Search Results (37,829)

Pine wilt disease (PWD) can spread rapidly after the disease occurs and often causes large-scale death of the pine. Therefore, the timely identification of infected trees is critical for forest conservation and effective disease management. However, early infected trees are difficult to distinguish in satellite remote sensing images. Their visual differences from healthy trees and complex background features are often subtle, and existing image-processing methods do not fully exploit heterogeneous information. To address this problem, we constructed the Naro dataset for satellite-based PWD detection and proposed SGM-RTDETR based on Real-Time Detection Transformer (RT-DETR). The proposed model consists of a Semantic–Visual Fusion Module (SVFM) and a Disease Feature Refinement Module (DFRM). In SVFM, ExG, VARI, and GLI are concatenated with RGB imagery to form a six-channel visual input, which enhances the spectral differences between diseased and non-diseased targets. In addition, textual prior knowledge is introduced into the decoder input through a Stackelberg game-based visual–text fusion strategy. This strategy helps the encoded memory features maintain clearer disease-related semantics in complex backgrounds. DFRM then performs channel recalibration, feature refinement, and residual enhancement on the fused memory features to better extract fine-grained disease cues in remote sensing scenes. Experiments on the Naro dataset show that SGM-RTDETR achieves 80.75% mAP@0.5 and 35.43% mAP@0.5:0.95, which is 2.74 percentage points higher than RT-DETR-L on mAP@0.5:0.95. Overall, the results indicate that the dual-module structure improves the precision and robustness of PWD detection in satellite remote sensing images. Full article

Diffusion-based generative models achieve high-fidelity image synthesis; however, controlling internal representations for abstract visual concepts remains challenging due to the ambiguity of textual descriptions. In this work, we propose a prompt-guided concept-vector learning framework for the controllable manipulation of such concepts without requiring external human-annotated image pairs, segmentation masks, identity labels, or manually annotated editing targets. The method introduces a learnable concept vector optimized in the bottleneck (mid-block) feature space of a pretrained Stable Diffusion U-Net, while keeping all pretrained model parameters frozen. A multi-prompt data generation strategy based on paired positive and neutral prompts provides weak semantic guidance for capturing the target concept direction and reducing dependence on a single prompt formulation. The learned vector is further applied in an image-to-image setting through controlled noise injection and concept-guided denoising, enabling the semantic modification of real images while preserving structural content. The concept strength is controlled by a scaling parameter $\gamma$, while the image-to-image noise strength is controlled by $\beta$, allowing for a practical balance between semantic modification and structural fidelity. Experiments are conducted on two main abstract concepts, perfect skin and peaceful lake, with additional qualitative analysis on subjective portrait-level concepts. Quantitative evaluation using SSIM, LPIPS, and CLIP similarity demonstrates that the proposed method improves semantic alignment while maintaining structural preservation compared with Stable Diffusion image-to-image baselines. A human preference study further shows that concept-injected outputs are preferred in 76.0% of responses for perfect skin and 85.7% for peaceful lake. Ablation studies further demonstrate the controllability and robustness of the proposed framework. Overall, the method provides a simple and parameter-efficient approach for interpretable concept-level manipulation in diffusion models. Full article

Background: Multimodal breast ultrasound, including B-mode imaging, color Doppler flow imaging, and elastography, provides complementary information for lesion characterization. However, effectively integrating heterogeneous modalities remains challenging due to inconsistent feature distributions, limited cross-modal interaction, computational cost in existing methods, and sensitivity to noise and missing data. Methods: We presented an efficient Cross-Modal Interaction and Dynamic Fusion Network (CIDFNet) for multimodal breast ultrasound analysis. The framework integrates a multi-scale feature enhancement module to improve modality-specific representations, a cross-modal interaction module to enable early-stage feature exchange across modalities, and a dynamic fusion strategy to adaptively combine modality information based on feature reliability estimation. In addition, an invertible neural network is incorporated to reconstruct missing modality features during training. Results: Experiments on an internal dataset of 248 patients with 1532 images show that CIDFNet obtains an AUC of 85.69%, accuracy of 75.51%, recall of 50.00%, F1-score of 62.50%, and precision of 83.33%, while requiring 49.51 M parameters and 79.79 G FLOPs, respectively. Under a simplified Gaussian noise perturbation setting, performance degradation is observed. Conclusions: CIDFNet presents a framework for multimodal breast ultrasound analysis that reflects a trade-off between performance and computational efficiency. Full article

