Search Results (40)

Background and Objectives: Access-site complications (ASCs) remain clinically relevant after peripheral endovascular procedures, particularly with large femoral sheaths and complex anatomy. While randomized coronary trials show non-inferiority of vascular closure devices (VCDs) versus manual compression (MC), real-world data in peripheral interventions performed under systematic ultrasound-guided access are limited. Materials and Methods: This retrospective single-center cohort included consecutive peripheral arterial revascularizations (2010–2023) performed via common femoral access under real-time ultrasound guidance. Hemostasis was achieved using MC or VCDs, categorized as collagen plug-based, suture-mediated, or clip-based systems. The primary endpoint was 30-day ASCs, defined as hematoma requiring management, pseudoaneurysm, bleeding requiring transfusion, access-site thrombosis/occlusion, arteriovenous fistula, or infection. The secondary endpoint was VCD failure, defined as unsuccessful hemostasis requiring adjunctive measures. Multivariable logistic regression adjusted for prespecified anatomical and procedural covariates, including sheath size > 6 Fr and puncture-site calcification. Results: Among 231 procedures, VCDs were used in 139 (60.2%) and MC in 92 (39.8%). ASC occurred in 28 cases (12.1%), with higher rates in the MC group compared with VCDs (18.5% vs. 9–14% across device types; p = 0.044). In adjusted analyses, MC (vs any VCD) (odds ratio [OR] 2.41, 95% confidence interval [CI] 1.06–5.47; p = 0.035), sheath size > 6 Fr, and puncture-site calcification were independently associated with ASCs. VCD failure occurred in 5 cases (3.6%) and was not observed with collagen plug-based devices. Conclusions: In this ultrasound-guided real-world peripheral cohort, VCD use was associated with lower 30-day ASC rates and low device failure rates compared with MC. Given the retrospective and non-randomized design, these findings should be considered hypothesis-generating and support individualized, imaging-guided strategies for femoral closure in peripheral interventions. Full article

The rising demand for real-time perception in aerial platforms has intensified the need for lightweight, hardware-efficient object detectors capable of reliable onboard operation. This survey provides a focused examination of real-time aerial object detection, emphasizing algorithms designed for edge devices and UAV onboard processors, where computation, memory, and power resources are severely constrained. We first review the major aerial and remote-sensing datasets and analyze the unique challenges they introduce, such as small objects, fine-grained variation, multiscale variation, and complex backgrounds, which directly shape detector design. Recent studies addressing these challenges are then grouped, covering advances in lightweight backbones, fine-grained feature representation, multi-scale fusion, and optimized Transformer modules adapted for embedded environments. The review further highlights hardware-aware optimization techniques, including quantization, pruning, and TensorRT acceleration, as well as emerging trends in automated NAS tailored to UAV constraints. We discuss the adaptation of large pretrained models, such as CLIP-based embeddings and compressed Transformers, to meet onboard real-time requirements. By unifying architectural strategies, model compression, and deployment-level optimization, this survey offers a comprehensive perspective on designing next-generation detectors that achieve both high accuracy and true real-time performance in aerial applications. Full article

All-sky imagers (ASIs) used in solar energy meteorology face an extreme intra-image dynamic range, with the circumsolar neighborhood orders of magnitude brighter than the diffuse dome. Many operational ASI pipelines address this gap with high-dynamic-range (HDR) bracketing inside the camera’s image signal processor (ISP), i.e., after demosaicing and color processing in a nonlinear 8-bit RGB domain. Near the Sun, such ISP-domain HDR can down-weight the shortest exposure, retain clipped or near-clipped samples from longer frames, and compress highlight contrast, thereby increasing circumsolar saturation and flattening aureole gradients. A radiance-linear HDR fusion in the sensor/RAW domain (RAW–HDR) is therefore contrasted with the vendor ISP-based HDR mode (ISP–HDR). Solar-based geometric calibration enables Sun-centered analysis. Paired, interleaved acquisitions under clear-sky and broken-cloud conditions are evaluated using two circumsolar performance criteria per RGB channel: (i) saturated-area fraction in concentric rings and (ii) a median-based radial gradient in defined arcs. All quantitative analyses operate on the radiance-linear HDR result; post-merge tone mapping is only used for visualization. Across conditions, ISP–HDR exhibits roughly double the near-saturation within 0–$4^{°}$ of the Sun and about a three- to fourfold weaker circumsolar radial gradient within 0–$6^{°}$ relative to RAW–HDR. These findings indicate that radiance-linear fusion in the RAW domain better preserves circumsolar structure than the examined ISP-domain HDR mode and thus provides more suitable input for downstream tasks such as cloud–edge detection, aerosol retrieval, and irradiance estimation. Full article

