Experimental Evaluation Research Articles

Evaluation of lower detection limit and performance analyses related to the incident angle of X-rays and absolute dose using a triple-type dosimeter

When using a wearable active-type dosimeter for real-time dosimetry during X-ray fluoroscopic procedures, medical staff should be aware of the potential for an inaccurate exposure dose due to the large angular dependence. To address this issue, our research group has proposed a triple-type active dosimeter. With this device, the incident angle of scattered X-rays can be determined based on the signal difference between two dosimeters placed side-by-side, in which the wedge filters on the entrance windows are arranged symmetrically. In addition, the absolute dose can be determined from the central dosimeter with angular dependence correction. Although the concept of using wedge filters was proven in a previous study, the wedge filter materials have not been sufficiently studied. In this study, therefore, we sought to determine which of several materials was best suited as a wedge filter material for medical applications: aluminum (Al), titanium (Ti), or copper (Cu). We analyzed the performance of dosimeters with the different wedge filter materials in terms of estimation accuracy and lower detection limit. Using a clinical fluoroscopic system with a tube voltage of 80 kV, the scattered X-rays generated from an elliptical cylinder water phantom were measured while varying the incident direction of the X-rays from −50° to +50° in 10-degree intervals. The experimental results showed that the incident angles were properly determined when using the Ti filter, but the estimation accuracy was slightly lower when using the Al and Cu filters. The accuracy of absolute dose was evaluated from the standard deviation of the absolute dose data between −50° and +50°. As a result, the accuracies when using the Al, Ti, and Cu filters were 3.6%, 3.9%, and 5.9%, respectively. The lower detection limit when using the Cu filter showed a significantly higher value (0.50 μSv/s), but that with the Ti filter was reasonable (0.07 μSv/s). Based on these experimental evaluations, we concluded that the Ti filter would be suitable for use in a triple-type dosimeter. Finally, we investigated the expected dose rates of scattered X-rays in various interventional radiology (IVR) procedures described in the literature and determined that our dosimeter could be used for actual dose measurement.

Label-free OIRD detection of protein microarrays on high dielectric constant substrate with enhanced intrinsic sensitivity

Oblique-incidence reflectivity difference (OIRD) is a dielectric constant-sensitive technique and exhibits intriguing applications in label-free and high-throughput detection of protein microarrays. With the outstanding advantage of being compatible with arbitrary substrates, however, the effect of the substrate, particularly its dielectric constant on the OIRD sensitivity has not been fully disclosed. In this paper, for the first time we investigated the dependence of OIRD sensitivity on the dielectric constant of the substrate under top-incident OIRD configuration by combining theoretical modeling and experimental evaluation. Optical modeling suggested that the higher dielectric constant substrate exhibits a higher intrinsic sensitivity. Experimentally, three substrates including glass, fluorine-doped tin oxide (FTO) and silicon (Si) with different dielectric constants were selected as microarray substrates and their detection performances were evaluated. In good agreement with the modeling, high dielectric constant Si-based microarray exhibited the highest sensitivity among three chips, reaching a detection limit of as low as 5 ng mL−1 with streptavidin as the model target. Quantification of captured targets on three chips with on-chip enzyme-linked immunosorbent assay (ELISA) further confirmed that the enhanced performance originates from the high dielectric constant enhanced intrinsic OIRD sensitivity. This work thus provides a new way to OIRD-based label-free microarrays with improved sensitivity.

Performance evaluation of small-scale modified MR damper under cyclic loading: experimental evaluation

Performance evaluation of small-scale modified MR damper under cyclic loading: experimental evaluation

Alkali-Ion-Assisted Activation of ε-VOPO4 as a Cathode Material for Mg-Ion Batteries.

Rechargeable multivalent-ion batteries are attractive alternatives to Li-ion batteries to mitigate their issues with metal resources and metal anodes. However, many challenges remain before they can be practically used due to the low solid-state mobility of multivalent ions. In this study, a promising material identified by high-throughput computational screening is investigated, ε-VOPO4, as a Mg cathode. The experimental and computational evaluation of ε-VOPO4 suggests that it may provide an energy density of >200Whkg-1 based on the average voltage of a complete cycle, significantly more than that of well-known Chevrel compounds. Furthermore, this study finds that Mg-ion diffusion can be enhanced by co-intercalation of Li or Na, pointing at interesting correlation dynamics of slow and fast ions.

