Search

Background: The Boston Naming Test (BNT) is the most widely used measure to assess anomia. However, it has been criticized for failing to differentiate the underlying cognitive process of anomia. Objective: We validated the color-picture version of BNT (CP-BNT) in a sample with diverse neurodegenerative dementia diseases (NDDs). We also verified the differential ability of the composite indices of CP-BNT across NDDs groups. Methods: The present study included Alzheimer’s disease (n = 132), semantic variant primary progressive aphasia (svPPA, n = 53), non-svPPA (n = 33), posterior cortical atrophy (PCA, n = 35), and normal controls (n = 110). We evaluated psychometric properties of CP-BNT for the spontaneous naming (SN), the percentage of correct responses on semantic cuing and word recognition cuing (% SC, % WR). Receiver operating characteristic analysis was used to examine the discriminatory power of SN alone and the composite indices (SN, % SC, and % WR). Results: The CP-BNT had sufficient internal consistency, good convergent, divergent validity, and criterion validity. Different indices of CP-BNT demonstrated distinct cognitive underpinnings. Category fluency was the strongest predictor of SN (β= 0.46, p < 0.001). Auditory comprehension tests highly associated with % WR (Sentence comprehension: β= 0.22, p = 0.001; Word comprehension: β= 0.20, p = 0.001), whereas a lower visuospatial score predicted % SC (β= – 0.2, p = 0.001). Composite indices had better predictability than the SN alone when differentiating between NDDs, especially for PCA versus non-svPPA (area under the curve increased from 63.9% to 81.2%). Conclusion: The CP-BNT is a highly linguistically relevant test with sufficient reliability and validity. Composite indices could provide more differential information beyond SN and should be used in clinical practice.

BACKGROUND There are six different formulae for estimating mean arterial pressure (MAP) from systolic and diastolic pressure readings. This study is to determine the optimum formula for calculating MAP when compared to the gold standard approach, which is the area under the curve of an invasively measured pulse waveform divided by the cardiac cycle duration. METHODS Eight live pigs were used as the experimental model for the invasive measurement of femoral artery pressure (AP) by a fluid filled catheter connected with a pressure transducer. In addition, intraocular pressure (IOP) and jugular vein pressure (JVP) were also recorded. The mean pressure (MP) was calculated from digital waveforms sampled at 1,000 points per second with the six formulae and area method for AP, IOP and JVP. RESULTS The absolute mean difference between the area MAP and each formula’s MAP ranged from 0.98 to 3.23 mm Hg. Our study also found that even under physiological conditions, area MAP can vary between successive pulses by up to 5 mm Hg. For mean IOP and JVP, the mean difference between a formula’s MP and the area method’s was less than 1 mm Hg for most formulae. With the pooled data, there was excellent agreement amongst all formulae for MAP with the intra-class correlation coefficient (ICC) ranging from 0.97 to 0.99, while the ICC of most formulae for IOP and JVP was 1.0. CONCLUSIONS Our study suggests that all current formulae are adequate for estimating MAP, though some formulae are not suitable for mean IOP and JVP.

Few-shot segmentation aims to segment objects belonging to a specific class under the guidance of a few annotated examples. Most existing approaches follow the prototype learning paradigm and generate category prototypes by squeezing masked feature maps extracted from images in the support set. These support prototypes may lead to inaccurate predictions when directly compared with features extracted from the query set due to the considerable distribution discrepancy between support and query features. We propose a query-guided prototype learning architecture to address this problem from two aspects: (i) We propose a cross-alignment loss for training the segmentation decoder. This loss function will help the decoder improve its robustness against the distribution discrepancy between support and query features. (ii) We build a dynamic fusion module to strengthen the original support prototype with another prototype extracted from query features. Experiments show that our method achieves promising results compared to previous prototype learning methods on PASCAL-5 i and COCO-20 i datasets.

