Verification Phase Research Articles

Development and verification of automatic recognition system for vocal performance styles based on ResNet

This paper studies the automatic recognition system of vocal performance style based on residual network (ResNet). The system can precisely recognize multiple vocal performance styles through automation technology. Firstly, an audio dataset covering multiple vocal styles was constructed based on the publicly available dataset GTZAN, and the data in the dataset was preprocessed to extract key music features. Subsequently, the ResNet model was chosen as the core recognition component of the system, which was trained to learn the complex relationship between audio features and vocal styles. Finally, the construction of the system was carried out. In the verification phase, the dataset was divided into two parts: the training set and the test set. After training the model using the training set, its performance was comprehensively evaluated on the test set. The experiment showed that the recognition system based on ResNet performed well in precision, recall, and accuracy, with specific values of 95.6%, 91.6%, and 96.7%, respectively. Meanwhile, the recognition accuracy of the ResNet-based system for 10 vocal performance styles was 95.96% on average. Applying ResNet to an automatic recognition system for vocal performance styles not only provides an efficient and reliable solution in this field, but also opens up new horizons for the further application of deep learning in the music industry.

Does a priming warm-up influence the incidence of [Formula: see text] during a ramp test and verification phase?

This study compared the effects of two different warm-up protocols (normal vs. priming) on the oxygen plateau ([Formula: see text]) incidence rate during a ramp test. It also compared the cardiopulmonary responses during the ramp test and subsequent verification phase. Eleven recreational cyclists performed two experimental visits. The first visit required a normal warm-up (cycling at 50 W for 10 min) followed by the ramp test (30 W.min-1) and supramaximal verification phase with 30 min rest between tests. The second visit required a priming warm-up (cycling at 50 W for 4 min increasing to 70% difference between the gas exchange threshold [GET] and maximum work rate [WRmax] for 6 min) followed by the same protocol as in the first visit. Physiological responses were collected during the exercise and compared. Oxygen kinetics ([Formula: see text] Kinetics) and [Formula: see text] incidence rate were determined during the ramp tests for both visits. As planned, following the warm-up the priming visit experienced greater physiological response. However, the incidence rate of [Formula: see text] during the ramp test was the same between visits (73%), and maximal oxygen uptake was not different between visits after the ramp test (normal, 4.0 ± 0.8; primed, 4.0 ± 0.7 L·min-1, p = 0.230) and verification phase (normal, 3.8 ± 0.6; primed, 3.8 ± 0.7 L·min-1, p = 0.924) using a Holm-Bonferroni correction for controlling family-wise error rate. [Formula: see text] Kinetics were not different between visits during the ramp test (normal, 10.8 ± 1.1; primed, 10.8 ± 1.2 mL·min-1·W-1, p = 0.407). The verification phase confirmed [Formula: see text] in 100% for both the normal and priming visits. Our hypothesis that a priming warm-up facilitates the incidence rate of [Formula: see text] during a ramp test is not supported by the results. The verification phase remains a prudent option when determining a 'true' [Formula: see text] is required.

Neutralization of Oxytocin by 355 nm Photocleavage in High Vacuum

Controlling the charge state of biomolecules in high vacuum enables experiments in molecular physics and physical chemistry, particularly in applications of mass spectrometry, trapping and cooling, spectroscopy and matter‐wave interferometry. Here we investigate the synthesis, verification and gas phase photolysis of nitrodibenzofuranyl aryl ether derivatives. These photocages are chosen because their absorption maximum can be tuned to near 355 nm, where intense pulsed laser light is available. We demonstrate efficient gas‐phase cleavage of a photoactive nitrodibenzofuranyl aryl ether derivative, both free and attached to oxytocin. This enables the neutralization of a polypeptide derivative in high vacuum using a wavelength outside the absorption spectrum of natural amino acids.