Background: Elongated styloid processes and ossifications of the stylohyoid chain can be observed on dental imaging modalities. In this study, we assessed the performance of artificial intelligence (AI) in identifying elongated styloid processes and ossifications of the stylohyoid chain. Methods: We performed a systematic review of relevant studies published between April 2020 and April 2026 on PubMed, Scopus, and Web of Science. Relevant data were extracted using predefined criteria. We assessed the risk of bias using categories derived from QUADAS-2, CLAIM and STARD-AI. Results: Four original studies met the inclusion criteria. Of these, only two specifically addressed elongated styloid processes on panoramic images (OPGs). For one study that utilized ML algorithms, both logistic regression and neural networks achieved 100% performance, while naive Bayes demonstrated substantially lower performance than either model. Another study using deep learning algorithms observed accuracy rates of 97.49% and 84.11%, and area under the curve values of 0.9825 and 0.8943 for EfficientNetB5 and InceptionV3 models. A broader study using OPG anomaly detection reported target-level data for stylohyoid ligament ossification. The fourth study used cone-beam computed tomography images, including stylohyoid ligament ossification as part of a multi-class soft tissue calcification/ossification detection task. Due to significant variability in target definitions, imaging modalities, validation methods, and performance metrics across studies, a meta-analysis was not feasible. Conclusions: The use of AI-based systems for detecting elongated styloid processes and stylohyoid chain ossification shows potential for future clinical utility; however, current evidence is insufficient to support independent clinical practice. Future research should incorporate larger-scale prospective multicenter validations as well as external validation on a patient-by-patient basis when possible. Additional research into the clinical implications associated with both false-positive and false-negative results is warranted. Full article

Glaucoma is a leading cause of irreversible vision loss, and its early diagnosis remains critically important yet challenging. Traditional assessment based on the cup-to-disc ratio is often insufficient at early stages, whereas the retinal vascular network can provide additional quantitative biomarkers. This study develops and validates a binary classification method for distinguishing healthy from glaucomatous fundus images by combining deep-learning-based vessel segmentation, fractal and multifractal analysis, and textural features. The public ORIGA dataset is utilized. Images are converted to grayscale using three alternative approaches, followed by Gray-Level Co-occurrence Matrix texture analysis and fractal analysis based on the differential box-counting method. Vessel segmentation is implemented via a U-Net neural network trained on a combination of public datasets, after which multifractal analysis is performed on the resulting binary masks. The extracted features are used to train and compare several machine learning models with hyperparameter optimization. The best-performing model among ONH-based features (Random Forest) achieves 75.00%; however, a logistic regression model using multifractal parameters and CDR reaches 86.17%, substantially outperforming the CDR-only baseline (66.15%). Notably, while classical fractal dimension shows only marginal differences (1–2% relative change) between groups, multifractal parameters reveal distinct changes: the multifractal spectrum width $\Delta \alpha$ increases markedly and the minimum singularity exponent ${\alpha }_{\min }$ decreases in glaucomatous eyes, indicating increased heterogeneity of the vascular network. These findings suggest that multifractal characteristics of the vascular network can serve as reliable and sensitive biomarkers for automated glaucoma screening, offering clear advantages over classical fractal analysis. Full article