Maize leaf disease image enhancement algorithm using TFEGAN

Maize leaf diseases have a profound impact on maize crop yield. However, due to practical constraints, gathering a substantial number of maize disease images is time-consuming and labor-intensive. Furthermore, low-light conditions and noise interference in maize leaf disease images present significant challenges for image generation. Therefore, this paper proposes TFEGAN, a novel algorithm designed to generate maize leaf disease images. TFEGAN leverages a Two-Way Extraction Module (TWEM) to enhance the perceptual domain of the model and effectively reuse maize leaf disease features. The algorithm employs two well-trained Generative Adversarial Networks (GANs) to enhance low-light and noisy maize leaf disease images. The low-light enhancement component utilizes a double discriminator and self-regularization to balance global and local enhancement while constraining the distance between the input image and the denoising component. Moreover, the denoising part employs a strong coordinated attention mechanism and an environmental encoder to learn noise regions and their surrounding structures, thus generating better local images and enabling more accurate evaluation by the discriminant network. The generators of both networks are encapsulated by two discriminators and combined with a bi-directional feature extraction structure to generate normal maize leaf disease images. Experimental evaluations demonstrate the effectiveness of TFEGAN in improving the quality of maize leaf disease images under low-light and noisy conditions, showcasing its potential for practical applications. Overall, this paper presents a promising solution to the challenges of low-light and noise interference in maize leaf disease images, utilizing bidirectional feature extraction and GANs.

Experimental evaluation of Adsorption Heat for Water vapour/Silica gel pair for chiller application

Isosteric adsorption heat is crucial in assessing adsorption cycle performance. Typically, adsorption heat is indirectly estimated from the isotherm equation fitted to the experimental uptake data. This indirect method is susceptible to the accuracy and choice of isotherm equation. Consistency of this indirect method with actual adsorption heat measurements is rarely examined. Present work addresses this research gap by simultaneously measuring adsorption uptake and heat. This is achieved by incorporating heat flux sensors within an existing constant volume variable pressure (CVVP) setup. The measurements are carried out for water vapour/RD-silica gel pair with temperature and pressure ranging from 298 to 363 K and 0.15 kPa to 6.3 kPa respectively. Two popular isotherms viz. Dubinin–Astakhov (D–A) and Tóth isotherm equation are used to estimate the indirect adsorption heat. Both isotherm equations show a good fit with the experimental uptake (< 5% deviation). Interestingly, indirect adsorption heat estimated using Tóth isotherm fails to predict (∼15% deviation) the uptake dependent adsorption heat trend shown by calorimetric data, whereas D–A isotherm agrees well (<5% deviation). Therefore, the present work forms fundamental basis for identifying suitable isotherm equation, which is consistent with experimental adsorption uptake as well as heat data. Subsequently, a comparison of theoretical coefficient of performance (COP) for water vapour/RD-silica gel adsorption chiller is undertaken, based on indirect adsorption heat values available in the literature. The comparison reveals that COP prediction is sensitive to adsorption heat. This further emphasizes the necessity for experimentally measuring adsorption heat.

Consensus-Based Information Filtering in Distributed LiDAR Sensor Network for Tracking Mobile Robots.

A distributed state observer is designed for state estimation and tracking of mobile robots amidst dynamic environments and occlusions within distributed LiDAR sensor networks. The proposed novel framework enhances three-dimensional bounding box detection and tracking utilizing a consensus-based information filter and a region of interest for state estimation of mobile robots. The framework enables the identification of the input to the dynamic process using remote sensing, enhancing the state prediction accuracy for low-visibility and occlusion scenarios in dynamic scenes. Experimental evaluations in indoor settings confirm the effectiveness of the framework in terms of accuracy and computational efficiency. These results highlight the benefit of integrating stationary LiDAR sensors' state estimates into a switching consensus information filter to enhance the reliability of tracking and to reduce estimation error in the sense of mean square and covariance.