A Preliminary Numerical Thermal Analysis of an Endometrial Ablation Procedure Based on Foley’s Catheter

Dysfunctional uterine bleeding (DUB) is an ailment affecting a substantial number of women during a significant part of their reproductive years. This condition usually leads to discomfort, mild or severe pain, or, in some extreme cases, anemia. It may be caused by diverse factors such as hormonal issues, structural abnormalities, or even cancer in a woman’s reproductive tract. A typical DUB treatment requires a pharmacological approach through the prescription of drugs. Nonetheless, some patients require a more involved treatment calling for surgical procedures. In order to avoid hysterectomy, which is the complete removal of the uterus, the medical community has devised, over the years, alternative surgical schemes to handle DUB issues. One such treatment is the “endometrial ablation technique” where the inner lining of the uterus is subjected to a controlled destruction by using a hot, approximately constant-temperature fluid inside a balloon, which is placed in contact with the uterine wall of the patient. As a means to avoid the high costs associated with such devices, the medical community has proposed a similar treatment in which the fluid inside the balloon is no longer held at a high constant temperature but rather diminishes as the treatment progresses. The main purpose of this contribution is to critically assess the effectiveness of this second, low-cost endometrial ablation treatment. Consequently, a mathematical model is devised in which Pennes’ bioheat equation is employed to predict the transient temperature field of the uterine wall together with an energy balance that yields the temperature of the saline solution inside the thermal balloon, throughout the duration of the treatment. These equations aresolved by utilizing the finite volume method and the numerical predictions are verified by comparing the present results with similar cases from the literature. Once the verification phase is authenticated, the uterine wall and fluid balloon transient temperature fields are obtained for typical situations reported in the medical literature. On general grounds, the simulations indicate that the extent of the affected tissue is strongly dependent on the volume,on the type of fluid inside the thermal balloon, and on the magnitude of the perfusion of the uterus. Moreover, it was found that the proposed low-cost treatment can indeed be a successful alternative to the standard endometrial ablation process under some viable circumstances that are described in the contribution.

Peak V̇O2 during handcycling in spinal cord injured athletes: incremental versus verification testing.

The present study aimed to compare peak oxygen uptake (V̇O2peak), peak heart rate (HRpeak) and peak O2 pulse during an incremental and a verification test performed on the same day in hand-cyclists with spinal cord injury (SCI). Eight competitive SCI hand-cyclists (age: 23 ± 2.7 years; V̇O2peak: 2.24 ± 0.79 l.min-1 or 36.3 ± 14.0 ml.kg-1.min-1) performed a maximal incremental handcycling test and a verification test to exhaustion at 100% of V̇O2peak speed on an oversized treadmill. The V̇O2peak, HRpeak and peak O2pulse (i.e. VO2/HR) were compared between incremental and verification tests. V̇O2peak obtained in the verification test (2.51 ± 0.96 l.min-1) was significantly higher than values obtained in the incremental test (2.24 ± 0.79 l.min-1; P = 0.018). There was no difference in HRpeak (Incremental: 169 ± 24 bpm; Verification 167 ± 25 bpm; P = 0.130). Peak O2pulse from the verification test (14.6 ± 4.7 ml.beat-1) was higher than incremental test (13.0 ± 3.8 ml.beat-1; P = 0.007). Six subjects presented variability greater than ± 4% in verification test compared with the incremental exercise. In conclusion, the verification test elicited higher V̇O2peak than the incremental test despite the similar HR, suggesting the relevance of verification phase to determine V̇O2peak of hand-cyclists with spinal cord injuries.

Identifying the impact of ARHGAP and MAP gene families on autism spectrum disorders.