Melody generation aims to learn the distribution of real melodies to generate new melodies conditioned on lyrics, which has been a very interesting topic in the area of artificial intelligence and music. However, a challenging issue still limits the quality and reliability of melody generation conditioned on lyrics: how to enhance the interpretability between the input lyrics and generated melodies so humans can understand their relationships. To solve this issue, in this article, we propose a model for melody generation from lyrics with local interpretability, which contains two significant contributions: (i) Mutual information between input lyrics and generated melody is exploited to instruct the training of the network, which avoids the loss of content consistency during the training stage. (ii) Transformer is explored to efficiently extract semantic features from lyrics sequences, which provides more interpretable correlations between different syllables in lyrics. Experiments on a large-scale dataset with paired lyrics-melodies demonstrate that the proposed approach can generate higher-quality melodies from lyrics compared with existing methods.

Cross-modal retrieval is to utilize one modality as a query to retrieve data from another modality, which has become a popular topic in information retrieval, machine learning, and databases. Finding a method to effectively measure the similarity between different modality data is the major challenge of cross-modal retrieval. Although several research works have calculated the correlation between different modality data via learning a common subspace representation, the encoder’s ability to extract features from multi-modal information is not satisfactory. In this article, we present a novel variational autoencoder architecture for audio–visual cross-modal retrieval by learning paired audio–visual correlation embedding and category correlation embedding as constraints to reinforce the mutuality of audio–visual information. On the one hand, audio encoder and visual encoder separately encode audio data and visual data into two different latent spaces. Further, two mutual latent spaces are respectively constructed by canonical correlation analysis. On the other hand, probabilistic modeling methods are used to deal with possible noise and missing information in the data. Additionally, in this way, the cross-modal discrepancies from intra-modal and inter-modal information are simultaneously eliminated in the joint embedding subspace. We conduct extensive experiments over two benchmark datasets. The experimental results confirm that the proposed architecture is effective in learning audio–visual correlation and is appreciably better than the existing cross-modal retrieval methods.

Audio-visual tracks in video contain rich semantic information with potential in many applications and research. Since the audio-visual data have inconsistent distributions and because of the heterogeneous nature of representations, the heterogeneous gap between modalities makes them impossible to compare directly. To bridge the modality gap, a frequently adopted approach is to simultaneously project audio-visual data into a common subspace to capture the commonalities and characteristics of modalities for measurement, which has been extensively studied in relation to the issues of modality-common and modality-specific feature learning in previous research. However, it is difficult for existing methods to address the tradeoff between both issues; e.g., the modality-common feature is learned from the latent commonalities of audio-visual data or the correlated features as aligned projections, in which the modality-specific feature can be lost. To solve the tradeoff, we propose a novel end-to-end architecture, which synchronously projects audio-visual data into the explicit and the implicit dual common subspaces. The explicit subspace is used to learn modality-common features and reduce the modality gap of explicitly paired audio-visual data, where the representation-specific details are abandoned to retain the common underlying structure of audio-visual data. The implicit subspace is used to learn modality-specific features, where each modality privately pulls apart the feature distances between different categories to maintain the category-based distinctions, by minimizing the distance between audio-visual features and corresponding labels. The comprehensive experimental results on two audio-visual datasets, VEGAS and AVE, demonstrate that our proposed model for using two different common subspaces for audio-visual cross-modal learning is effective and significantly outperforms the state-of-the-art cross-modal models that learn features from a single common subspace by 4.30% and 2.30% in terms of average MAP on the VEGAS and AVE datasets, respectively.

Carbon–carbon bond cleavage reactions, adapted to deconstruct aliphatic hydrocarbon polymers and recover the intrinsic energy and carbon value in plastic waste, have typically been catalysed by metal nanoparticles or air-sensitive organometallics. Metal oxides that serve as supports for these catalysts are typically considered to be inert. Here we show that Earth-abundant, non-reducible zirconia catalyses the hydrogenolysis of polyolefins with activity rivalling that of precious metal nanoparticles. To harness this unusual reactivity, our catalytic architecture localizes ultrasmall amorphous zirconia nanoparticles between two fused platelets of mesoporous silica. Macromolecules translocate from bulk through radial mesopores to the highly active zirconia particles, where the chains undergo selective hydrogenolytic cleavage into a narrow, C18-centred distribution. Calculations indicated that C–H bond heterolysis across a Zr–O bond of a Zr(O)2 adatom model for unsaturated surface sites gives a zirconium hydrocarbyl, which cleaves a C–C bond via β-alkyl elimination.