Exploring Flexural Strength Variation in Polymeric Materials for Provisional Fixed Prosthetic Structures: Comparative Analysis with and without Reinforcement through Laboratory Experimentation and Statistical Evaluation

Provisional fixed partial dentures represent a critical phase in dental treatment, necessitating heightened mechanical durability, particularly in comprehensive and extended treatment plans. Strengthening these structures with various reinforcing materials offers a method to enhance their resilience. Utilizing a three-point testing methodology on standardized trial specimens allows for a comparative assessment of various materials and reinforcement techniques for pre-prosthetic applications. This study aims to validate and assess the significance of integrating different reinforcing materials into standardized test bodies. The study focuses on test specimens comprising three types of unreinforced laboratory and clinical polymers for provisional constructions (n = 6)—heat-cured PMMA (Superpont C+B, Spofa Dental, Czech Republic), CAD-CAM prefabricated PMMA (DD temp MED, Dental Direkt, Germany), CAD-CAM printing resin (Temporary CB Resin, FormLabs, USA), self-polymerizing PEMA (DENTALON plus, Kulzer, Germany), light-polymerizing composite (Revotek LC, GC, Japan), and dual-polymerizing composite (TempSpan, Pentron, USA). Additionally, laboratory polymers are evaluated in groups with five types of reinforcing filaments (n = 15)—Glass Fiber (Fiber Splint One-Layer, Polydentia, Switzerland), Polyethylene thread (Ribbond Regular 4.0 mm, Ribbond Inc., USA), triple-stranded chrome-cobalt wire for splinting 015″ (Leone S.p.a., Italy), Aesthetic ligature wire 012″ (Leone S.p.a., Italy), and Glass Fiber coated with light-cured composite 8.5 × 0.2 mm (Interlig, Angelus, Brazil). Analysis of the data using Generalized Linear Models (GLMs) reveals that the experimental bodies, produced via the subtractive digital method using PMMA (DD temp MED, Dental Direkt GmbH, Germany) as the polymer and glass filaments as the reinforcement, exhibit superior mechanical properties, particularly when pre-wetted with Interlig liquid composite (Angelus, Brazil).

Production, characterization, and application of Pseudoxanthomonas taiwanensis biosurfactant: a green chemical for microbial enhanced oil recovery (MEOR)

Biosurfactants, as microbial bioproducts, have significant potential in the field of microbial enhanced oil recovery (MEOR). Biosurfactants are microbial bioproducts with the potential to reduce the interfacial tension (IFT) between crude oil and water, thus enhancing oil recovery. This study aims to investigate the production and characterization of biosurfactants and evaluate their effectiveness in increasing oil recovery. Pseudoxanthomonas taiwanensis was cultured on SMSS medium to produce biosurfactants. Crude oil was found to be the most effective carbon source for biosurfactant production. The biosurfactants exhibited comparable activity to sodium dodecyl sulfate (SDS) at a concentration of 400 ppm in reducing IFT. It was characterized as glycolipids, showing stability in emulsions at high temperatures (up to 120 °C), pH levels ranging from 3 to 9, and NaCl concentrations up to 10% (w/v). Response surface methodology revealed the optimized conditions for the most stable biosurfactants (pH 7, temperature of 40 °C, and salinity of 2%), resulting in an EI24 value of 64.45%. Experimental evaluations included sand pack column and core flooding studies, which demonstrated additional oil recovery of 36.04% and 12.92%, respectively. These results indicate the potential application of P. taiwanensis biosurfactants as sustainable and environmentally friendly approaches to enhance oil recovery in MEOR processes.

Few-Shot Federated Learning: A Federated Learning Model for Small-Sample Scenarios

Traditional federated learning relies heavily on mature datasets, which typically consist of large volumes of uniformly distributed data. While acquiring extensive datasets is relatively straightforward in academic research, it becomes prohibitively expensive in practical applications, especially in emerging or specialized medical fields characterized by data scarcity. This poses a significant challenge. To address this issue, our study introduces a federated learning model that integrates few-shot learning techniques and is complemented by personalized knowledge distillation to further enhance the model’s classification accuracy. This innovative approach significantly reduces the dependence on large-scale datasets, enabling efficient model training under limited data conditions. Our experimental evaluations conducted on small-scale datasets, including Omniglot, FC100, and mini-ImageNet, indicate that our model surpasses existing state-of-the-art federated learning models in terms of accuracy, achieving a substantial improvement. Specifically, on the FC100 dataset, the classification accuracy of the conventional federated learning algorithm FedAvg was merely 19.6%, whereas the method proposed in this study achieved a classification accuracy of 41%, representing an improvement of more than double. This advancement not only highlights our model’s superiority in alleviating the challenges of limited data availability, but also expands the applicability of federated learning to a broader range of applications.