The rising incidence of Autism Spectrum Disorder (ASD) has become a major concern, affecting children's psychological well-being and placing a significant strain on healthcare systems. Despite its impact, the etiological mechanisms underpinning ASD remain elusive. This study leveraged dorsolateral prefrontal cortex gene data from 452 individuals of European descent, sourced from the CommonMindConsortium, and examined ASD-related gene expression data from the Gene Expression Omnibus (GEO) database (GSE18123), along with Genome-Wide Association Studies (GWAS) data from the Lundbeck Foundation Integrated Psychiatric Research and Psychiatric Genomics Consortium. Expression quantitative trait loci data were sourced from the GTExv8 database. We employed Transcriptome-Wide Association Studies (TWAS) and Weighted Gene Co-expression Network Analysis (WGCNA) to pinpoint genes within ASD-associated susceptibility gene families (ARHGAP, MAP). Four genes-ARHGAP27, MAPT, ARHGAP19, and MAP1B-were scrutinized, and their biological implications were elucidated through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Protein-Protein Interaction (PPI) analysis and conditional analysis within the TWAS framework helped identify pivotal genes (ARHGAP27, MAPT). A subsequent verification phase involving Mendelian Randomization (MR) evaluated the potential causal links between the identified genes and ASD. The findings revealed no causal association between ARHGAP19, MAP1B, and ASD. In contrast, significant causal relationships were established for ARHGAP27 and MAPT, suggesting that ARHGAP27 may elevate ASD risk as a susceptibility gene, whereas MAPT appears to reduce the risk as a protective gene.

Light and normal weight concretes shear strength estimation using tree-based tunned frameworks

Concrete component shear strength is a complex phenomenon that depends on a number of governing bodies and regulatory mechanisms. Moreover, the addition of new concrete variations like fibered concrete and light-weight concrete increases the complexity of the topic. Consequently, in order to estimate the shear resistance (Vu) of normal-weight concrete (NWC) and light-weight concrete (LWC) components, three machine learning-based methods were developed. The work gives researchers significant tools to predict concrete structural integrity. This helps design safer, more efficient buildings by accounting for complicated material qualities and concrete variants like fibered or lightweight concrete in real-world construction settings. Extreme gradient boosting (XGB) was taken into consideration for this aim. The Chaos Game Optimization (CGO), Giant Trevally Optimizer (GTO), and Mountain Gazelle Optimizer (MGO) algorithms were used in order to identify the hyperparameters effectively (abbreviated as XGBMGO,XGBGTO,andXGBCGO). The collected big data was divided into three subsets: training (75 %: 1194), validating (15 %: 256), and testing (15 %: 256). The XGBGTO, XGBMGOand XGBCGO algorithms have a considerable capacity to precisely forecast the Vuof NWC and LWC. The R2 values for theXGBCGO net were determined to be 0.9959, 0.9957, and 0.9961, respectively, throughout the training, verification, and testing phases. The reasons put forward generally suggest that the XGBCGO analysis can achieve more than the XGBGTO, and XGBMGOframeworks.

Modelling absorption and emission profiles from accretion disc winds with WINE

Context. Fast and massive winds are ubiquitously observed in the UV and X-ray spectra of active galactic nuclei (AGNs) and other accretion-powered sources. Several theoretical and observational pieces of evidence suggest they are launched at accretion disc scales, carrying significant mass and angular momentum. Thanks to such high-energy output, they may play an important role in transferring the energy released by accretion to the surrounding environment. In the case of AGNs, this process can help to set the so-called co-evolution between an AGN and its host galaxy, which mutually regulates their growth across cosmic time. To precisely assess the effective role of UV and X-ray winds at accretion disc scales, it is necessary to accurately measure their properties, including mass and energy rates. However, this is a challenging task, due to both the limited signal-to-noise ratio of available observations and the limitations of the models currently used in the spectral analysis. Aims. We aim to maximise the scientific return of current and future observations by improving the theoretical modelling of these winds through our Winds in the Ionised Nuclear Environment (WINE) model. WINE is a spectroscopic model specifically designed for disc winds in AGNs and compact accreting sources, which couples photoionisation and radiative transfer with special relativistic effects and a three-dimensional model of the emission profiles. Methods. We explore with WINE the main spectral features associated with the disc winds in AGNs, with a particular emphasis on the detectability of the wind emission in the total transmitted spectrum. We explore the impact of the wind ionisation, column density, velocity field, and geometry in shaping the emission profiles. We simulated observations with the X-ray microcalorimeter Resolve on board the recently launched XRISM satellite and the X-IFU on board the future Athena mission. This allows us to assess the capabilities of these telescopes in the study of disc winds in X-ray spectra of AGNs for the typical physical properties and exposure times of the sources included in the XRISM performance verification phase. Results. The wind kinematic and geometry (together with the ionisation and column density) deeply affect both shape and strength of the wind spectral features. Thanks to this, both Resolve and, on a longer timescale, X-IFU will be able to accurately constrain the main properties of disc winds over a broad range of ionisation, column densities, and covering factors. We also investigate the impact of the spectral energy distribution (SED) on the resulting appearance of the wind. Our findings reveal a dramatic difference in the gas opacity when using a soft, Narrow Line Seyfert 1-like SED compared to a canonical powerlaw SED with a spectral index of Γ ≈ 2.