Background: Neuroendocrine tumors—are relatively rare but increasingly diagnosed malignancies originating from diffuse neuroendocrine cells, most commonly affecting the gastroenteropancreatic system. Due to their long asymptomatic development and low incidence, pose a diagnostic and therapeutic challenge for physicians. Recently, the role of nuclear medicine has been growing not only in the diagnostic stage but also in treatment. Systemic radionuclide therapy using somatostatin analogs labelled with the radioisotope lutetium-177 is becoming increasingly common in patients with advanced-stage disease. Currently, most patients receive a standard activity of therapeutic radiopharmaceuticals. Recent clinical studies provide increasing evidence of a close relationship between the absorbed radiation dose in pathological lesions and the therapeutic effect of radioisotope therapy. Internal dosimetry is used to measure the doses of ionising radiation absorbed by the patient after administration of the radiopharmaceutical. The lack of individual internal dosimetry prior to therapy means that only a small fraction of patients receive optimal doses of radioactivity, which is markedly different from external beam radiotherapy planning. Methods: A narrative literature review was conducted using the PubMed/MEDLINE and Embase databases, focusing primarily on publications from the last years. The search strategy included combinations of keywords related to peptide receptor radionuclide therapy and dosimetry, such as “Lutetium-177”, “neuroendocrine tumors”, “dosimetry”, “PRRT”, “systemic radionuclide therapy” and “artificial intelligence”. Particular emphasis was placed on recent prospective clinical studies, multicenter investigations, systematic reviews and consensus documents published by major nuclear medicine societies, including the European Association of Nuclear Medicine (EANM) and the Society of Nuclear Medicine and Molecular Imaging (SNMMI). Seminal earlier publications considered essential for understanding the development of dosimetry concepts and clinical implementation were also included. Results: This study confirms the existence of a clinically significant dose-response relationship in ¹⁷⁷Lu-PRRT. Higher absorbed doses to tumour lesions are associated with longer progression-free survival. The lack of individualized internal dosimetry prior to therapy means that only a small proportion of patients receive optimal radiation doses. Simplified dosimetric approaches with a reduced number of imaging time points, together with emerging artificial intelligence–based tools, appear promising for reducing the complexity of the dosimetry process. Conclusions: The aim of this study was to analyse the current literature on the role of internal dosimetry in the treatment of neuroendocrine tumors using the radioisotope lutetium-177. Available data support the clinical relevance of individualized dosimetry and highlight its potential to optimize both therapeutic efficacy and treatment safety. Full article

Robust perception in adverse weather conditions remains a significant challenge for autonomous vehicles. Short-wave infrared (SWIR) sensors offer specific physical properties that enable them to penetrate atmospheric disturbances like fog, rain, and snow. However, effectively combining this robustness with the textural and colour information of visible (VIS) cameras is difficult due to signal decorrelation and the limitations of static fusion schemes. To address this, we present VISWIR (Visible and SWIR Weighted Image Reconstruction), a pixel-level fusion method based on a multi-scale pyramid architecture. We introduce an automated strategy for scheduling parameters based on weather conditions using an optimisation framework. Rather than relying on static weights, our method applies offline parameter scheduling to adjust fusion hyperparameters based on the meteorological context. We focus on a multi-objective optimisation approach that maximises perceptual image quality via No-Reference Image Quality Assessment (NR-IQA) metrics. Validated in controlled environment scenarios with varying weather severities, our results confirm the potential of VISWIR as a robust, lightweight algorithmic baseline to enhance the perception capabilities of autonomous vehicles in adverse weather conditions. Full article

Single-photon detection is a promising approach for low–slow–small Unmanned Aerial Vehicle (UAV) detection, holding great value in urban air defense and security monitoring. In complex urban environments, intense non-uniform building clutter and multi-echo aliasing easily submerge weak target signals, severely limiting traditional single-photon systems under low signal-to-background ratios. To address this, this paper proposes an urban-oriented detection strategy based on a free-running single-photon array, and designs a dual-optimized pulse position modulation laser detection and range image enhancement algorithm. By establishing temporal correlations via pulse sequence convolution, the algorithm effectively isolates weak UAV echoes from strong background clutter to break through detection limitations. Compared with the popular Markov correction method that often suppresses overlapping weak targets under strong reflections, the proposed method significantly improves small-target feature retention, successfully balancing background elimination and detection sensitivity. Field tests and quantitative evaluations demonstrate that the system reliably eliminates building clutter and achieves stable continuous tracking of weak UAV signals within 1.5 km, providing a highly robust and effective technical solution for urban low-altitude surveillance. Full article