To address the challenge of collaborative strategy optimization caused by non-independent and identically distributed data in cross-institutional scenarios, a federated quantitative learning framework integrating Path Attention Aggregation Module (PAAM), Gradient Clipping and Compression (GCC), and a Heterogeneity-Aware Adaptive Optimizer (HAAO) is proposed to achieve efficient return optimization and robust risk control. The framework is validated across multi-market and multi-institutional environments, with experiments covering three key dimensions: return performance, risk management, and communication efficiency. The results demonstrate that the proposed model achieves an annualized return (AR) of 16.57%, representing an approximate 19.7% improvement over the traditional FedAvg model; the Sharpe ratio (SR) increases to 1.25, while the maximum drawdown (MDD) decreases to 15.92% and volatility remains controlled at 8.83%, indicating superior balance between return and risk. In the communication efficiency evaluation, when the number of communication rounds is reduced to 50 and 25, the model maintains accuracy at 94.2% and 92.8%, recall at 93.5% and 91.7%, and precision at 94.8% and 92.3%, respectively. Overall, the proposed framework achieves a dynamic balance between global convergence and risk constraints through path weighting, gradient sparsification, and frequency-domain learning rate adjustment. This research provides a novel and scalable paradigm for distributed financial prediction that ensures both privacy preservation and communication efficiency, demonstrating substantial engineering feasibility and practical applicability in intelligent financial modeling. Full article

Few-shot image classification aims to recognize novel classes from only a handful of labeled examples, a challenge in domains where data collection is costly or impractical. Existing solutions often rely on meta learning, fine tuning, or data augmentation, introducing computational overhead, risk of overfitting, or are not highly efficient. This paper introduces ProbaCLIP, a simple training-free approach that leverages Kernel Density Estimation (KDE) within the embedding space of Contrastive Language-Image Pre-training (CLIP). Unlike other CLIP-based methods, the proposed approach operates solely on visual embeddings and does not require text labels. Class-conditional probability densities were estimated from few-shot support examples, and queries were classified by likelihood evaluation, where Principal Component Analysis (PCA) was used for dimensionality reduction, compressing the dissimilarities between classes on each episode. We further introduced an optional bandwidth optimization strategy and a consensus decision mechanism through cross-validation, while addressing the special case of one-shot classification with distance-based measures. Extensive experiments on multiple datasets demonstrated that our method achieved competitive or superior accuracy compared to the state-of-the-art few-shot classifiers, reaching up to 98.37% accuracy in five-shot tasks and up to 99.80% in a 16-shot framework with ViT-L/14@336px. We proved our methodology by achieving high performance without gradient-based training, text supervision, or auxiliary meta-training datasets, emphasizing the effectiveness of combining pre-trained embeddings with statistical density estimation for data-scarce classification. Full article

Background and Clinical Significance: Tandem pathology at the dominant-hemisphere middle cerebral artery (MCA)—combining a wide-neck bifurcation aneurysm that shares the neck with both M2 origins and a short dorsal M1 fusiform dilation embedded in the lenticulostriate belt—compresses the therapeutic margin and complicates device-first pathways. We aimed to describe an anatomy-led, microscope-only sequence designed to secure an immediate branch-definitive result at the fork and to remodel dorsal M1 without perforator compromise, and to place these decisions within a pragmatic perioperative framework. Case Presentation: A 37-year-old right-handed man with reproducible, load-sensitive cortical association and capsulostriate signs underwent high-fidelity digital subtraction angiography (DSA) with 3D rotational reconstructions. Through a left pterional approach, vein-respecting Sylvian dissection achieved gravity relaxation. Reconstruction proceeded in sequence: a fenestrated straight clip across the bifurcation neck with the superior M2 encircled to preserve both M2 ostia, followed by a short longitudinal clip parallel to M1 to reshape the fusiform segment while keeping each lenticulostriate mouth visible and free. Temporary occlusion windows were brief (bifurcation 2 min 30 s; M1 < 2 min). No neuronavigation, intraoperative fluorescence, micro-Doppler, or intraoperative angiography was used. No perioperative antiplatelets or systemic anticoagulation were administered and venous thromboembolism prophylaxis followed institutional practice. The bifurcation dome collapsed immediately with round, mobile M2 orifices, and dorsal M1 regained near-cylindrical geometry with patent perforator ostia under direct inspection. Emergence was neurologically intact, headaches abated, and preoperative micro-asymmetries resolved without new deficits. The early course was uncomplicated. Non-contrast CT at three months showed structurally preserved dominant-hemisphere parenchyma without infarction or hemorrhage. Lumen confirmation was scheduled at 12 months. Conclusions: In dominant-hemisphere tandem MCA disease, staged, perforator-sparing clip reconstruction can restore physiologic branch and perforator behavior while avoiding prolonged antiplatelet exposure and device-related branch uncertainty. A future-facing pathway pairs subtle clinical latency metrics with high-fidelity angiography, reports outcomes in branch- and perforator-centric terms, and, where available, incorporates patient-specific hemodynamic simulation and noninvasive lumen surveillance to guide timing, technique, and follow-up. Full article

