus recovery from membrane separation are reviewed. Membrane fouling, energy consumption, and the selectivity of toxic substances in membrane separation processes were evaluated. This work will serve as a basis for future research and development of phosphorus recovery by membrane separation processes and as a response to the increasingly pressing issues of eutrophication and the growing depletion of phosphorus resources.

In the two-atom or multiatom system, the atoms can interact with each other through exchange of virtual photon. This kind of energy exchange is often referred as the dipole-dipole interaction (DDI). Here we consider this DDI system consisting of a pair of two-level atoms strongly coupled with a bimodal whispering-gallery-mode (WGM) microresonator which is driven by an external laser field. Our aim is to explore the photon correlation characteristics of the proposed architecture using realistic experimental parameter values. We compare in detail the quality of photon antibunching (i.e., the smallness of the second-order correlation function) from three involved configurations in cavity quantum electrodynamics (QED): (i) only one two-level atom, (ii) two far apart two-level atoms without DDI, and (iii) two DDI (dipole-coupled) two-level atoms are respectively coupled to the driven WGM microresonator through the evanescent field. We clearly show that the DDI between both atoms can distinctly enhance the photon antibunching even in the weak-coupling regime in configuration (iii) with feature-rich line shapes. We also find that the photon antibunching can be modulated by properly adjusting the atom-cavity coupling strength. In addition, we display that this strong photon antibunching is robust against the cooperative atomic decay. Our DDI-based cavity QED scheme may provide an alternative way to the construction of integrated on-chip single-photon sources.

Quantum control of photons has become an area of great interest in the development of quantum technology. Here we explore how an external magnetic field can be exploited to control the statistical properties of photons in a bimodal cavity quantum electrodynamics (CQED) system. Our CQED system is composed of a bimodal micropillar Fabry-P\'erot cavity containing a single $V$-level quantum dot (QD) involving fine-structure-splitting neutral exciton transitions under the effect of the applied magnetic field in the presence of an incident pump laser field driving one cavity mode. With the introduction of the external magnetic field, not only is the conversion of polarization characteristics of photons achieved, but the statistics properties of photons are well engineered and the switching between photon bunching and antibunching is realized in the weak-coupling regime of CQED. The superiority of our system is manifested in the following aspects. (i) Using experimental parameters, we can manipulate the statistical properties of photons by appropriately tuning the strength of the magnetic field under the weak-coupling CQED condition, accompanied by either enhanced antibunching or a transition from antibunching to strong bunching of the light and vice versa. So we can arrive at selective photon statistics. (ii) The strong photon antibunching effect, under nonresonant scenarios but without the need to match the resonant or near-resonant conditions for the cavity and QD, can appear. Furthermore, we find that the multifrequency photon blockade can be generated in the present system, which is different from previous studies about photon blockade only for specific optical detuning. (iii) The high photon occupations or transmissions, simultaneously accompanied by strong photon antibunching in the system, are beneficial to the correlation measurement in practical experiments. This process works with a high quality of photon antibunching and bunching over a wide range of parameters and has potential applications in on-chip quantum information processing. It is hoped that the proposed scheme provides a possibility for the generation of tunable single-photon sources or controllable photon quantum gates.

The efficiency of image enhancement algorithms depends on the quality and processing speed of image enhancement. There are many algorithms to implement image enhancement using wavelet theory. These algorithms have one thing in common: they all capture image details by decomposing low frequency sub-images. In fact, a lot of details in high-frequency sub-images are also found. Enlightened by the above-mentioned facts, a novel medical image enhancement method based on wavelet decomposition is proposed by adding details from the high-frequency sub-images and decomposing the image specially with ant-symmetric biorthogonal wavelet instead of some traditional wavelets. It not only improves the image enhancement, but also overcomes the shortcomings of large computation with faster computational speed and satisfies the real-time requirement in edge detection. Simulation experiments of mammographic images are implemented by Matlab with several different methods, the results show that the proposed method is superior to some popular methods, such as histogram equalization and wavelet nonlinear enhancement.

