Distinguishing Ability Research Articles

Red-emitting ionic fluorophores: anion-dependent tunable fluorescence, chlorinated solvent sensing, white light emission and latent fingerprinting

TPA-P ionic fluorophores exhibited anion-dependent tunable fluorescence, chlorinated solvent sensing with isomer distinguishing ability, white light emission with blue-emitting PSS and excellent LFPs imaging on various substrates.

Serum metabolomics reveals the metabolic profile and potential biomarkers of ankylosing spondylitis.

Ankylosing spondylitis (AS) is a chronic systemic inflammatory disease that significantly impairs physical function in young individuals. However, the identification of radiographic changes in AS is frequently delayed, and the diagnostic efficacy of biomarkers like HLA-B27 remains moderately effective, with unsatisfactory sensitivity and specificity. In contrast to existing literature, our current experiment utilized a larger sample size and employed both untargeted and targeted UHPLC-QTOF-MS/MS based metabolomics to identify the metabolite profile and potential biomarkers of AS. The results indicated a notable divergence between the two groups, and a total of 170 different metabolites were identified, which were associated with the 6 primary metabolic pathways exhibiting a correlation with AS. Among these, 26 metabolites exhibited high sensitivity and specificity with area under curve (AUC) values greater than 0.8. Subsequent targeted quantitative analysis discovered 3 metabolites, namely 3-amino-2-piperidone, hypoxanthine and octadecylamine, exhibiting excellent distinguishing ability based on the results of the ROC curve and the Random Forest model, thus qualifying as potential biomarkers for AS. Summarily, our untargeted and targeted metabolomics investigation offers novel and precise insights into potential biomarkers for AS, potentially enhancing diagnostic capabilities and furthering the comprehension of the condition's pathophysiology.

Substantiation of source data on the parametric algorithms for the classification of weather hazards

The meteorological situation is one of the decisive factors determining the safety and frequency of civil aviation flights. Weather hazards (WH), associated with cumulonimbus clouds, such as a heavy shower, thunderstorm, hail, combined with high atmosphere turbulence, quite often lead to aviation events and even accidents. Currently, a domestic weather radar system of the near airfield zone (WR) “Monocle” has been developed and successfully operated. The criteria for the classification of meteorological phenomena (MP), used in the WR, have been developed individually for each phenomenon and have some heuristic character. These criteria are cumbersome and complicate the process of automating the WH classification. In this case, there is a natural desire to generalize the criteria and optimize them in accordance with the theory of distinguishing statistical hypotheses. This article discusses the application of the Bayesian approach to the WH classification. The statistical Bayesian decision theory assumes decision-making in terms of the probability theory when all significant probabilistic values, so-called sufficient statistics, are known. In order to obtain statistical descriptions of the probability distributions of reflectivity and the eddy dissipation rate (EDR), an analysis of radar signals, reflected from such MP as a rain shower, thunderstorm, hail was carried out. The article provides brief descriptions of the methods of conducting experiments to form statistical database and its analysis. Based on the above methods, the statistical parameter H(EDRmax) analysis for a rain shower, the amplitude distribution of reflectivity parameters and the EDR (Zmax, EDRmax) for thunderstorms and hail was carried out, which showed the low distinguishing ability of each individual parameter when solving the problem to classify MP within the assigned alphabet. The obvious solution is dictated by the theory of recognition. To increase the classification confidence, it is essential to share information parameters, for example, in the form of multidimensional distribution densities of the probabilities of random parameters. The article presents a parametric description of the MP “rain shower-thunderstorm-hail” classification features. An analysis to evaluate the probabilistic characteristics of the WH classification for the adopted empirical classification criteria in the WR shows that the adopted criteria are far from optimal in terms of the probabilities of the correct classification, especially in the rain shower case. It is obvious that a problem solution of the assigned classification confidence is associated with the optimization of the feature space and classification criteria. Based on the data obtained, it is necessary to build an algorithm to classify the WH “rain shower-thunderstorm-hail”.