Background/Objectives: Hip-level amputees face ambulatory challenges due to the lack of a lower limb and prosthetic hip power. Some hip-level amputees restore mobility by using a prosthesis with hip, knee, and ankle joints. Powered prosthetic joints contain an actuator that provides external flexion-extension moments to assist with movement. Powered knee and powered ankle-foot units are on the market, but no viable powered hip unit is commercially available. This research details the development of a novel powered four-bar prosthetic hip joint that can be integrated into a full-leg prosthesis. Methods: The hip joint design consisted of a four-bar linkage with a harmonic drive DC motor placed in the inferior link and an additional linkage to transfer torque from the motor to the hip center of rotation. Link lengths were determined through engineering optimization. Device strength was demonstrated with force and finite element analysis and with ISO 15032:2000 A100 static compression tests. Walking tests with a wearable hip-knee-ankle-foot prosthesis simulator, containing the novel powered hip, were conducted with three able-bodied participants. Each participant walked back and forth on a level 10 m walkway. Custom hardware and software captured joint angles. Spatiotemporal parameters were determined from video clips processed in the Kinovea software (ver. 0.9.5). Results: The powered hip passed all force and finite element checks and ISO 15032:2000 A100 static compression tests. The participants, weighing 96 ± 2 kg, achieved steady gait at 0.45 ± 0.11 m/s with the powered hip. Participant kinematic gait profiles resembled those seen in transfemoral amputee gait. Some gait asymmetries occurred between the sound and prosthetic legs. No signs of mechanical failure were seen. Most design requirements were met. Areas for powered hip improvement include hip flexion range, mechanical advantage at high hip flexion, and device mass. Conclusions: The novel powered four-bar hip provides safe level-ground walking with a full-leg prosthesis simulator and is viable for future testing with hip-level amputees. Full article

Background and Clinical Significance: Spinal arachnoid cysts are rare lesions that may become symptomatic through progressive spinal cord compression. We present a complex case of a thoracic extradural SAC in a 17-year-old male, managed through a stepwise, multidisciplinary approach. Case Presentation: The patient presented with progressive lower limb weakness, right knee paresthesia, and urinary hesitancy following physical exertion. MRI revealed a large posterior extradural SAC extending from T2–T3 to T8, with associated spinal cord compression. Initial management involved T8 laminectomy and cyst fenestration under intraoperative neurophysiological monitoring, with partial clinical improvement. However, early recurrence with pseudomeningocele formation prompted a second surgery, including external CSF drainage. Persistent cerebrospinal fluid (CSF) leakage led to targeted epidural blood patching, followed by temporary stabilization. Due to continued cyst enlargement and spinal cord compression, definitive surgical repair was undertaken: fistula clipping at T3 and embolization with platinum coils inside the cystic cavity, combined with a new blood patch. This novel technique resulted in radiological improvement and clinical stabilization. Conclusions: This case highlights the diagnostic and therapeutic challenges of managing symptomatic extradural SACs, particularly in young patients. Our experience underscores the utility of a staged approach involving surgical decompression, neuroimaging-guided interventions, and definitive dural repair. The combination of fistula clipping and coil embolization may offer a promising strategy for refractory cases, potentially reducing recurrence and preserving neurological function. Full article