Jointly stochastic fully symmetric interpolatory rules and local approximation for scalable Gaussian process regression

When exploring the broad application prospects of large-scale Gaussian process regression (GPR), three core challenges significantly constrain its full effectiveness: firstly, the O(n3) time complexity of computing the inverse covariance matrix of n training points becomes an insurmountable performance bottleneck when processing large-scale datasets; Secondly, although traditional local approximation methods are widely used, they are often limited by the inconsistency of prediction results; The third issue is that many aggregation strategies lack discrimination when evaluating the importance of experts (i.e. local models), resulting in a loss of overall prediction accuracy. In response to the above challenges, this article innovatively proposes a comprehensive method that integrates third-degree stochastic fully symmetric interpolatory rules (TDSFSI), local approximation, and Tsallis mutual information (TDSFSIRLA), aiming to fundamentally break through existing limitations. Specifically, TDSFSIRLA first introduces an efficient third-degree stochastic fully symmetric interpolatory rules, which achieves accurate approximation of Gaussian kernel functions by generating adaptive dimensional feature maps. This innovation not only significantly reduces the number of required orthogonal nodes and effectively lowers computational costs, but also maintains extremely high approximation accuracy, providing a solid theoretical foundation for processing large-scale datasets. Furthermore, in order to overcome the inconsistency of local approximation methods, this paper adopts the Generalized Robust Bayesian Committee Machine (GRBCM) as the aggregation framework for local experts. GRBCM ensures the harmonious unity of the prediction results of each local model through its inherent consistency and robustness, significantly improving the stability and reliability of the overall prediction. More importantly, in response to the issue of uneven distribution of expert weights, this article creatively introduces Tsallis mutual information as a metric for weight allocation. Tsallis mutual information, with its sensitive ability to capture information complexity, assigns weights to different local experts that match their contribution, effectively solving the problem of prediction bias caused by uneven weight distribution and further improving prediction accuracy. In the experimental verification phase, this article conducted comprehensive testing on multiple synthetic datasets and seven representative real datasets. The results show that the TDSFSIRLA method not only achieves significant reduction in time complexity, but also demonstrates excellent performance in prediction accuracy, fully verifying its significant advantages and broad application prospects in the field of large-scale Gaussian process regression.

Model‐based systems engineering and safety assessment: A workflow for mechatronic systems design

AbstractMechatronic systems become ever more complex because of their increasing number of interconnected safety critical components and sophistication. MBSE (Model‐based Systems Engineering) and MBSA (Model‐Based Safety Assessment) are the most commonly adopted approaches to deal with the design and safety analysis of mechatronic systems. Unfortunately, both approaches are normally adopted separately, especially in the earlier phases of system design, thus leading to a lack of communication between system engineers and the safety team. This work aims to fill that gap at a high level, that is, through process interaction. This paper proposes an enhanced V‐model for the design of safety‐critical mechatronic systems. It relates a system development process with specific safety assessment methods. Specifically, the proposed workflow details exchange flows between the RFLP (Requirements, Functional, Logical, Physical) method, the FHA (Functional Hazard Analysis), the FMEA (Failure Mode and Effects Analysis), the MBSA and simulation, and the FTA (Fault Tree Analysis). These analyses are complemented with multiphysics modeling and simulation to observe system behavior in functional and failure scenarios, with the aim of requirements verification. The design workflow has been applied to a winged Unmanned Aerial Vehicle to apply the parallel process and the necessary interaction of MBSE and MBSA approaches. The information flows between the individual activities proved effective for designing a safe system before the verification phase. The main benefit of the proposed workflow is providing both the design and safety team with some interaction points, thus avoiding a lack of safety‐critical analysis in the early phases of system design.

