Search Results (4,920)

Breast cancer is a prevalent malignancy characterized by the uncontrolled growth of glandular epithelial cells, which can metastasize through the blood and lymphatic systems. Microcalcifications, small calcium deposits within breast tissue, are critical markers for early detection of breast cancer, especially in non-palpable carcinomas. These microcalcifications, appearing as small white spots on mammograms, are challenging to identify due to potential confusion with other tissues. This study hypothesizes that a hybrid feature extraction approach combined with Convolutional Neural Networks (CNNs) can significantly enhance the detection and localization of microcalcifications in mammograms. The proposed algorithm employs Gabor, Prewitt, and Gray Level Co-occurrence Matrix (GLCM) kernels for feature extraction. These features are input to a CNN architecture designed with maxpooling layers, Rectified Linear Unit (ReLU) activation functions, and a sigmoid response for binary classification. Additionally, the Top Hat filter is used for precise localization of microcalcifications. The preprocessing stage includes enhancing contrast using the Volume of Interest Look-Up Table (VOI LUT) technique and segmenting regions of interest. The CNN architecture comprises three convolutional layers, three ReLU layers, and three maxpooling layers. The training was conducted using a balanced dataset of digital mammograms, with the Adam optimizer and binary cross-entropy loss function. Our method achieved an accuracy of 89.56%, a sensitivity of 82.14%, and a specificity of 91.47%, outperforming related works, which typically report accuracies around 85–87% and sensitivities between 76 and 81%. These results underscore the potential of combining traditional feature extraction techniques with deep learning models to improve the detection and localization of microcalcifications. This system may serve as an auxiliary tool for radiologists, enhancing early detection capabilities and potentially reducing diagnostic errors in mass screening programs. Full article

The point clouds obtained directly from three-dimensional scanning devices are often sparse and noisy. Therefore, point cloud upsampling plays an increasingly crucial role in fields such as point cloud reconstruction and rendering. However, point cloud upsampling methods are primarily supervised and fixed-rate, which restricts their applicability in various scenarios. In this paper, we propose a novel point cloud upsampling method, named RE-PU, which is based on the point cloud reconstruction and achieves self-supervised upsampling at arbitrary rates. The proposed method consists of two main stages: the first stage is to train a network to reconstruct the original point cloud from a prior distribution, and the second stage is to upsample the point cloud data by increasing the number of sampled points on the prior distribution with the trained model. The experimental results demonstrate that the proposed method can achieve comparable outcomes to supervised methods in terms of both visual quality and quantitative metrics. Full article

The sustainable development of tourism in cultural heritage sites benefits from the active participation of tourists in the co-creation process. Location-based AR games show great potential in tourists’ participation in creation and positive experiences. This study explores the relationship between the stage factors of the co-creation experience and the overall co-creation. Combining the service-dominant logic and process perspective of value co-creation theory, this research proposes a conceptual framework for co-creating experiences in cultural heritage tourism using augmented reality technology through two studies. In the first phase of the study, quantitative research was conducted on 256 visitors to measure the impact of factors in the three processes of pre-co-creation experience, on-site experience, and post-co-creation experience on the overall co-creation experience. In the study’s second phase, follow-up qualitative interviews were conducted based on multiple linear regression analysis results to expand the interpretation of the relationship and importance of factors affecting the co-creation experience process. The results show that psychological engagement, awareness, knowledge, and social relations during pre-visitation help enhance the overall co-creation experience. In contrast, the overall co-creation experience is enriched by real-time storytelling, interaction, and emotional resonance in both on-site and post-experience processes. The proposal of this framework model advances the discussion of augmented reality technology and co-creation experience to the empirical level. It provides a basis for further tourism co-creation experience design practice. Full article