Text-to-image generation aims to produce images that match the semantic content of a text prompt. In style transfer tasks, the model must further integrate reference styles while preserving prompt semantics. However, balancing semantic consistency and style fidelity remains challenging. Existing methods commonly rely on fixed feature weights and lack adaptive control, which often leads to style over-injection and content distortion. To address these issues, we propose a novel framework that performs dynamic regulation at both the feature and temporal levels. At the feature level, we propose an Entropy-Aware Adaptive Fusion (EAAF) module. It incorporates a bidirectional distribution transformation mechanism to enhance the statistical correlation between content and style features. The module further uses information entropy as a dynamic control signal to adaptively adjust the strength of style injection, thereby achieving a balance between semantic consistency and style fidelity. At the temporal level, we design a Progressive Feature Reweighting (PFR) strategy. By applying stage-wise weighting to content and style features at different diffusion steps, this strategy effectively improves structural stability and color consistency. In addition, our framework is modular and can be integrated into existing diffusion-based style transfer models without additional fine-tuning or retraining. Experimental results demonstrate that applying our approach to current state-of-the-art models, such as StyleStudio and CSGO, significantly enhances their performance, particularly in maintaining strong prompt alignment while achieving high-fidelity style transfer. Full article

This study presents a novel and accessible method for determining the specific heat capacity of cylindrical lithium-ion batteries without the need for specialized equipment for cell disassembly, climatic chambers, or expensive calorimeters. The proposed approach does not require disassembly of the cell. Since specific heat capacity is a key parameter in thermal modeling and is often unavailable in manufacturer datasheets, the method addresses an important practical gap. The measurement principle is based on recording the change in surface temperature caused by a short 30 s discharge pulse of 9 A. A thermographic camera captures infrared images at fixed time intervals, while an electromechanical module rotates the battery around its longitudinal axis, providing an accurate estimation of the average surface temperature during and after the pulse. The resulting temperature–time profiles are used to evaluate heat losses and compute the specific heat capacity. To validate the methodology, experiments were conducted on an aluminum cylinder of identical dimensions to an 18650 cell, made of Al 6082-T6 (Cp ≈ 896 J·kg⁻¹·K⁻¹). The results show a maximum deviation of 2.68% from the reference value, confirming the reliability of the proposed method for determining the specific heat capacity of cylindrical Li-ion batteries. Full article

“Interbedded water in thin coal seams” is characterized by its high degree of concealment and complex hydraulic connections. However, due to the confined space of underground mine tunnels and severe electromagnetic interference from metal structures, traditional geophysical methods struggle to accurately delineate the boundaries of water accumulation, making this a major and challenging water hazard in coal mines. Taking the Qinhua Coal Mine in Xinjiang, China, as the engineering context, this paper investigates the detection of water accumulation in interbedded coal seams within goaf areas using the cross-hole resistivity method. It proposes a cross-hole resistivity tomography scanning approach characterized by “progressive depth penetration and layer-by-layer traversal,” and employs an inversion method based on inequality constraints to obtain relatively detailed and reliable imaging results. Through resistivity imaging analysis, low-resistivity water accumulation anomalies were successfully delineated, and water accumulation dead zones were identified. Based on the detection results, effective drainage was carried out beneath the water-filled zones. Subsequent follow-up surveys confirmed the disappearance of the low-resistivity anomalies, thereby validating the reliability and engineering practicality of the cross-hole resistivity tomography method for precisely detecting water body boundaries under complex geological conditions in coal seams. Full article