The eROSITA Final Equatorial Depth Survey (eFEDS). The hard X-ray selected sample

During its calibration and performance verification phase, the eROSITA instrument aboard the Spectrum-RG satellite performed a uniform wide-area X-ray survey of approximately 140 deg2, known as the eROSITA Final Equatorial Depth Survey (eFEDS). The primary aim of eFEDS is to demonstrate the scientific performance to be expected at the end of the eight--pass eROSITA all--sky survey. This survey will provide the first focussed image of the whole sky in the hard X-ray ($>2$ keV) bandpass. The expected source population in this energy range is thus of great interest, particularly for AGN studies. We used a 2.3--5 keV selection to construct a sample of 246 point-like hard X-ray sources for further study and characterisation. These sources are classified as either extragalactic ($ 90$ <!PCT!>) or Galactic ($ 10$ <!PCT!>), with the former consisting overwhelmingly of AGN and the latter active stars. We concentrated our further analysis on the extragalactic AGN sample, describing their X-ray and multi-wavelength properties and comparing them to the eFEDS main AGN sample selected in the softer 0.2--2.3 keV band. The eROSITA hard band selects a subsample of sources that is a factor of more than ten brighter than the eFEDS main sample. The AGN within the hard population reach up to $z=3.2$ but on the whole, they are relatively nearby, with median $z$=0.34 compared to $z$=0.94 for the main sample. The hard survey probes typical luminosities in the range $ X = 43-46$. The X-ray spectral analysis shows significant intrinsic absorption (with $ H >21$) in $ 20$ <!PCT!> of the sources, with a hard X-ray power law continuum with mean $< which is typical of AGN, but slightly harder than the soft-selected eROSITA sample. Around $10$ <!PCT!> of the hard sample show a significant `soft excess' component. The sampled black hole mass distribution in the eFEDS broad-line AGN population is consistent with that of the deeper COSMOS survey that probes a higher redshift population. On the other hand, the Eddington ratios appear systematically lower, which is consistent with the idea that the decline in SMBH activity since $z 1$ is due to a reduction in the typical accretion rate, rather than a shift towards activity in lower-mass black holes. The eFEDS hard sample provides a preview of what can be expected from the eRASS final survey in terms of data quality. This pilot survey indicates the power of eROSITA to shed new light on the demographics and evolution of AGN, and the potential for discovery of new and rare populations.

On the use open BIM procedures for acoustic problems of the buildings: the new IDS standard for design stage and verification stage

Recent European directives provide for the use of digital procedures based on Building Information Modeling (BIM) as a tool for the design and management of the entire life cycle of public buildings. This directive has been implemented by the various Member States. Even acoustic performance must now be evaluated with the use of digital procedures based on BIM. The purpose of this work is to analyze the standards currently available on the market for the Desing phase, the verification phase and the acoustic classification of the buildings and evaluate the applicability through the use of the Open BIM models is the format and specifications based on IDS format. An Information Delivery Specification (IdS) is a computer interpretable document that defines the Exchange Requirements of model based exchange. This can be a combination of Industry Foundation Classes (IFC), Domain Extensions, and additional classifications and properties (national agreements or company specific ones; either stored in bSDD or somewhere else).

A Method for Conveying Confidence in iNaturalist Observations: A Case Study Using Non-Native Marine Species.