The adaptability of bermudagrass genotypes to high-pH saline–alkali conditions was investigated through a comprehensive evaluation of 38 genotypes during the seedling stage. For this purpose, two distinct treatments were established: exposure to saline–alkali solution composed of 45% NaCl, 5% Na₂SO₄, 5% NaHCO₃, and 45% Na₂CO₃ (pH 10.0), and exposure to distilled water as control. On 6th day of treatment, eight physiological indicators were measured. Compared with the control, the net photosynthetic rates, leaf water content, and chlorophyll content of the test genotypes decreased under stress. In contrast, the soluble protein content, proline levels, malondialdehyde concentration, and conductivity exhibited an increase. The salt–alkali tolerance coefficients of each indicator ranged from 0.24 to 8.54, and the variable coefficient was from 9.77% to 62.82%. Based on the salt–alkali tolerance coefficients, the comprehensive evaluation value (D) and resistance coefficient (CSAC) for each genotype were calculated. Subsequently, 38 genotypes were classified into three salt–alkali tolerance clusters by hierarchical clustering analysis, with Cluster I consisting of 10 genotypes with the most salt–alkali tolerance, and Cluster II with intermediate tolerance. Cluster III was comprised of 18 genotypes showing the lowest tolerance. The predictive model for assessing salt–alkali tolerance in bermudagrass is (D) = −0.238 + 0.106 × SAC_Chlb + 0.209 × SAC_RWC + 0.015 × SAC_Pro + 0.284 × SAC_Protein + 0.051 × SAC_Pn. Notably, Cluster I genotypes were more vigorous and showed lower damage under saline stress compared to Cluster III. Moreover, stepwise regression analysis pinpointed Chlb, RWC, and Pro as crucial indicators for evaluating salt–alkali tolerance in bermudagrass genotypes. Full article

Band selection (BS) aims to reduce redundancy in hyperspectral imagery (HSI). Existing BS approaches typically model HSI only in a single dimension, either spectral or spatial, without exploring the interactions between different dimensions. To this end, we propose an unsupervised BS method based on a spectral–spatial cross-dimensional attention network, named SSANet-BS. This network is comprised of three stages: a band attention module (BAM) that employs an attention mechanism to adaptively identify and select highly significant bands; two parallel spectral–spatial attention modules (SSAMs), which fuse complex spectral–spatial structural information across dimensions in HSI; a multi-scale reconstruction network that learns spectral–spatial nonlinear dependencies in the SSAM-fusion image at various scales and guides the BAM weights to automatically converge to the target bands via backpropagation. The three-stage structure of SSANet-BS enables the BAM weights to fully represent the saliency of the bands, thereby valuable bands are obtained automatically. Experimental results on four real hyperspectral datasets demonstrate the effectiveness of SSANet-BS. Full article

The primary objective of this paper focuses on developing a control approach to improve the operational performance of a three-level neutral point clamped (3LNPC) shunt active power filter (SAPF) within a grid-tied PV system configuration. Indeed, this developed control approach, based on the used 3LNPC-SAPF topology, aims to ensure the seamless integration of a photovoltaic system into the three-phase four-wire grid while effectively mitigating grid harmonics, grid current unbalance, ensuring grid unit power factor by compensating the load reactive power, and allowing power sharing with the grid in case of an excess of generated power from the PV system, leading to overall high power quality at the grid side. This developed approach is based initially on the application of the four-wire instantaneous p-q theory for the identification of the reference currents that have to be injected by the 3LNPC-SAPF in the grid point of common coupling (PCC). Whereas, the 3LNPC is controlled based on using the finite control set model predictive control (FCS-MPC), which can be accomplished by determining the convenient set of switch states leading to the voltage vector, which is the most suitable to ensure the minimization of the selected cost function. Furthermore, the used topology requires a constant DC-link voltage and balanced split-capacitor voltages at the input side of the 3LNPN. Hence, the cost function is adjusted by the addition of another term with a selected weighting factor related to these voltages to ensure their precise control following the required reference values. However, due to the random changes in solar irradiance and, furthermore, to ensure efficient operation of the proposed topology, the PV system is connected to the 3LNPN-SAPF via a DC/DC boost converter to ensure the stability of the 3LNPN input voltage within the reference value, which is achieved in this paper based on the use of the maximum power point tracking (MPPT) technique. For the validation of the proposed control technique and the functionality of the used topology, a set of simulations has been presented and investigated in this paper following different irradiance profile scenarios such as a constant irradiance profile and a variables irradiance profile where the main aim is to prove the effectiveness and flexibility of the proposed approach under variable irradiance conditions. The obtained results based on the simulations carried out in this study demonstrate that the proposed control approach with the used topology under different loads such as linear, non-linear, and unbalanced can effectively reduce the harmonics, eliminating the unbalance in the currents and compensating for the reactive component contained in the grid side. The obtained results prove also that the proposed control ensures a consistent flow of power based on the sharing principle between the grid and the PV system as well as enabling the efficient satisfaction of the load demand. It can be said that the proposal presented in this paper has been proven to have many dominant features such as the ability to accurately estimate the power sharing between the grid and the PV system for ensuring the harmonics elimination, the reactive power compensation, and the elimination of the neutral current based on the zero-sequence component compensation, even under variable irradiance conditions. This feature makes the used topology and the developed control a valuable tool for power quality improvement and grid stability enhancement with low cost and under clean energy. Full article