The proposal for single-photon generation utilizing quantum interference of two coupled cavity modes with weak Kerr nonlinearity under resonant coherent driving has been successfully put forward in previous work [T. C. H. Liew and V. Savona, Phys. Rev. Lett. 104, 183601 (2010)]. Here we suggest an alternative scheme to explore the statistical properties of the photons and their control in a parametrically amplified photonic molecule that is composed of a pair of coupled cavities, both containing the degenerate optical parametric amplifier (DOPA), by means of extending the model of Liew and Savona. Due to the destructive quantum interference effect between different paths for two-photon excitation, the photons in both cavity modes can exhibit a strong antibunching effect when one of the two cavities is coherently driven by an external laser field. In order to intuitively understand the strong photon antibunching phenomenon occurring in our DOPA photonic molecule, the approximate analytical solutions of the system are provided via the Schr\"odinger equation and further are compared with the full numerical solutions via the master equation. The in-depth results reveal that the two solutions are in good agreement. For the two cavity modes, we achieve the optimal antibunching conditions in terms of the nonlinear gain and the strength ratio between the two pump laser fields based on analytical calculations, gaining deep insight into the rich physics of photon statistics. We also show that the statistical properties of the photons can be well controlled by regulating the relative phase between the two pump laser fields and the tunneling rate between the two cavity modes. Furthermore, we address the second-order cross-correlation function between the two cavity modes and find that a strong intermodal antibunching can be efficiently generated. Our approach via the DOPA photonic molecule architecture can be useful to generate tunable single-photon sources or controlled photonic quantum gates.

We present a detailed study of a three-mode-coupling cavity optomechanical system where one mechanical mode and two optical whispering-gallery modes are coupled together. We extend the earlier investigation of regular optomechanically induced transparency (OMIT) [Science330, 1520 (2010)SCIEAS0036-807510.1126/science.1195596] in a more general case. A complete analytical description of the present system is obtained, and quantitative analyses of the Fano-shape OMIT spectrum are provided. It is clearly shown that a sharp and asymmetric Fano-shape OMIT lineshape, which is decided by the phase difference between the probe field and the mechanical driving, can be observed from the probe output.

Dual frequency comb generation has been the focus of many recent research efforts owing to its viability in a myriad of applications, ranging from precision measurement to high-speed optical communications. Here, with the state of the art of a single quantum dot (QD) coupled to a pillar microcavity [Nat. Nanotechnol. 12, 663 (2017)], we put forward an alternative setup for the generation of orthogonally polarized dual frequency combs and their corresponding control by using coherent two-tone cavity driving, i.e., a control laser and a probe laser. The present QD-pillar microcavity system operates in the weak-coupling regime, and this makes the experimental realization of the scheme easy. The underlying nonlinear process of the system, thanks to the QD-pillar interaction, for example, degenerate and nondegenerate four-wave mixing, results in efficient frequency comb formation. The dual combs consist of equidistant spectral lines whose spectral spacing coincides with the frequency difference between the control and probe lasers. So, programming this frequency difference between them enables us to tune both the comb-line spacing (i.e., repetition rate) and comb-line number in both vertical ($V$) and horizontal ($H$) polarizations. As compared with the previous approaches for dual microcomb generation, the present scheme has attractive features such as a tunable, slow repetition rate (as slow as a few megahertz), low input power (as low as a few nanowatts), and integrated compact nature in experimentally realistic operating regimes. The achievable low repetition rates are desirable to resolve narrower spectral lines and to enable use of low-power signal processing electronics.

This study proposes an innovative M-L (Multiple-channel local binary fitting) model for medical image segmentation. Designed to improve upon existing image segmentation models, the M-L model introduces a regional limit function to the multi-band active contour model to enable multilayer image segmentation. The Gaussian kernel function is used to improve the previous model's robustness, necessitating the use of a new initialization curve which enhances the accuracy of segmentation results. Compared to existing image segmentation methods, the proposed M-L model improves numerical stability and efficiency through the introduction of a new penalty term and an increased step length. This simulation experiment verifies the advantages of the new M-L model for improved medical image segmentation, including increased efficiency and usability of the model.