Concerns and limitations relating to data quality, reliability and accuracy hamper the use of citizen science initiatives in research and conservation. Valued for their cost-effective and large data acquisition potential, citizen science platforms such as iNaturalist have been highlighted as beneficial tools to supplement monitoring using traditional data sources. However, intrinsic uncertainties in unverified observations stem from the nature of species being identified, the quality of uploaded media and georeferencing; these factors can limit the value of the data as they can result in inaccurate records. Verification of data prior to use is critical. This process can, however, be laborious and time-consuming, with bias associated with the individual responsible for the task. To address this challenge this study developed a protocol for assigning confidence in iNaturalist observations, using marine alien and cryptogenic species observations from South Africa as a case study. A positive relationship was found between the accuracy of observations and confidence score. The inherent data quality assessment in iNaturalist, termed quality grade, was found to be an inadequate proxy for accuracy. The results of this study highlight the importance of the expert verification phase when using citizen science data. The confidence score facilitates a streamlined approach to the verification process by reducing the time taken to validate records, while assessing the three levels of uncertainty within observations and reducing researcher bias. It is recommended that this confidence score be used as an essential tool when using citizen science derived data.

Generating Java code pairing with ChatGPT

The Large Language Models (LLMs) like ChatGPT 3.5 have created a new era of automatic code generation. However, the existing research primarily focuses on generating simple code based on datasets (such as HumanEval, etc.). Most of approaches pay less attention to complex and practical code generation. Therefore, in this paper, we propose a new approach called “Xd-CodeGen” which can be used to generate large scale Java code. This approach is composed of four phases: requirement analysis, modeling, code generation, and code verification. In the requirement analysis phase, ChatGPT 3.5 is utilized to decompose and restate user requirements. To do so, a knowledge graph is developed to describe entities and their relationship in detail. Further, Propositional Projection Temporal Logic (PPTL) formulas are employed to define the properties of requirements. In the modeling phase, we use knowledge graphs to enhance prompts and generate UML class and activity diagrams for each sub-requirement using ChatGPT 3.5. In the code generation phase, based on established UML models, we make use of prompt engineering and knowledge graph to generate Java code. In the code verification phase, a runtime verification at code level approach is employed to verify generated Java code. Finally, we apply the proposed approach to develop a practical Java web project.

Effective design of sustainable energy productivity based on the experimental investigation of the humidification-dehumidification-desalination system using hybrid optimization

Reliable and cost-effective industrial-based sustainable desalination technologies are crucial for efficient and effective desalination of saline water. Seawater desalination emerges as a vital treatment, with humidification-dehumidification identified as a promising technology due to its simple, low-maintenance design and compatibility with renewable energy sources. The present work was developed using a large-scale experimental test rig based on the performance of a humidification-dehumidification system operated by the heat pump. As such, the study presents an advanced approach for optimizing humidification-dehumidification desalination systems using hybrid machine learning optimization techniques. The effectiveness of a neuro-fuzzy model, a decision algorithm based on a boosted tree, and a simple averaging ensemble in enhancing the performance of humidification-dehumidification systems were investigated. Experimental data includes cold water flowrate (L/min), hot water flowrate (L/min), inlet air temperature (°C), inlet cold water temperature (°C), inlet hot water temperature (°C), and inlet air relative humidity (%) as input variables combination and freshwater productivity (L/h), recovery ratio (%) as target variables. Several statistical indicators were used to evaluate the models’ prediction skills; similarly, a 2-dimensional Taylor diagram and error-radial plot were also utilized. In the calibration phase, boosted tree models slightly outperform neuro-fuzzy models, particularly for recovery ratio, demonstrating high accuracy and low bias. However, all models exhibit robust predictive accuracy in the verification phase, especially for recovery ratio, emphasizing their potential in generalizing to unseen data. The result indicated that the boosted tree outperformed others with Nash–Sutcliffe Efficiency = 94 % and 88 % for recovery ratio and freshwater productivity, respectively. The simple averaging ensemble shows marginal improvement with 95 % accuracy for the recovery ratio. Integrating hybrid artificial intelligence optimization with humidification-dehumidification desalination systems marks a transformative step in addressing freshwater shortages. By utilizing cutting-edge machine learning techniques, we achieve a more efficient, accurate, and reliable prediction of desalination performance.