Fast intensive validation on blockchain with scale-out dispute resolution

Blockchain heralds the dawn of decentralized applications that coordinate proper computations without requiring prior trust. Existing blockchain solutions, however, are incapable of dealing with intensive validation. Duplicated execution results in limited throughput and unacceptably high costs. Furthermore, the absence of secure incentive mechanisms derives undesired dilemmas among rational verifiers.In this work, we present Lever-FS, a practical blockchain validation framework that makes intensive validation cost-efficient and incentive-compatible among rational verifiers. It is faster than previous constructions since full-fledged scalability is achieved over optimistic execution, dispute resolution, and backbone confirmation of every potential workload. Lever-FS first curtails the scale of each validation to a single node and introduces novel challenge-response games between potential adversaries and rational participants, optimistically optimizing validation redundancy according to the practical adversarial capability confronted. When there is a rich and stubborn adversary, the backstop protocol is then activated to resolve intricate disputes via a threshold voting supported by concurrent redundant executions. Throughout the game, compelling incentive design efficiently transfers the adversary’s budget to proliferated task rewards for subsequent executions, therefore allowing the user to lever sufficient endorsements for the correct verdict with minimum expense. In addition, fair incentive distribution mechanisms are designed to circumvent the well-known Verifier’s Dilemma. Finally, we accelerate Lever-FS with sharding to enable scale-out backbone confirmation, seamless workload balancing, and optimized unanimous assertion across multiple independent validation instances.Experiments reveal that Lever-FS significantly improves the throughput while lowering expenses of intensive validation with a slight tradeoff in latency. Compared to state-of-the-art alternatives, it removes their brittle reliance on altruism, dense interactions, or massive computational power. It is also robust to conceivable attacks on validation and performs distinguishable ability to purify Byzantine participants.

Discrimination of artificial strawberry aroma by electronic nose based on nanocomposites

AbstractAn electronic nose containing an array of gas sensors with interdigitated electrodes coated with nanocomposites based on polyaniline (Pani) combined with carboxylated multiwalled carbon nanotubes doped with different acids was used to discriminate artificial strawberry aromas. Principal component analysis (PCA) and linear discriminant analysis (LDA) were used to investigate if the electronic nose was able to distinguish between different strawberry aromas. In addition, a nonlinear projection of the interactive document map technique (IDMAP) was applied to evaluate the distinction ability of the gas sensor array. The sensing layers showed tubular characteristics for Pani/MWCNT_COOH morphology and the group's representative doped state. The main and most common compound identified by gas chromatography in strawberry aromas was ethyl butyrate. The gas sensors showed high sensitivity (>6.55 mV/ppm), low limit of detection (LD <0.31 ppm), limit of quantification (LQ <5.43 ppm), short response time (<63 s) and recovery time (<10.15 s) for different artificial strawberry aromas. The results demonstrate that the electronic nose could differentiate among the artificial strawberry aromas. The volatile profile of strawberry aromas was distinguished into distinct groups using both LDA and PCA analysis. The use of IDMAP permitted the identification of the major contributors to the distinguishing ability of gas sensors with a high silhouette coefficient (0.91). The sensor array was efficient at aroma discrimination, demonstrating excellent potential for food analysis in industrial applications.Practical applicationsThe distinction between natural and artificial flavors has significantly shrunk as volatile chemical technology has advanced. Historically, it was difficult to industrially replicate natural aromas. Identification and analysis of aromas are challenging tasks. Targeted aroma compounds are frequently present in low concentrations (parts per million), which makes it very challenging to identify them. The electronic nose system is a promising new analytical method that can quickly and accurately, without pre‐treating the sample, detect and identify the full information of aroma compounds in real time. It is generally accepted as a nondestructive, affordable, and portable method with high feasibility, speed, and ease of use. For qualitative differentiation and distinguishing, PCA, LDA, and IDMAP methods were also used.

G-triplex/hemin DNAzyme mediated colorimetric aptasensor for Escherichia coli O157:H7 detection based on exonuclease III-assisted amplification and aptamers-functionalized magnetic beads

Escherichia coli O157: H7 (E. coli O157: H7) is one of the most common foodborne pathogens and is widespread in food and the environment. Thus, it is significant for rapidly detecting E. coli O157: H7. In this study, a colorimetric aptasensor based on aptamer-functionalized magnetic beads, exonuclease III (Exo III), and G-triplex/hemin was proposed for the detection of E. coli O157: H7. The functional hairpin HP was designed in the system, which includes two parts of a stem containing the G-triplex sequence and a tail complementary to cDNA. E. coli O157: H7 competed to bind the aptamer (Apt) in the Apt-cDNA complex to obtain cDNA. The cDNA then bound to the tail of HP to trigger Exo III digestion and release the single-stranded DNA containing the G-triplex sequence. G-triplex/hemin DNAzyme could catalyze TMB to produce visible color changes and detectable absorbance signals in the presence of H2O2. Based on the optimal conditions, E. coli O157: H7 could be detected down to 1.3 × 103 CFU/mL, with a wide linear range from 1.3 × 103 to 1.3 × 107 CFU/mL. This method had a distinguished ability to non-target bacteria, which showed good specificity. In addition, the system was successfully applied to detect E. coli O157: H7 in milk samples.