Creating an effective deep learning technique for accurately diagnosing leak signals across diverse environments is crucial for integrating artificial intelligence (AI) into the power plant industry. We propose an automatic weight redistribution ensemble model based on transfer learning (TL) for detecting leaks in diverse power plant environments, overcoming the challenges of site-specific AI methods. This innovative model processes time series acoustic data collected from multiple homogeneous sensors located at different positions into three-dimensional root-mean-square (RMS) and frequency volume features, enabling accurate leak detection. Utilizing a TL-driven, two-stage learning process, we first train residual-network-based models for each domain using these preprocessed features. Subsequently, these models are retrained in an ensemble for comprehensive leak detection across domains, with control weight ratios finely adjusted through a softmax score-based approach. The experiment results demonstrate that the proposed method effectively distinguishes low-level leaks and noise compared to existing techniques, even when the data available for model training are very limited. Full article

The development of a crop production strategy through the use of a crop model represents a crucial method for the assurance of a stable agricultural yield and the subsequent enhancement thereof. There are currently no studies evaluating the suitability of the AquaCrop model for the drip irrigation of Gossypium barbadense in Southern Xinjiang, which is the primary planting region for Gossypium barbadense in China. In order to investigate the performance of the AquaCrop model in simulating the growth of cotton under mulched drip irrigation, the model was locally calibrated and validated according to different irrigation thresholds during a key growth period of two years. The results of the simulation for total soil water (TSW), crop evapotranspiration (ET_c), canopy coverage (CC), aboveground biomass (Bio), and seed cotton yield demonstrated a high degree of correlation with the observed data, with a root mean square error (RMSE) of <11.58%. The Bio and yield simulations demonstrated a high degree of concordance with the corresponding measured values, with root mean square error (RMSE) values of 1.23 t ha⁻¹ and 0.15 t ha⁻¹, respectively. However, the predicted yield declined in the verification year, though the prediction error remained below 15%. Furthermore, the estimated evapotranspiration (ET_c) value demonstrated a slight degree of overestimation. Generally, the middle and late stages of cotton growth led to an overestimation of the TSW content. However, the prediction error was less than 13.99%. Through the calculation of each performance index of the AquaCrop model, it is found that they are in the acceptable range. In conclusion, the AquaCrop model can be employed as a viable tool for predicting the water response of cotton to drip irrigation under mulched film in Southern Xinjiang. Based on 64 years of historical meteorological data, three years were selected as scenarios for simulation. Principal component analysis (PCA) showed that, in a local wet year in Southern Xinjiang, the irrigation quota was 520 mm, and the irrigation cycle was 6 days/time. In normal years, the irrigation quota was 520 mm, with an irrigation cycle of 6 days/time. In dry years, the irrigation quota was 595 mm, with an irrigation cycle of 10 days/time. This allowed for higher seed cotton yields and irrigation water productivity, as well as the maximization of cotton yields and net revenue in the arid oasis area of Southern Xinjiang. Full article

A cost-effective solution to address the challenges posed by sensitive instrumentation in next-gen turbomachinery components is to reduce the number of measurement samples required to assess complex flows. This study investigates Gaussian Process (GP) modeling approaches within the framework of a data-driven hybrid measurement technique for turbomachinery applications. Three different modeling approaches—Baseline GP, CFD to Experiments GP, and Multi-Fidelity GP—are evaluated, and their performance in predicting mean flow characteristics and associated uncertainties on a low aspect ratio axial compressor stage, representative of the last stage of a high-pressure compressor, are focused on. The Baseline GP demonstrates robust accuracy, while the integration of CFD data in CFD into Experiments GP introduces complexities and more errors. The Multi-Fidelity GP, leveraging both CFD and experimental data, emerges as a promising solution, exhibiting enhanced accuracy in critical flow features. A sensitivity analysis underscores its stability and accuracy, even with reduced measurements. The Multi-Fidelity GP, therefore, stands as a reliable data fusion method for the proposed hybrid measurement technique, offering a potential reduction in instrumentation effort and testing times. Full article