With the increasing demand for privacy preservation and strategy sharing in global financial markets, traditional centralized modeling approaches have become inadequate for multi-institutional collaborative tasks, particularly under the realistic challenges of multi-source heterogeneity and non-independent and identically distributed (non-IID) data. To address these limitations, a heterogeneity-aware Federated Quantitative Learning framework, Federated Quantitative Learning, is proposed to enable efficient cross-market financial strategy modeling while preserving data privacy. This framework integrates a Path Quality-Aware Aggregation Mechanism, a Gradient Clipping and Compression Module, and a Heterogeneity-Adaptive Optimizer, collectively enhancing model robustness and generalization. Empirical studies conducted on multiple real-world financial datasets, including those from the United States, European Union, and Asia-Pacific markets, demonstrate that Federated Quantitative Learning outperforms existing mainstream methods in key performance indicators such as annualized return, Sharpe ratio, maximum drawdown, and volatility. Under the full model configuration, Federated Quantitative Learning achieves an annualized return of 12.72%, a Sharpe ratio of 1.12, a maximum drawdown limited to 10.3%, and a reduced volatility of 9.7%, showing significant improvements over methods such as Federated Averaging, Federated Proximal Optimization, and Model-Contrastive Federated Learning. Moreover, module ablation studies and attention mechanism comparisons further validate the effectiveness of each core component in enhancing model performance. This study introduces a novel paradigm for secure strategy sharing and high-quality modeling in multi-institutional quantitative systems, offering practical feasibility and broad applicability. Full article

With the rapid advancement of global digital transformation, enterprises and financial institutions face increasing challenges in managing and processing receipt-like financial documents. Traditional manual document processing methods can no longer meet the demands of modern office operations and business expansion. To address these issues, automated document recognition systems based on computer vision and deep learning technologies have emerged. This paper proposes a receipt recognition technology based on multimodal alignment and lightweight sequence modeling, integrating the CLIP (Contrastive Language-Image Pretraining) and Bidirectional Gated Recurrent Unit (BiGRU) framework. The framework aims to achieve synergistic optimization of image and text information through semantic correction. By leveraging dynamic threshold classification, geometric regression loss, and multimodal feature alignment, the framework significantly improves text detection and recognition accuracy in complex layouts and low-quality images. Experimental results show that the model achieves a detection F1 score of 93.1% and a Character Error Rate (CER) of 5.1% on the CORD dataset. Through a three-stage compression strategy of quantization, pruning, and distillation, the model size is reduced to 18 MB, achieving real-time inference speeds of 25 FPS on the Jetson AGX Orin edge device, with power consumption stabilized below 12 W. This framework provides an efficient, accurate, and edge-computing-friendly solution for automated receipt processing. Practical implications include its potential to enhance the efficiency of financial audits, improve tax compliance, and streamline the operational management of financial institutions, making it a valuable tool for real-world applications in receipt automation. Full article

Background/Objectives: Prior studies have noted varied, spontaneous motor recovery in rat strains after spinal cord injury (SCI), but systematic comparisons of different locomotor measurements across different severity and sexes are lacking. Hence, we quantified hindlimb utilization in male and female Sprague-Dawley (SD) and Wistar rats following moderate and severe SCI. Methods: Compression SCI was induced using a 15-g clip for 180 s for moderate SCI or a 50-g aneurysm clip for 60 s for severe SCI in male and female SD and Wistar rats. Measures of locomotor performance using the Basso–Beattie–Bresnahan (BBB), CatWalk gait analysis, and horizontal ladder tests were taken postoperatively and weekly for seven weeks. Results: BBB scores indicated greater spontaneous recovery in SD rats, with females showing higher scores than males following moderate and severe SCI. No sex or strain differences were observed in the horizontal ladder test. The CatWalk results indicated greater average hindlimb swing speed in SD rats following moderate SCI, but greater print area was observed in Wistar rats after severe SCI, although female SD rats had greater print area than either male SD or female Wistar rats following moderate SCI. Conclusions: The findings that SD rats, especially females, exhibited greater spontaneous motor recovery following moderate SCI indicate the need to consider the sex and strain of rats when conducting therapeutic testing following moderate SCI. The significance of these findings is that they should facilitate the use of appropriate rat models for translational research in SCI that can be applied to future clinical trials. Full article