Benzimidazolium ionic liquid functionalized star-polycaprolactone stationary phase for capillary gas chromatography

In this work, an innovative ionic liquid functionalized star-shaped polymer is synthesized by coupling star-polycaprolactone with benzimidazolium (star-PCL-Bim), and statically coated into a capillary column for gas chromatographic (GC) separations. The fabricated star-PCL-Bim capillary column with moderate-polar nature reveals a high column efficiency of 3598 plates/m for n-dodecane at 120 °C, and advantageous distinguishing ability ( R > 1.5) for 9 aromatic isomer mixtures and 12 cis-/ trans -isomers. Its excellent resolving capacity can be credited to the architecture of star-PCL-Bim, the PCL arms, and π-electron-rich benzimidazolium cations, and it results in a quite different retention behavior of star-PCL-Bim from that of star-PCL and commercial HP-35 columns, with “dual-nature” selectivity in separating a mixture including 29 compounds. These favorable chromatographic characteristics demonstrate the great application potential of this material in GC, and provide a prospective strategy to obtain high-performance stationary phases.

A no-reference underwater image quality evaluator via quality-aware features

In this paper, we propose a novel no-reference evaluator based on quality-aware features, called QA-UIQE, for underwater image quality assessment. QA-UIQE extracts and fuses a set of quality-aware features including naturalness, color, contrast, sharpness, and structure. Technically, we first present a new color-cast weighted colorfulness measurement as well as color consistency measurement to characterize color, and design a saliency-weighted contrast measurement to improve the distinguishing ability of measuring contrast. Also, the locally mean subtracted and contrast normalized, maximum local variation, and local entropy are incorporated to measure naturalness, sharpness and structure, respectively. Afterward, we integrate the feature vectors extracted from the training set into Gaussian process regression to predict the image quality. Moreover, we collect a real-world underwater image dataset for testing the generalization ability of our method. The experimental results illustrate that our QA-UIQE has a superior prediction accuracy and is highly consistent with human visual perception.

NCL++: Nested Collaborative Learning for long-tailed visual recognition

Long-tailed visual recognition has received increasing attention in recent years. Due to the extremely imbalanced data distribution in long-tailed learning, the learning process shows great uncertainties. For example, the predictions of different experts on the same image vary remarkably despite the same training settings. To alleviate the uncertainty, we propose a Nested Collaborative Learning (NCL++) which tackles the long-tailed learning problem by a collaborative learning. To be specific, the collaborative learning consists of two folds, namely inter-expert collaborative learning (InterCL) and intra-expert collaborative learning (IntraCL). InterCL learns multiple experts collaboratively and concurrently, aiming to transfer the knowledge among different experts. IntraCL is similar to InterCL, but it aims to conduct the collaborative learning on multiple augmented copies of the same image within the single expert. To achieve the collaborative learning in long-tailed learning, the balanced online distillation is proposed to force the consistent predictions among different experts and augmented copies, which reduces the learning uncertainties. Moreover, in order to improve the meticulous distinguishing ability on the confusing categories, we further propose a Hard Category Mining (HCM), which selects the negative categories with high predicted scores as the hard categories. Then, the collaborative learning is formulated in a nested way, in which the learning is conducted on not just all categories from a full perspective but some hard categories from a partial perspective. Extensive experiments manifest the superiority of our method with outperforming the state-of-the-art whether with using a single model or an ensemble. The code will be publicly released.

Plasma Extracellular Vesicle LncRNA H19 as a Potential Diagnostic Biomarker for Inflammatory Bowel Diseases.