Detecting early plant diseases autonomously poses a significant challenge for self-navigating robots and automated systems utilizing Artificial Intelligence (AI) imaging. For instance, Botrytis cinerea, also known as gray mold disease, is a major threat to agriculture, particularly impacting significant crops in the Cucurbitaceae and Solanaceae families, making early and accurate detection essential for effective disease management. This study focuses on the improvement of deep learning (DL) segmentation models capable of early detecting B. cinerea on Cucurbitaceae crops utilizing Vision Transformer (ViT) encoders, which have shown promising segmentation performance, in systemic use with the Cut-and-Paste method that further improves accuracy and efficiency addressing dataset imbalance. Furthermore, to enhance the robustness of AI models for early detection in real-world settings, an advanced imagery dataset was employed. The dataset consists of healthy and artificially inoculated cucumber plants with B. cinerea and captures the disease progression through multi-spectral imaging over the course of days, depicting the full spectrum of symptoms of the infection, ranging from early, non-visible stages to advanced disease manifestations. Research findings, based on a three-class system, identify the combination of U-Net++ with MobileViTV2-125 as the best-performing model. This model achieved a mean Dice Similarity Coefficient (mDSC) of 0.792, a mean Intersection over Union (mIoU) of 0.816, and a recall rate of 0.885, with a high accuracy of 92%. Analyzing the detection capabilities during the initial days post-inoculation demonstrates the ability to identify invisible B. cinerea infections as early as day 2 and increasing up to day 6, reaching an IoU of 67.1%. This study assesses various infection stages, distinguishing them from abiotic stress responses or physiological deterioration, which is crucial for accurate disease management as it separates pathogenic from non-pathogenic stress factors. The findings of this study indicate a significant advancement in agricultural disease monitoring and control, with the potential for adoption in on-site digital systems (robots, mobile apps, etc.) operating in real settings, showcasing the effectiveness of ViT-based DL segmentation models for prompt and precise botrytis detection. Full article

Deepfake has become an emerging technology affecting cyber-security with its illegal applications in recent years. Most deepfake detectors utilize CNN-based models such as the Xception Network to distinguish real or fake media; however, their performance on cross-datasets is not ideal because they suffer from over-fitting in the current stage. Therefore, this paper proposed a spatial consistency learning method to relieve this issue in three aspects. Firstly, we increased the selections of data augmentation methods to 5, which is more than our previous study’s data augmentation methods. Specifically, we captured several equal video frames of one video and randomly selected five different data augmentations to obtain different data views to enrich the input variety. Secondly, we chose Swin Transformer as the feature extractor instead of a CNN-based backbone, which means that our approach did not utilize it for downstream tasks, and could encode these data using an end-to-end Swin Transformer, aiming to learn the correlation between different image patches. Finally, this was combined with consistency learning in our study, and consistency learning was able to determine more data relationships than supervised classification. We explored the consistency of video frames’ features by calculating their cosine distance and applied traditional cross-entropy loss to regulate this classification loss. Extensive in-dataset and cross-dataset experiments demonstrated that Swin-Fake could produce relatively good results on some open-source deepfake datasets, including FaceForensics++, DFDC, Celeb-DF and FaceShifter. By comparing our model with several benchmark models, our approach shows relatively strong robustness in detecting deepfake media. Full article

In this study, finite element (FE) simulation by the software Abaqus was relied on to investigate the roll forming process of a wheel rim made of an innovative dual-phase steel, i.e., DP590, after flash butt welding (FBW). In the simulation, an FE model was generated, including the design of the dies for flaring, three-roll forming, and expansion, and detailed key processing parameters based on practical production of the selected DP590. Combined with the microstructures and properties of the weld zone (WZ) and heat-affected zones (HAZs) after FBW, the distribution of stress/strain and the change in thickness of the base metal (BM), WZ and HAZs were analyzed, and compared in the important stages of roll forming. Theoretically, the variation in the microstructure and the corresponding stress–strain behaviors of the BM, WZ, and HAZs after FBW have led to the thickness reduction of DP590 that originated from softening behaviors occurring at the region of subcritical HAZs (SCHAZs), and a small amount of tempered martensite has evidently reduced the hardness and strength of the SCHAZ. Meanwhile, the distribution of stress/strain has been influenced to some extent. Further, the study includes the influence of the friction coefficient on the forming quality of the wheel rim to guarantee the simulation accuracy in practical applications. In sum, the dual-phase steel has to be carefully applied to the wheel rim, which needs to experience the processes of FBW and roll forming, focusing on the performance of SCHAZs. Full article