Innovative Financial Management in Higher Education: A Multi-Scale Deep Learning Approach for Risk Reduction and Quality Enhancement

This paper presents a novel university financial management system leveraging multi-scale deep learning. With rising college enrollment and teaching complexities, traditional financial models require adaptation to mitigate risks and improve management quality. The system integrates hardware and software innovations: multiple sensors enhance data scanning, coordinated by a central coordinator, ensuring comprehensive financial database coverage. Software-wise, a structured database establishes attribute-based financial connections, crucial for weight assignment. Employing a multilayer perceptual network topology, a full interconnection model based on multi-scale deep learning facilitates profound data extraction. Experimental evaluations demonstrate the system's superior financial risk assessment capabilities compared to traditional approaches, extracting a broader spectrum of financial parameters for comprehensive risk warnings. By embracing multi-scale deep learning, this system promises significant advancements in university financial management, enhancing adaptability and risk mitigation in college finance departments.

Multispectral Pedestrian Detection Based on Prior-Saliency Attention and Image Fusion

Detecting pedestrians in varying illumination conditions poses a significant challenge, necessitating the development of innovative solutions. In response to this, we introduce Prior-AttentionNet, a pedestrian detection model featuring a Prior-Attention mechanism. This model leverages the stark contrast between thermal objects and their backgrounds in far-infrared (FIR) images by employing saliency attention derived from FIR images via UNet. However, extracting salient regions of diverse scales from FIR images poses a challenge for saliency attention. To address this, we integrate Simple Linear Iterative Clustering (SLIC) superpixel segmentation, embedding the segmentation feature map as prior knowledge into UNet’s decoding stage for comprehensive end-to-end training and detection. This integration enhances the extraction of focused attention regions, with the synergy of segmentation prior and saliency attention forming the core of Prior-AttentionNet. Moreover, to enrich pedestrian details and contour visibility in low-light conditions, we implement multispectral image fusion. Experimental evaluations were conducted on the KAIST and OTCBVS datasets. Applying Prior-Attention mode to FIR-RGB images significantly improves the delineation and focus on multi-scale pedestrians. Prior-AttentionNet’s general detector demonstrates the capability of detecting pedestrians with minimal computational resources. The ablation studies indicate that the FIR-RGB+ Prior-Attention mode markedly enhances detection robustness over other modes. When compared to conventional multispectral pedestrian detection models, Prior-AttentionNet consistently surpasses them by achieving higher mean average precision and lower miss rates in diverse scenarios, during both day and night.

Density-based clustering with boundary samples verification

Density-based clustering is a widely explored domain in the field of machine learning, with numerous methods proposed to address challenges in clustering under varying density conditions. These methods primarily categorize samples based on their local density. However, challenges arise when dealing with boundary samples. The density of certain boundary samples is typically lower, which can lead to misclassification as noise. Moreover, in situations where two clusters have similar densities and are in close proximity, accurately classifying the boundary points between them becomes a challenging task. In this paper, we introduce an enhanced approach based on k-nearest neighbors to address this challenge within density-based clustering. In our method, upon the formation of a new cluster, we identify boundary samples by examining the spatial relationships between each sample and its k-nearest neighbors, as well as their connections to the newly established cluster. Once all clusters are formed, we refine the classification of these boundary samples by adjusting or retaining their assigned labels based on their k-nearest neighbors. Experimental evaluations on synthetic and real-world datasets demonstrate the effectiveness of our proposed method.

Bifonazole caffeate: The first molecular salt of bifonazole with enhanced biopharmaceutical property based on experiments and quantum chemistry research