Inflammatory bowel disease (IBD) is a complex gastrointestinal disease with 2 main subtypes of Crohn's disease (CD) and ulcerative colitis (UC), whose diagnosis mainly depends on the medical history, clinical symptoms, endoscopic, histologic, radiological, and serological findings. Extracellular vesicles (EVs) are now considered an additional mechanism for intercellular communication, allowing cells to exchange biomolecules. Long noncoding RNAs (lncRNAs) that are enriched in EVs have been defined as an ideal diagnostic biomarker for diseases. In this study, we investigated the expression differences of 5 lncRNAs in tissue and plasma EVs of active IBD patients compared with patients in the remission phase and healthy controls to introduce an EV-lncRNA as a noninvasive IBD diagnostic biomarker. Twenty-two active IBD patients, 14 patients in the remission phase, 10 active rheumatoid arthritis (RA) patients, 14 irritable bowel syndrome (IBS) patients, and 22 healthy individuals were recruited in the discovery cohort. In addition, 16 patients with active IBD, 16 healthy controls, 10 inactive IBD patients, 12 active RA patients, and 14 IBS patients were also included in the validation cohort. The expression levels of 5 lncRNAs in tissue and EV-plasma were evaluated by quantitative real-time polymerase chain reaction (qRT-PCR) . Machine learning and receiver operating characteristic (ROC) curve analysis were performed to investigate the distinguishing ability of the candidate biomarkers. While the expression levels of lncRNAs CDKN2B-AS1, GAS5, and TUG1 were significantly downregulated, lncRNAs H19 and CRNDE were overexpressed in active IBD lesions. Expression of H19 was detected in plasma EVs whose isolation had been confirmed via dynamic light scattering, microscopy images, and western blotting. The classification results demonstrated the excellent ability of H19 in distinguishing IBD/active from IBD/remission, healthy control, RA, and IBS (area under the ROC curve = 0.95, 0.97,1, and 0.97 respectively). Our study suggests that circulating EV-lncRNA H19 exhibited promising potential for the diagnosis of active IBD.

Genomic diversity and taxonomic marker for Arcobacter species.

Arcobacter was recognized as an emerging enteropathogen and controversies regarding its classification persisted. This study aimed to reevaluate the taxonomy of Arcobacter utilizing the 16S rRNA gene, 23S rRNA gene, single-copy orthologous genes, as well as genomic indices such as Average Nucleotide Identity (ANI) and in silico DNA-DNA hybridization (isDDH). The taxonomy of this genus was reevaluated in this study using multiple indices with a dataset of 371 genomes comprising 34 known species and 14 potentially new species. Good discrimination could be achieved only in some species but not for the species with higher sequence similarity using the comparisons of the 16S rRNA gene and 23S rRNA gene sequences. A high-accuracy phylogenomic approach for Arcobacter was established using 84 single-copy orthologous genes obtained through various bioinformatics methods. One marker gene (gene711), which was found to possess the same distinguishing ability as ANI, isDDH, and single-copy orthologous methods, was identified as a reliable locus for inferring the phylogeny of the genus. The effective species classification was achieved by employing gene711 with a sequence similarity exceeding 96%, even for species like A. cloacae, A. lanthieri, and A. skirrowii, which exhibited ambiguous classification using ANI and isDDH. Additionally, excellent subspecies categorizing among A. cryaerophilus could be distinguished using gene711. In conclusion, this framework strategy had the potential advantage of developing rapid species identification, particularly for highly variable species, providing a novel insight into the behavior and characteristics of Arcobacter.

Towards a metrics suite for evaluating cache side-channel vulnerability: Case studies on an open-source RISC-V processor

Caches are key microarchitectural components of modern processors to improve memory access speed. However, attackers can readily implement timing side-channel attacks by exploiting the inherent time difference between cache hits and misses, which brings serious security threats to computer systems. In recent years, cache attacks have achieved fruitful consequences regarding attack techniques, targets, and platforms. At the same time, in order to measure the cache vulnerability, researchers have been looking at verifying the security policies and have proposed a variety of evaluation metrics. However, there are certain limitations in the current metrics, such as harsh assumptions, the distinguishing ability cannot be maintained all the time, the false positive and false negative noises can not be explicitly quantified, other tools are needed, and the verification environment is difficult to reproduce. To face these issues, we proposed a set of evaluation metrics, which includes the highest score (HS), minimum rounds to disclosure (MRD), and key score scissors differential (KSSD); we give the formal expressions in practice and theoretical, respectively. In order to facilitate the accurate reproduction of the experimental environment and results, we build a dual-core RISC-V bare-metal system based on the open-source framework Chipyard. On this basis, we conduct comprehensive evaluations to exploit vulnerabilities within the RISC-V cache architecture to verify the validity of the metrics suite. Furthermore, we compared the metrics with the commonly used success rate (SR) and guessing entropy (GE). The comparison results show that our proposed metrics suite can accurately depict the effects of attacks. Moreover, KSSD can maintain discrimination and converge quickly; MRD has more practical guidance; HS can specifically describe false negative noises. Finally, we perform a theoretical evaluation of AES algorithms with different T-table implementations using the proposed metrics, analyze the system's false positive and false negative noises, and suggest how the number of cache evictions can be determined under the random replacement policy.