APPS: Authentication-enabled privacy protection scheme for secure data transfer in Internet of Things

The Internet of Things (IoT) is an emerging field that encompasses several heterogeneous devices and smart objects that are integrated with the network. In open platforms, these objects are deployed to present advanced services in numerous applications. Innumerable security-sensitive data is generated by the IoT device and therefore, the security of these devices is an important task. This work formulates a secure data transfer technique in IoT, named Authentication enabled Privacy Protection (APPS) scheme for resource-constrained IoT devices. The proposed scheme demonstrates resilience against various attacks; such as resisting reply attacks, device anonymity, untracebility, session key establishment, quantum attacks and resisting MITM attacks. For the privacy protection scheme, the secured data transfer is initiated between the entities, like IoT devices, servers, and registration centers, by using various phases, namely registration phase, key generation phase, data encryption, authentication, verification, and data retrieval phase. Here, a mathematical model is designed for protecting data privacy using hashing, encryption, secret keys, etc. Finally, performance of proposed APPS model is analyzed; wherein the outcomes reveal that the proposed APPS model attained the maximum detection rate of 0.85, minimal memory usage of 0.497MB, and minimal computational time of 112. 79 sec and minimal turnaround time 131.91 sec.

Development of a Fit-For-Purpose Multi-Marker Panel for Early Diagnosis of Pancreatic Ductal Adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) suffers from a lack of an effective diagnostic method, which hampers improvement in patient survival. Carbohydrate antigen 19-9 (CA19-9) is the only FDA-approved blood biomarker for PDAC, yet its clinical utility is limited due to suboptimal performance. Liquid chromatography-mass spectrometry (LC-MS) has emerged as a burgeoning technology in clinical proteomics for the discovery, verification, and validation of novel biomarkers. A plethora of protein biomarker candidates for PDAC have been identified using LC-MS, yet few has successfully transitioned into clinical practice. This translational standstill is owed partly to insufficient considerations of practical needs and perspectives of clinical implementation during biomarker development pipelines, such as demonstrating the analytical robustness of proposed biomarkers which is critical for transitioning from research-grade to clinical-grade assays. Moreover, the throughput and cost-effectiveness of proposed assays ought to be considered concomitantly from the early phases of the biomarker pipelines for enhancing widespread adoption in clinical settings. Here, we developed a fit-for-purpose multi-marker panel for PDAC diagnosis by consolidating analytically robust biomarkers as well as employing a relatively simple LC-MS protocol. In the discovery phase, we comprehensively surveyed putative PDAC biomarkers from both in-house data and prior studies. In the verification phase, we developed a multiple-reaction monitoring (MRM)-MS-based proteomic assay using surrogate peptides that passed stringent analytical validation tests. We adopted a high-throughput protocol including a short gradient (<10min) and simple sample preparation (no depletion or enrichment steps). Additionally, we developed our assay using serum samples, which are usually the preferred biospecimen in clinical settings. We developed predictive models based on our final panel of 12 protein biomarkers combined with CA19-9, which showed improved diagnostic performance compared to using CA19-9 alone in discriminating PDAC from non-PDAC controls including healthy individuals and patients with benign pancreatic diseases. A large-scale clinical validation is underway to demonstrate the clinical validity of our novel panel.

Unambiguous and High-Fidelity Backdoor Watermarking for Deep Neural Networks.