In order to make novel breakthroughs in molecular salt studies of BCS class-IV antifungal medication bifonazole (BIF), a salification-driven strategy towards ameliorating attributes and aiding augment efficiency is raised. This strategy fully harnesses structural characters together attributes and benefits of caffeic acid (CAF) to concurrently enhance dissolvability and permeability of BIF by introducing the two ingredients into the identical molecular salt lattice through the salification reaction, which, coupled with the aroused potential activity of CAF significantly amplifies the antifungal efficacy of BIF. Guided by this route, the first BIF-organic molecular salt, BIF-CAF, is directionally designed and synthesized with satisfactorily structural characterizations and integrated theoretical and experimental explorations on the pharmaceutical properties. Single-crystal X-ray diffraction resolving confirms that there is a lipid-water amphiphilic sandwich structure constructed by robust charge-assistant hydrogen bonds in the salt crystal, endowing the molecular salt with the potential to enhance both dissolvability and permeability relative to the parent drug, which is validated by experimental evaluations. Remarkably, the comprehensive DFT-based theoretical investigations covering frontier molecular orbital, molecular electrostatic potential, Hirshfeld surface analysis, reduced density gradient, topology, sphericity and planarity analysis strongly support these observations, thereby allowing some positive relationships between macroscopic properties and microstructures of the molecular salt can be made. Intriguingly, the optimal properties, together with the stimulated activity of CAF markedly augment in vitro antifungal ability of the molecular salt, with magnifying inhibition zones and reducing minimum inhibitory concentrations. These findings fill in the gaps on researches of BIF-organic molecular salt, and adequately exemplify the feasibility and validity by integrating theoretical and experimental approaches to resolve BIF’s problems via the salification-driven tactic.

Variate Associated Domain Adaptation for Unsupervised Multivariate Time Series Anomaly Detection

Multivariate Time Series Anomaly Detection (MTS-AD) is crucial for the effective management and maintenance of devices in complex systems such as server clusters, spacecrafts and financial systems etc. However, upgrade or cross-platform deployment of these devices will introduce the issue of cross-domain distribution shift, which leads to the prototypical problem of Domain Adaptation for MTS-AD. Compared with general domain adaptation problems, MTS-AD domain adaptation presents two peculiar challenges: 1) The dimensions of data from the source domain and the target domain are usually different, so alignment without losing any information is necessary. 2) The association between different variates plays a vital role in the MTS-AD task, which is overlooked by traditional domain adaptation approaches. Aiming at addressing the above issues, we propose a V ariate A ssociated domai N a D aptation method combined with a Gr A ph Deviation Network (abbreviated as VANDA ) for MTS-AD, which includes two major contributions. First, we characterize the intra-domain variate associations of the source domain by a graph deviation network (GDN), which can share parameters across domains without dimension alignment. Second, we propose a sliding similarity to measure the inter-domain variate associations and perform joint training by minimizing the optimal transport distance between source and target data for transferring variate associations across domains. VANDA achieves domain adaptation by transferring both variate associations and GDN parameters from the source domain to the target domain. We construct two pairs of MTS-AD datasets from existing MTS-AD data and combine three domain adaptation strategies with six MTS-AD backbones as the benchmark methods for experimental evaluation and comparison. Extensive experiments demonstrate the effectiveness of our approach, which outperforms the benchmark methods, and significantly improves the AD performance of the target domain by effectively utilizing the source domain knowledge.

Dual-branch network with fused Mel features for logic-manipulated speech detection

In recent years, logic-manipulated speech detection has predominantly relied on single-feature and single-channel networks, which fail to consider a broader range of acoustic features. To address this limitation, we propose a novel approach based on a dual-branch network that incorporates fused Mel features for detecting logic-manipulated speech. The structure of our network consists of three main components. Firstly, we extract Mel spectrograms and Mel-frequency cepstral coefficients (MFCC) to address the deficiencies caused by the reduction in the information dimension. Then, to better utilize these two types of features, we construct a dual-branch network framework that enables the separate learning of temporal and frequency domain characteristics. Finally, at the backend of the network, we fuse the feature vectors for decision-making. Experimental evaluations conducted on the ASVspoof2019LA dataset demonstrate the effectiveness and superiority of our proposed method compared to state-of-the-art models and other acoustic features.