Development and Validation of a Novel Prognosis Model Based on a Panel of Three Immunogenic Cell Death-Related Genes for Non-Cirrhotic Hepatocellular Carcinoma.

The accurate prediction of non-cirrhotic hepatocellular carcinoma (NCHCC) risk facilitates improved surveillance strategy and decreases cancer-related mortality. This study aimed to explore the correlation between immunogenic cell death (ICD) and NCHCC prognosis using The Cancer Genome Atlas (TCGA) datasets, and the potential prognostic value of ICD-related genes in NCHCC. Clinical and transcriptomic data of patients with NCHCC patients were retrieved from TCGA database. Weighted gene co-expression network analysis was performed to obtain the NCHCC phenotype-related module genes. Consensus clustering analysis was performed to classify the patients into two clusters based on intersection genes among differentially expressed genes (DEGs) between cancer and adjacent tissues, NCHCC phenotype-related genes, and ICD-related genes. NCHCC-derived tissue microarray was used to evaluate the correlation of the expression levels of key genes with NCHCC prognosis using immunohistochemical staining. Cox regression analyses were performed to construct a prognostic risk score model comprising three genes (TMC7, GRAMD1C, and GNPDA1) based on DEGs between two clusters. The model stratified patients with NCHCC into two risk groups. The overall survival (OS) of the high-risk group was significantly lower than that of the low-risk group. Univariable and multivariable Cox regression analyses revealed that these signature genes are independent predictors of OS. Functional analysis revealed differential immune status between the two risk groups. Next, a nomogram was constructed, which demonstrated the potent distinguishing ability of the developed model based on receiver operating characteristic curves. In vitro functional validation revealed that the migration and invasion abilities of HepG2 and Huh7 cells were upregulated upon GRAMD1C knockdown but downregulated upon TMC7 knockdown. This study developed a prognostic model comprising three genes, which can aid in predicting the survival of patients with NCHCC and guide the selection of drugs and molecular markers for NCHCC.

Multiparametric MRI-Based Radiomics Signature with Machine Learning for Preoperative Prediction of Prognosis Stratification in Pediatric Medulloblastoma

Despite advances in risk-stratified treatment strategies for children with medulloblastoma (MB), the prognosis for MB with short-term recurrence is extremely poor, and there is still a lack of evaluation of short-term recurrence risk or short-term survival. This study aimed to construct and validate a radiomics model for predicting the outcome of MB based on preoperative multiparametric magnetic resonance images (MRIs) and to provide an objective for clinical decision-making. The clinical and imaging data of 64 patients with MB admitted to Shenzhen Children's Hospital from December 2012 to December 2021 and confirmed by pathology were retrospectively collected. According to the 18-month progression-free survival, the cases were classified into a good prognosis group and a poor prognosis group, and all cases were divided into training group (70%) and validation group (30%) randomly. Radiomics features were extracted from MRI of each child. The consistency test, t-test, and the least absolute shrinkage and selection operator were used for feature selection. The support vector machine (SVM) and receiver operator characteristic were used to evaluate the distinguishing ability of the selected features to the prognostic groups. RAD score was calculated based on the selected features. The clinical characteristics and RAD score were included in the multivariate logistic regression, and prediction models were constructed by screening out independent influences. The radiomics nomogram was constructed, and its clinical significance was evaluated. A total of 1930 radiomic features were extracted from the images of each patient, and 11 features were included in the construction of radiomics score after selected. The area under the curve (AUC) values of the SVM model in the training and validation groups were 0.946 and 0.797, respectively. The radiomics nomogram was constructed based on the training cohort, and the AUC values in the training group and the validation group were 0.926 and 0.835, respectively. The results of clinical decision curve analysis showed that a good net benefit could be obtained from the nomogram. The radiomics nomogram established based on MRI can be used as a noninvasive predictive tool to evaluate the prognosis of children with MB, which is expected to help neurosurgeons better conduct preoperative planning and patient follow-up management.