This paper proposes the “ultimate compression” method as a solution to the expansive computation and high storage costs required by state-of-the-art neural network models in inference. Our approach uniquely combines tensor decomposition techniques with binary neural networks to create efficient deep neural network models optimized for edge inference. The process includes training floating-point models, applying tensor decomposition algorithms, binarizing the decomposed layers, and fine tuning the resulting models. We evaluated our approach in various state-of-the-art deep neural network architectures on multiple datasets, such as MNIST, CIFAR-10, CIFAR-100, and ImageNet. Our results demonstrate compression ratios of up to 169×, with only a small degradation in accuracy (1–2%) compared to binary models. We employed different optimizers for training and fine tuning, including Adam and AdamW, and used norm grad clipping to address the exploding gradient problem in decomposed binary models. A key contribution of this work is a novel layer sensitivity-based rank selection algorithm for tensor decomposition, which outperforms existing methods such as random selection and Variational Bayes Matrix Factorization (VBMF). We conducted comprehensive experiments using six different models and present a case study on crowd-counting applications, demonstrating the practical applicability of our method. The ultimate compression method outperforms binary neural networks and tensor decomposition when applied individually in terms of storage and computation costs. This positions it as one of the most effective options for deploying compact and efficient models in edge devices with limited computational resources and energy constraints. Full article

In recent years, deep-learning methods have significantly improved the classification results in the field of plant-leaf recognition. However, limited by the model input, the original image needs to be compressed to a certain size before it can be input into the convolutional neural network. This results in great changes in the shape and texture information of some samples, thus affecting the classification accuracy of the model to a certain extent. Therefore, a minimum enclosing quadrate (MEQ) method is proposed to standardize the sample datasets. First, the minimum enclosing rectangle (MER) of the leaf is obtained in the original image, and the target area is clipped. Then, the minimum enclosing quadrate of the leaf is obtained by extending the short side of the rectangle. Finally, the sample is compressed to fit the input requirements of the model. In addition, in order to further improve the classification accuracy of plant-leaf recognition, an EC-ResNet50 model based on transfer-learning strategy is proposed and further combined with the MEQ method. The Swedish leaf, Flavia leaf, and MEW2012 leaf datasets are used to test the performance of the proposed methods, respectively. The experimental results show that using the MEQ method to standardize datasets can significantly improve the classification accuracy of neural networks. The Grad-CAM visual analysis reveals that the convolutional neural network exhibits a higher degree of attention towards the leaf surface features and utilizes more comprehensive feature regions during recognition of the leaf samples processed by MEQ method. In addition, the proposed MEQ + EC-ResNet50 method also achieved the best classification results among all the compared methods. This experiment provides a widely applicable sample standardization method for leaf recognition research, which can avoid the problem of sample deformation caused by compression processing and reduce the interference of redundant information in the image to the classification results to a certain degree. Full article

The clip-induced spinal cord injury (SCI) rat model is pivotal in preclinical SCI research. However, the literature exhibits variability in compression duration and limited attention to clip deformation-related loss of closure force. We aimed to investigate the impact of compression duration on SCI severity and the influence of clip deformation on closure force. Rats received T10-level clip-induced SCI with durations of 1, 5, 10, 20, and 30 s, and a separate group underwent T10 transection. Outcomes included functional, histological, electrophysiological assessments, and inflammatory cytokine analysis. A tactile pressure mapping system quantified clip closure force after open–close cycles. Our results showed a positive correlation between compression duration and the severity of functional, histological, and electrophysiological deficits. Remarkably, even a brief 1-s compression caused significant deficits comparable to moderate-to-severe SCI. SSEP waveforms were abolished with durations over 20 s. Decreased clip closure force appeared after five open–close cycles. This study offers critical insights into regulating SCI severity in rat models, aiding researchers. Understanding compression duration and clip fatigue is essential for experiment design and interpretation using the clip-induced SCI model. Full article

Among various network compression methods, network quantization has developed rapidly due to its superior compression performance. However, trivial activation quantization schemes limit the compression performance of network quantization. Most conventional activation quantization methods directly utilize the rectified activation functions to quantize models, yet their unbounded outputs generally yield drastic accuracy degradation. To tackle this problem, we propose a comprehensive activation quantization technique namely Bilateral Clipping Parametric Rectified Linear Unit (BCPReLU) as a generalized version of all rectified activation functions, which limits the quantization range more flexibly during training. Specifically, trainable slopes and thresholds are introduced for both positive and negative inputs to find more flexible quantization scales. We theoretically demonstrate that BCPReLU has approximately the same expressive power as the corresponding unbounded version and establish its convergence in low-bit quantization networks. Extensive experiments on a variety of datasets and network architectures demonstrate the effectiveness of our trainable clipping activation function. Full article