The link dynamic vehicle count is a spatial variable that measures the traffic state of road sections, which reflects the actual traffic demand. This paper presents a hybrid deep learning method that combines the gated recurrent unit (GRU) neural network model with automatic hyperparameter tuning based on Bayesian optimization (BO) and the improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) model. There are four steps in this hybrid approach. First, the ICEEMDAN is employed to decompose the link dynamic vehicle count time series data into several intrinsic components. Second, the components are predicted by the GRU model. At the same time, the Bayesian optimization method is utilized to automatically optimize the hyperparameters of the GRU model. Finally, the predicted subcomponents are reconstructed to obtain the final prediction results. The proposed hybrid deep learning method is tested on two roads of Hangzhou, China. Results show that, compared with the 12 benchmark models, the proposed hybrid deep learning model achieves the best performance in link dynamic vehicle count forecasting.

The problem of voltage restoration and current balancing has been a hot topic in the study of DC microgrids (MGs), and how to engineer a secondary controller to solve the conflict between these two is the concern of this paper. With the above consideration, an optimal distributed data-driven secondary control strategy is proposed in this paper, which relies entirely on the input and output (current measurements) data of DC MGs. The proposed framework completes the design of a secondary control strategy for DC MGs based on a data model for the first time, which incorporates the physical coupling relationship between voltage and current into the controller design and is independent of the global circuit parameters and topology information. As a result of the proposed cost function with future current prediction, the designed secondary control strategy not only retains the advantages of consensus control, such as facilitating the plug-and-play of generation units, but also enables DC MGs to achieve better output current quality at steady state, fast dynamic response and improves robustness to external disturbances. Unlike the existing schemes based on average voltage observer, this paper resorts to a combination of limiting the voltage reference value generated by the secondary controller and a current cooperative strategy based on the learned data model to achieve an easy-to-use trade-off between the two objectives of voltage regulation and current sharing in DC MGs. From the design and implementation point of view, the solution proposed in this paper is going to be simpler. Then, an analytical model of a closed-loop DC MG system under the proposed control scheme is developed, and the analysis of stability and consensus of weighted output current is given accordingly. Finally, the proposed control strategy is tested on a hardware DC MG system with solar panels and an MPPT controller to verify its feasibility.

Pt/SrTiO3 nanoparticle catalysts have been synthesized by surface organometallic chemistry in solution on a 5 g scale. Pt/SrTiO3 selectively and repeatedly upcycles isotactic polypropylene into uniform liquid products with Mn ∼ 200 Da.

Twenty DAB-peptide and DAB-dipeptide derivatives were synthesized from d-tartrate-derived nitrone, and five of them were found to be bovine liver β-galactosidase inhibitors.

In this paper, a converter-based multi-bus DC microgrid (MG) in series is studied, in which the trade-off between voltage recovery and current equalization has been a hot topic of interest. To solve this problem, a distributed learning-based high-order fully actuated (DL-HOFA) secondary control is proposed in this paper, and the simple structure of this control technique facilitates its application in various modern DC MGs with different configurations. Before designing the secondary control protocol, this paper provides a comprehensive description of DC MGs in advance. In the suggested control strategy, the controller design is tightly related to the underlying physical characteristics of the MG, and this prominent feature represents a significant improvement in its adaptability. In addition, the DL-HOFA control obtains fast dynamic and accurate current sharing performance by virtue of the high-order fully actuated DC MG model. The effectiveness of the proposed control method is verified on a real photovoltaic- and battery-based hardware system with maximum power point tracking (MPPT) controller.

N/F co-doped carbon materials prepared with an aniline-type benzoxazine as a carbon precursor show a better catalytic performance for the selective oxidation of 2-methylnapthalene.

Acquiring metal-free horizontal single-walled carbon nanotube (SWNT) arrays is of paramount importance for the development of stable nanodevices. However, the majority of SWNTs are prepared with transition metal-based catalysts, which will inevitably leave metallic residuals and deteriorate the device performance. Here, green and low-cost NaCl is developed as a metal-free catalyst. By employing a strategy of rapid nucleation at a higher temperature followed by steady growth at a lower temperature, the production of a well-defined NaCl catalyst capable of growing metal-free horizontal SWNT arrays with an average density of ∼100 tubes per 100 μm is realized. Besides, we prove that the as-grown metal-free SWNT arrays have a unique advantage in preparing stable devices for eliminating the potential risk of local mass catalyst residuals. Hence, the current study can offer a feasible solution to promote practical applications of SWNT-based next-generation nanodevices.