Automatic depression diagnosis through hybrid EEG and near-infrared spectroscopy features using support vector machine.

Depression is a common mental disorder that seriously affects patients' social function and daily life. Its accurate diagnosis remains a big challenge in depression treatment. In this study, we used electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) and measured the whole brain EEG signals and forehead hemodynamic signals from 25 depression patients and 30 healthy subjects during the resting state. On one hand, we explored the EEG brain functional network properties, and found that the clustering coefficient and local efficiency of the delta and theta bands in patients were significantly higher than those in normal subjects. On the other hand, we extracted brain network properties, asymmetry, and brain oxygen entropy as alternative features, used a data-driven automated method to select features, and established a support vector machine model for automatic depression classification. The results showed the classification accuracy was 81.8% when using EEG features alone and increased to 92.7% when using hybrid EEG and fNIRS features. The brain network local efficiency in the delta band, hemispheric asymmetry in the theta band and brain oxygen sample entropy features differed significantly between the two groups (p < 0.05) and showed high depression distinguishing ability indicating that they may be effective biological markers for identifying depression. EEG, fNIRS and machine learning constitute an effective method for classifying depression at the individual level.

Adsorption/Desorption of Cationic-Hydrophobic Peptides on Zwitterionic Lipid Bilayer Is Associated with the Possibility of Proton Transfer.

Cell-penetrating peptides (CPPs) are short peptides built up from dominantly cationic and hydrophobic amino acid residues with a distinguished ability to pass through the cell membrane. Due to the possibility of linking and delivering the appropriate cargo at the desired location, CPPs are considered an economic and less invasive alternative to antibiotics. Besides knowing that their membrane passage mechanism is a complex function of CPP chemical composition, the ionic strength of the solution, and the membrane composition, all other details on how they penetrate cell membranes are rather vague. The aim of this study is to elucidate the ad(de)sorption of arginine-/lysine- and phenylalanine-rich peptides on a lipid membrane composed of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) lipids. DSC and temperature-dependent UV-Vis measurements confirmed the impact of the adsorbed peptides on thermotropic properties of DPPC, but in an inconclusive way. On the other hand, FTIR spectra acquired at 30 °C and 50 °C (when DPPC lipids are found in the gel and fluid phase, respectively) unambiguously confirmed the proton transfer between particular titratable functional groups of R5F2/K5F2 that highly depend on their immediate surroundings (DPPC or a phosphate buffer). Molecular dynamic simulations showed that both peptides may adsorb onto the bilayer, but K5F2 desorbs more easily and favors the solvent, while R5F2 remains attached. The results obtained in this work highlight the importance of proton transfer in the design of CPPs with their desired cargo, as its charge and composition dictates the possibility of entering the cell.

Multi‐label feature selection based on relative entropy and fuzzy neighborhood mutual discrimination index

AbstractMulti‐label feature selection eliminates irrelevant and redundant features, and then improves the performance of multi‐label classification models. Most multi‐label feature selection algorithms assume that the training set contains logical labels, which means that labels are equally important for instances. However, in practical applications, there are different importances with respect to labels. To solve the problem, a multi‐label feature selection method based on relative entropy and fuzzy neighborhood mutual discriminant index is proposed. Firstly, logical labels are converted to label distribution through label enhancement. Secondly, the neighborhood and relative entropy are introduced into the label distribution, the label neighborhood similarity matrix is constructed to describe the similarity of samples under label space. Finally, the fuzzy neighborhood mutual discrimination index is used to combine the candidate features with the label neighborhood similarity matrix, which is used to judge the distinguishing ability of the candidate features. Comprehensive experiment of eight multi‐label datasets shows that the proposed algorithm has better classification performance than other compared algorithms.