Background: anatomic lung segmentectomies allow accurate resection of pulmonary lesions, maximizing healthy tissue preservation, and reducing unnecessary loss of lung function. In this setting, accurate preoperative planning is crucial. We present our early experience, detailing the successful use of 3D-CT models in tailoring therapeutic strategies for three patients undergoing complex anatomical lung resections due to neoplastic diseases. Case Presentation: (1) 60-year-old male patient with significant pulmonary functional impairment underwent successful right lower lobe bi-segmentectomy (S7–S8) for carcinoid, stage IA1. (2) 65-year-old female patient with previous left lung resection and functional impairment underwent uneventful right upper lobe bi-segmentectomy (S1–S2) for double lung adenocarcinoma, stage IIb. (3) 67-year-old male with previous ipsilateral lung resection underwent left lower lobe segmentectomy (S8) for metastatic colic adenocarcinoma without any complications. Conclusion: 3D-CT imaging, particularly through VP^TM platform, enhances the safety and precision of complex lung segmentectomy, providing a valuable surgical map for improved outcomes. Full article

PURPOSE: We aim to compare the performance of three different radiomics models (logistic regression (LR), random forest (RF), and support vector machine (SVM)) and clinical nomograms (Briganti, MSKCC, Yale, and Roach) for predicting lymph node involvement (LNI) in prostate cancer (PCa) patients. MATERIALS AND METHODS: The retrospective study includes 95 patients who underwent mp-MRI and radical prostatectomy for PCa with pelvic lymphadenectomy. Imaging data (intensity in T2, DWI, ADC, and PIRADS), clinical data (age and pre-MRI PSA), histological data (Gleason score, TNM staging, histological type, capsule invasion, seminal vesicle invasion, and neurovascular bundle involvement), and clinical nomograms (Yale, Roach, MSKCC, and Briganti) were collected for each patient. Manual segmentation of the index lesions was performed for each patient using an open-source program (3D SLICER). Radiomic features were extracted for each segmentation using the Pyradiomics library for each sequence (T2, DWI, and ADC). The features were then selected and used to train and test three different radiomics models (LR, RF, and SVM) independently using ChatGPT software (v 4o). The coefficient value of each feature was calculated (significant value for coefficient ≥ ±0.5). The predictive performance of the radiomics models and clinical nomograms was assessed using accuracy and area under the curve (AUC) (significant value for p ≤ 0.05). Thus, the diagnostic accuracy between the radiomics and clinical models were compared. RESULTS: This study identified 343 features per patient (330 radiomics features and 13 clinical features). The most significant features were T2_nodulofirstordervariance and T2_nodulofirstorderkurtosis. The highest predictive performance was achieved by the RF model with DWI (accuracy 86%, AUC 0.89) and ADC (accuracy 89%, AUC 0.67). Clinical nomograms demonstrated satisfactory but lower predictive performance compared to the RF model in the DWI sequences. CONCLUSIONS: Among the prediction models developed using integrated data (radiomics and semantics), RF shows slightly higher diagnostic accuracy in terms of AUC compared to clinical nomograms in PCa lymph node involvement prediction. Full article

Ultrasound elastography has been available on most modern systems; however, the implementation of quality processes tends to be ad hoc. It is essential for a medical physicist to benchmark elastography measurements on each system and track them over time, especially after major software upgrades or repairs. This study aims to establish baseline data using phantoms and monitor them for quality assurance in elastography. In this paper, we utilized two phantoms: a set of cylinders, each with a composite material with varying Young’s moduli, and an anthropomorphic abdominal phantom containing a liver modeled to represent early-stage fibrosis. These phantoms were imaged using three ultrasound manufacturers’ elastography functions with either point or 2D elastography. The abdominal phantom was also imaged using magnetic resonance elastography (MRE) as it is recognized as the non-invasive gold standard for staging liver fibrosis. The scaling factor was determined based on the data acquired using MR and US elastography from the same vendor. The ultrasound elastography measurements showed inconsistency between different manufacturers, but within the same manufacturer, the measurements showed high repeatability. In conclusion, we have established baseline data for quality assurance procedures and specified the criteria for the acceptable range in liver fibrosis phantoms during routine testing. Full article