Peer prediction mechanisms incentivize self-interested agents to truthfully report their signals even in the absence of verification by comparing agents’ reports with their peers. We propose two new mechanisms, Source and Target Differential Peer Prediction, and prove very strong guarantees for a very general setting. Our Differential Peer Prediction mechanisms are strongly truthful : Truth-telling is a strict Bayesian Nash equilibrium. Also, truth-telling pays strictly higher than any other equilibria, excluding permutation equilibria, which pays the same amount as truth-telling. The guarantees hold for asymmetric priors among agents, which the mechanisms need not know ( detail-free ) in the single question setting . Moreover, they only require three agents , each of which submits a single item report : two report their signals (answers), and the other reports her forecast (prediction of one of the other agent’s reports). Our proof technique is straightforward, conceptually motivated, and turns on the logarithmic scoring rule’s special properties. Moreover, we can recast the Bayesian Truth Serum mechanism [ 20 ] into our framework. We can also extend our results to the setting of continuous signals with a slightly weaker guarantee on the optimality of the truthful equilibrium.

Inter-organelle interactions play a vital role in diverse biological processes. Thus, chemical tools are highly desirable for understanding the spatiotemporal dynamic interplay among organelles in live cells and

Studying the insertion process of cochlear implant (CI) electrode array (EA) is important to ensure successful, sufficient, and safe implantation. A three-dimensional finite element (FE) model was developed to simulate the insertion process. The cochlear structures were reconstructed from an average statistical shape model (SSM) of human cochlea. The electrode is simplified as a long and tapered beam of homogeneous elastic materials, contacting and interacting with the stiff cochlear structures. A quasi-static insertion simulation was conducted, the insertion force and the contact pressure between the electrode and the cochlear wall, were calculated to evaluate the smoothness of insertion and the risk of potential cochlear trauma. Based on this model, different EA designs were analyzed, including the Young's modulus, the straight or bended shape, the normal or a more tapped section size. The influence of the insertion angle was also discussed. Our simulations indicate that reducing the EA Young's modulus, tapering and pre-bending are effective ways to ensure safe and successful EA implantation. This model is beneficial for optimizing EA designs and is potentially useful for designing patient-specific CI surgery.

Inflammation can contribute to the initiation and progression of atrial fibrillation (AF), and pinocembrin can suppress downstream inflammatory cytokine production by inhibiting the inflammation pathway. In our previous studies, pinocembrin was also beneficial in ameliorating cardiac arrhythmia in different models of rats, such as depression, myocardial infarction, and heart failure. This study aims to investigate the effect of pinocembrin on the susceptibility to AF in isoproterenol-induced rats.Rats were randomly divided into four groups. Pinocembrin was injected through the tail vein. Isoproterenol was treated by intraperitoneal injection for one week (5 mg/kg/day). We evaluated the susceptibility to AF by atrial electrophysiological experiments. Masson staining was used to evaluate the fibrosis area. The protein levels of connexin (Cx) 40, Cav1.2, Kv4.2, collagen I, collagen III, α-SMA, transforming growth factor (TGF)-β, NLRP3, caspase 1, and interleukin (IL)-1β were detected by western blot.Our data demonstrated that pinocembrin could prolong the atrial effective refractory period (ERP) and action potential duration (APD), and decrease AF inducibility. Isoproterenol increased the expression of Cav1.2 and Kv4.2 ion channels whereas pinocembrin could alleviate this change. Pinocembrin could reduce the fibrosis area, fibrosis-related protein collagen I, collagen III, α-SMA, and TGF-β and upregulate gap junction protein Cx40. In addition, pinocembrin reduced the expression of NLRP3, caspase 1, and IL-1β.Our study indicated that pinocembrin was beneficial to alleviate atrial electrical remodeling and fibrosis. Accompanied the downregulation of ion channels and upregulation of gap junction protein Cx40. Pinocembrin may produce these effects by inhibiting the NLRP3 pathway.