Congenital heart diseases with airway stenosis: a predictive nomogram to risk-stratify patients without airway intervention

BackgroundThis study focused on congenital heart disease (CHD) patients complicated with airway stenosis (AS) without airway intervention and aimed to identify the patients with potential risks.MethodsPatients diagnosed with CHD and AS were enrolled in this retrospective study. The primary outcome was defined as a postoperative mechanical ventilation duration of more than two weeks. We constructed a prediction model to predict the risk of prolonged mechanical ventilation (PMV).ResultsA total of 185 patients diagnosed with CHD and AS in Fuwai Hospital from July 2009 to December 2022 were included in the study. Weight at CHD surgery, cardiopulmonary bypass (CPB) duration, complex CHD and comorbid tracheobronchomalacia were identified as risk factors and included in the model. The ROC curve showed a good distinguishing ability, with an AUC of 0.847 (95% CI: 0.786–0.908). According to the optimal cut-off value of the ROC curve, patients were divided into high- and low-risk groups, and the subsequent analysis showed significant differences in peri-operative characteristics and in-hospital deaths.ConclusionsWith the predictive model, several factors could be used to assess the risky patients with PMV. More attention should be paid to these patients by early identification and routine surveillance.

DB-DCAFN: dual-branch deformable cross-attention fusion network for bacterial segmentation

Sputum smear tests are critical for the diagnosis of respiratory diseases. Automatic segmentation of bacteria from sputum smear images is important for improving diagnostic efficiency. However, this remains a challenging task owing to the high interclass similarity among different categories of bacteria and the low contrast of the bacterial edges. To explore more levels of global pattern features to promote the distinguishing ability of bacterial categories and maintain sufficient local fine-grained features to ensure accurate localization of ambiguous bacteria simultaneously, we propose a novel dual-branch deformable cross-attention fusion network (DB-DCAFN) for accurate bacterial segmentation. Specifically, we first designed a dual-branch encoder consisting of multiple convolution and transformer blocks in parallel to simultaneously extract multilevel local and global features. We then designed a sparse and deformable cross-attention module to capture the semantic dependencies between local and global features, which can bridge the semantic gap and fuse features effectively. Furthermore, we designed a feature assignment fusion module to enhance meaningful features using an adaptive feature weighting strategy to obtain more accurate segmentation. We conducted extensive experiments to evaluate the effectiveness of DB-DCAFN on a clinical dataset comprising three bacterial categories: Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. The experimental results demonstrate that the proposed DB-DCAFN outperforms other state-of-the-art methods and is effective at segmenting bacteria from sputum smear images.

Recognize enhanced temporal-spatial-spectral features with a parallel multi-branch CNN and GRU.

Deep learning has been applied to the recognition of motor imagery electroencephalograms (MI-EEG) in brain-computer interface, and the performance results depend on data representation as well as neural network structure. Especially, MI-EEG is so complex with the characteristics of non-stationarity, specific rhythms, and uneven distribution; however, its multidimensional feature information is difficult to be fused and enhanced simultaneously in the existing recognition methods. In this paper, a novel channel importance (NCI) based on time-frequency analysis is proposed to develop an image sequence generation method (NCI-ISG) for enhancing the integrity of data representation and highlighting the contribution inequalities of different channels as well. Each electrode of MI-EEG is converted to a time-frequency spectrum by utilizing short-time Fourier transform; the corresponding part to 8-30Hz is combined with random forest algorithm for computing NCI; and it is further divided into three sub-images covered by α (8-13Hz), β1 (13-21Hz), and β2 (21-30Hz) bands; their spectral powers are further weighted by NCI and interpolated to 2-dimensional electrode coordinates, producing three main sub-band image sequences. Then, a parallel multi-branch convolutional neural network and gate recurrent unit (PMBCG) is designed to successively extract and identify the spatial-spectral and temporal features from the image sequences. Two public four-class MI-EEG datasets are adopted; the proposed classification method respectively achieves the average accuracies of 98.26% and 80.62% by 10-fold cross-validation experiment; and its statistical performance is also evaluated by multi-indexes, such as Kappa value, confusion matrix, and ROC curve. Extensive experiment results show that NCI-ISG + PMBCG can yield great performance on MI-EEG classification compared to state-of-the-art methods. The proposed NCI-ISG can enhance the feature representation of time-frequency-space domains and match well with PMBCG, which improves the recognition accuracies of MI tasks and demonstrates the preferable reliability and distinguishable ability. This paper proposes a novel channel importance (NCI) based ontime-frequencyanalysis to develop an image sequences generation method (NCI-ISG) for enhancing the integrity of data representation and highlighting the contribution inequalities of different channels as well. Then, a parallel multi-branch convolutional neural network and gate recurrent unit (PMBCG) is designed to successively extract and identify the spatial-spectral and temporal features from the image sequences.