Multi-piezoelectric energy harvesters array based on wind-induced vibration: Design, simulation, and experimental evaluation

The efficiency of a single piezoelectric energy harvester (PEH) is inadequate to meet the energy requirements of equipment, requiring the simultaneous operation of multiple PEHs. Consequently, this work conducts research on multi-piezoelectric energy harvesters (MPEH) array based on wind-induced vibration. Analyze the output characteristics of array configurations using simulation and experimental methods, including dual parallel array (PEH-2p), dual series array (PEH-2s), four-square array (PEH-4), and nine-square array (PEH-9) energy harvesters. The results indicate that the spacing ratio (L/D) is the primary factor influencing the output performance of the PEH array. Both of the starting and peak wind speeds of PEH-2p were below the reference value, but the power peak has increased. The starting wind speed of PEH-2s has been reduced, with a more significant improvement in the output performance of the downstream energy harvester compared to the upstream energy harvester. The output power of the downstream energy harvester has increased by 323.35 % in comparison to the reference value. Additionally, the PEH-4 demonstrates a boosting effect on the output power of both the upstream and downstream energy harvesters. The energy harvester in the center of the PEH-9 has the highest power peak, being 349.44 % higher than the reference value.

MSO‐DETR: Metric space optimization for few‐shot object detection

AbstractIn the metric‐based meta‐learning detection model, the distribution of training samples in the metric space has great influence on the detection performance, and this influence is usually ignored by traditional meta‐detectors. In addition, the design of metric space might be interfered with by the background noise of training samples. To tackle these issues, we propose a metric space optimisation method based on hyperbolic geometry attention and class‐agnostic activation maps. First, the geometric properties of hyperbolic spaces to establish a structured metric space are used. A variety of feature samples of different classes are embedded into the hyperbolic space with extremely low distortion. This metric space is more suitable for representing tree‐like structures between categories for image scene analysis. Meanwhile, a novel similarity measure function based on Poincaré distance is proposed to evaluate the distance of various types of objects in the feature space. In addition, the class‐agnostic activation maps (CCAMs) are employed to re‐calibrate the weight of foreground feature information and suppress background information. Finally, the decoder processes the high‐level feature information as the decoding of the query object and detects objects by predicting their locations and corresponding task encodings. Experimental evaluation is conducted on Pascal VOC and MS COCO datasets. The experiment results show that the effectiveness of the authors’ method surpasses the performance baseline of the excellent few‐shot detection models.

A Sentence-Embedding-Based Dashboard to Support Teacher Analysis of Learner Concept Maps

Concept mapping is a valuable method to represent a domain of knowledge, also with the aim of supporting educational needs. Students are called upon to construct their own knowledge through a meaningful learning process, linking new concepts to concepts they have already learned, i.e., connecting new knowledge to knowledge they already possess. Moreover, the particular graphic form of a concept map makes it easy for the teacher to construct and interpret both. Consequently, for an educator, the ability to assess concept maps offered by students, facilitated by an automated system, can prove invaluable. This becomes even more apparent in educational settings where there is a large number of students, such as in Massive Open Online Courses. Here, we propose two new measures devised to evaluate the similarity between concept maps based on two deep-learning embedding models: InferSent and Universal Sentence Encoder. An experimental evaluation with a sample of teachers confirms the validity of one such deep-learning model as the baseline of the new similarity measure. Subsequently, we present a proof-of-concept dashboard where the measures are used to encode a concept map in a 2D space point, with the aim of helping teachers monitor students’ concept-mapping activity.

A Fully Automated Online Enrichment and Separation System for Highly Reproducible and In-Depth Analysis of Intact Glycopeptide.

A fully automated online enrichment and separation system for intact glycopeptides, named AutoGP, was developed in this study by integrating three different columns in a nano-LC system. Specifically, the peptide mixture from the enzymatic digestion of a complex biological sample was first loaded on a hydrophilic interaction chromatography (HILIC) column. The nonglycopeptides in the sample were washed off the column, and the glycopeptides retained by the HILIC column were eluted to a C18 trap column to achieve an automated glycopeptide enrichment. The enriched glycopeptides were further eluted to a C18 column for separation, and the separated glycopeptides were eventually analyzed by using an orbitrap mass spectrometer (MS). The optimal operating conditions for AutoGP were systemically studied, and the performance of the fully optimized AutoGP was compared with a conventional manual system used for glycopeptide analysis. The experimental evaluation shows that the total number of glycopeptides identified is at least 1.5-fold higher, and the median coefficient of variation for the analyses is at least 50% lower by using AutoGP, as compared to the results acquired by using the manual system. In addition, AutoGP can perform effective analysis even with a 1-μg sample amount, while a 10-μg sample at least will be needed by the manual system, implying an order of magnitude better sensitivity of AutoGP. All the experimental results have consistently proven that AutoGP can be used for much better characterization of intact glycopeptides.