The unprecedented success of deep learning could not be achieved without the synergy of big data, computing power, and human knowledge, among which none is free. This calls for the copyright protection of deep neural networks (DNNs), which has been tackled via DNN watermarking. Due to the special structure of DNNs, backdoor watermarks have been one of the popular solutions. In this article, we first present a big picture of DNN watermarking scenarios with rigorous definitions unifying the black- and white-box concepts across watermark embedding, attack, and verification phases. Then, from the perspective of data diversity, especially adversarial and open set examples overlooked in the existing works, we rigorously reveal the vulnerability of backdoor watermarks against black-box ambiguity attacks. To solve this problem, we propose an unambiguous backdoor watermarking scheme via the design of deterministically dependent trigger samples and labels, showing that the cost of ambiguity attacks will increase from the existing linear complexity to exponential complexity. Furthermore, noting that the existing definition of backdoor fidelity is solely concerned with classification accuracy, we propose to more rigorously evaluate fidelity via examining training data feature distributions and decision boundaries before and after backdoor embedding. Incorporating the proposed prototype guided regularizer (PGR) and fine-tune all layers (FTAL) strategy, we show that backdoor fidelity can be substantially improved. Experimental results using two versions of the basic ResNet18, advanced wide residual network (WRN28_10) and EfficientNet-B0, on MNIST, CIFAR-10, CIFAR-100, and FOOD-101 classification tasks, respectively, illustrate the advantages of the proposed method.

Privacy-preserving fair outsourcing polynomial computation without FHE and FPR

With the development of cloud computing, verifiable outsourcing computation (VC) has received much more attention. The polynomial is a fundamental mathematical function with widespread applications. Plenty of VC schemes for polynomials have been proposed recently. However, most previous schemes focus on ensuring that the client can get a valid result returned by the cloud service provide (CSP) before payment, while often ignoring the CSP’s interest. To the best of our knowledge, Guan et al. (2021) proposed a pioneering framework for building fair outsourcing polynomial computation, which serves as the state of the art. However, it discloses the privacy of outsourced polynomials, inputs, and outputs. Furthermore, it suffers from a false positive rate (FPR) in the verification phase due to the sampling technique. As a result, it breaks the fairness between the client and the CSP.To solve these problems, we propose a privacy-preserving fair outsourcing polynomial computation without FPR. To avoid expensive Fully Homomorphic Encryption (FHE), we utilize Paillier encryption and blind technique to ensure privacy. Our proposed scheme can guarantee fairness with an overwhelming probability by applying the SGX technique. The comprehensive performance evaluation and extensive simulations show that our protocol is more practical in cloud computing.

A Mendelian randomization analysis reveals the multifaceted role of the skin microbiota in liver cancer.

Hepatocellular carcinoma (HCC, or hepatic cancer, HC) and cholangiocarcinoma (CCA, or hepatic bile duct cancer, HBDC) are two major types of primary liver cancer (PLC). Previous studies have suggested that microbiota can either act as risk factors or preventive factors in PLC. However, no study has reported the relationship between skin microbiota and PLC. Therefore, we conducted a two-sample Mendelian randomization (MR) study to assess the causality between skin microbiota and PLC. Data from the genome-wide association study (GWAS) on skin microbiota were collected. The GWAS summary data of GCST90018803 (HBDC) and GCST90018858 (HC) were utilized in the discovery and verification phases, respectively. The inverse variance weighted (IVW) method was utilized as the principal method in our MR study. The MR-Egger intercept test, Cochran's Q-test, MR-Pleiotropy RESidual Sum and Outlier (MR-PRESSO), and leave-one-out analysis were conducted to identify the heterogeneity and pleiotropy. The results showed that Veillonella (unc.) plays a protective role in HBDC, while the family Neisseriaceae has a positive association with HBDC risk. The class Betaproteobacteria, Veillonella (unc.), and the phylum Bacillota (Firmicutes) play a protective role in HC. Staphylococcus epidermidis, Corynebacterium (unc.), the family Neisseriaceae, and Pasteurellaceae sp. were associated with an increased risk of HC. This study provided new evidence regarding the association between skin microbiota and PLC, suggesting that skin microbiota plays a role in PLC progression. Skin microbiota could be a novel and effective way for PLC diagnosis and treatment.