The interfacial transition zone (ITZ) in the aggregate periphery is often regarded as the weakest area in concrete. In this study, the results of extensive image analysis provide a new insight. First, fluorescence microscopy (FM) was adopted to obtain the microstructure of “complete ITZ”, which overcomes several limitations of the scanning electron microscope method. Then, an ITZ recognition interactive algorithm was proposed, which quantitatively characterizes the pore distributions in the ITZ and cement paste in both lateral and longitudinal directions. Finally, based on the experimental and statistical results, the pore distributions around aggregates, coarse sand and fine sand were characterized. Along the lateral direction, a high non-uniformity was observed in the porosity between units, taken at the same distance from the aggregate/sand surface. On the contrary, along the longitudinal direction, statistical results show minimal increases in the porosity within the ITZ. This is also applicable even in the innermost ITZs. Full article

As a next-generation automotive display technology, Augmented Reality Head-Up Display (AR-HUD) has demonstrated immense potential in reshaping driving safety and enhancing the human–computer interaction experience. To address the challenges of barrel distortion and perspective distortion inherent in HUD systems, we propose a joint-learning-based dual-path pre-correction method. This approach employs a shared encoder to extract image features, which are then decoupled into two parallel branches: a classification branch and a distortion flow prediction branch. Building upon this architecture, a model-fitting method is introduced to estimate the distortion model parameters in the parameter space using the predicted distortion types and flows, thereby reconstructing a refined distortion flow. Finally, image rectification is achieved through a resampling method. On the ARHDD dataset, the proposed method achieves a PSNR of 24.617 dB (barrel) and 25.062 dB (perspective), an SSIM of 0.845 and 0.873, and an NRMSE of 0.163 and 0.157, respectively. On the Places 365 dataset, it achieves a PSNR of 23.914 dB (barrel) and 21.870 dB (perspective), an SSIM of 0.812 and 0.748, and an NRMSE of 0.174 and 0.211, respectively. Both quantitative and qualitative comparative experiments against other state-of-the-art methods demonstrate that the proposed approach achieves superior correction performance for both types of distortion. Finally, the simulation verification of the HUD system proved that this correction method demonstrated excellent potential, but further verification is still needed in a real or semi-real environment. Full article

Reconstructing building models from urban SAR tomography (TomoSAR) point clouds is often constrained by limited resolution, low positioning accuracy in elevation, as well as data incompleteness and artifacts caused by microwave imaging mechanisms. These challenges seriously restrict the extraction of high-accuracy building models with structural details from TomoSAR point clouds. This paper proposes a refined urban building modeling method that effectively utilizes structural priors, including directionality, orthogonality, and potential symmetry. First, a piecewise fitting strategy integrated with density-based segmentation is employed to iteratively estimate the main directions of the buildings and capture finer geometric variations of complex façade footprints than simple-plane approximations. Second, a roof extraction algorithm combining an adaptive Doug-las–Peucker approach with symmetry evaluation and constraints is developed to regularize roof outlines and repair data defects. Crucially, to handle extreme cases where roof data are entirely missing, a novel building width estimation method based on building shadow analysis is proposed. Experiments conducted on the SARMV3D-1.0 and SARMV3D-3.0 point cloud datasets demonstrate that the proposed method significantly enhances reconstruction accuracy and geometric fidelity in urban regions compared to state-of-the-art approaches. Full article

Recent studies have proposed improving Laser Speckle Contrast Imaging (LSCI) by using polarimetric filtering to isolate multiply scattered photons from moving red blood cells (RBCs), an approach referred to as Laser Speckle Orthogonal Contrast Imaging (LSOCI). This reliance on multiple scattering enables the development of a calibration method based on a moving reference sample, chosen to generate dynamic circular Gaussian speckle fields that replicate the statistical properties of RBC scattering in both intensity and the distribution of polarization states. Assuming that multiply scattered photons from both RBCs and the reference sample exhibit a homogeneous distribution of polarization states over the Poincaré sphere, the proposed calibration links in vivo speckle contrast reduction bijectively to an equivalent speed of the reference sample. We demonstrate that this equivalent-velocity metric yields consistent in vivo measurements across distinct instruments despite the use of different laser spectral widths, thereby providing a standardized and transferable means to quantify microcirculatory activity. Full article