Search Results (5,804)

Automatic defect segmentation of atmospheric plasma sprayed (APS) thermal barrier coatings (TBCs) from scanning electron microscopy (SEM) images is essential for quantitative microstructural assessment; however, existing instance segmentation models often exhibit insufficient multi-scale representation and excessive computational cost. To address these limitations, we propose an improved YOLO11-seg–based framework that jointly optimizes accuracy and efficiency through two architectural enhancements: a Boundary-aware Recalibrated Bidirectional Pyramid Network (BRB-FPN) for strengthened cross-scale feature fusion and precise contour localization, and an Efficient Shared-Convolution Segmentation Head (ESCH) for reducing parameters and FLOPs while accelerating inference. Without introducing additional training strategies, the proposed method achieves consistent performance gains on a self-constructed APS–TBC SEM defect dataset, improving Mask mAP@50 by 6.9%, Mask mAP@50:95 by 4.4%, and recall by 5.5% over the baseline, while simultaneously lowering computational complexity and increasing FPS, and outperforming several mainstream segmentation approaches. These results demonstrate that the proposed framework provides an accurate and computationally efficient solution for industrial microscopic defect inspection. Full article

Background/Objectives: Computer-aided vessel occlusion detection for acute ischemic stroke is typically developed on early-phase CT angiography (CTA). In late-phase CTA, enhanced venous contrast can mimic arterial occlusions and lead to incorrect occlusion detections. To counteract this, we propose a constraint-based post-processing framework that leverages anatomical and temporal vascular information from CT perfusion (CTP) scans. Methods: Vascular anatomy and relative contrast-arrival times were encoded as time–vessel maps. Three complementary constraints worked on these maps to remove incorrect detections based on distance-to-vessel truncation points, contrast-arrival time, and spatial priors. Constraints were fixed on a 32-case late-phase development subset and applied to an external held-out cohort of late-phase CTA (N = 354). To enable application without CTP, an auxiliary CTA-to-time–vessel map generator was designed to estimate maps directly from CTA. Results: In the external cohort, the framework reduced incorrect occlusions per scan from 3.9 (95% CI 3.6–4.2) to 2.1 (95% CI 1.9–2.3), preserving occlusion-level sensitivity at 80% (95% CI 69–91). The patient-level AUROC was 83 for the baseline versus 88 with post-processing. Improvements were also observed in detectors exposed to late-phase CTA, where post-processing reduced incorrect occlusions per scan from 2.8 to 1.9. As only 51 occlusion-positive cases were present in the external cohort, confidence intervals were wide. Consequently, these estimates should be interpreted with caution. Conclusions: Anatomically and temporally informed post-processing reduced incorrect occlusion detections in late-phase CTA without measurable loss of sensitivity, potentially improving robustness against CTA phase changes in external institutions. Full article

Background: Frailty and impaired muscle quality are strong determinants of adverse outcomes after coronary artery bypass grafting (CABG), yet conventional risk scores overlook this biological dimension. The Hounsfield Unit Average Calculation (HUAC) obtained from routine preoperative computed tomography (CT) scans quantifies skeletal muscle attenuation, reflecting intramuscular fat infiltration and metabolic reserve. This study evaluated the prognostic value of HUAC measured at two thoracic levels and explored sex-specific differences in its predictive performance. Methods: In this retrospective single-center study, 479 patients (382 men, 97 women) undergoing isolated CABG with available preoperative CT scans were analyzed. Mean HUAC was measured bilaterally at the fourth (TH4) and twelfth (TH12) thoracic vertebral levels. Sex-specific HUAC quartiles were generated, with the lower quartile defining low thoracic muscle attenuation, representing a radiographic phenotype of impaired muscle quality. Associations with 1-, 3-, and 5-year mortality were assessed by Kaplan–Meier analysis, receiver-operating-characteristic (ROC) curves, and sex-stratified Cox regression models incorporating inverse probability weighting (IPW). Incremental prognostic value beyond EuroSCORE II was assessed by comparing Harrell’s C-statistics of Cox models with and without low HUAC. Results: Patients with low HUAC showed reduced survival at all time points. HUAC measured at TH4 demonstrated higher discriminative ability compared with TH12. In women, TH4 yielded area under the curve (AUC) of 0.773 and 0.807 for 1- and 3-year survival, respectively, compared with 0.670 and 0.640 in men. In sex-specific Cox analyses, women in the lower HUAC-TH4 quartile had markedly increased mortality risk (unweighted HRs 6.6, 9.5, and 6.4 for 1-, 3-, and 5-year follow-up; all p ≤ 0.03), which remained significant after IPW adjustment. In men, the association was weaker and lost significance after weighting. Formal interaction testing demonstrated significant sex × low-HUAC interactions at TH4 for 3- and 5-year mortality (p interaction = 0.046 and 0.045, respectively), whereas no significant interaction was observed at TH12. Addition of low TH4 HUAC to EuroSCORE II improved model discrimination in the overall cohort, with C-statistics increasing from 0.665 to 0.711 at 3 years and from 0.658 to 0.690 at 5 years (both likelihood-ratio p < 0.001); the largest incremental gains were observed in women (ΔC = +0.115 and +0.081, respectively). Conclusions: Low thoracic muscle attenuation, particularly HUAC at TH4, independently predicts long-term mortality after CABG. Formal interaction testing supports an exploratory sex-dependent prognostic association at TH4, with a stronger association of low muscle attenuation with mortality in women than in men. HUAC represents a simple, objective biomarker of physiological reserve that can be derived from existing preoperative CT scans and may enhance individualized risk stratification in contemporary cardiac surgery. Full article

X-ray computed tomography (CT) is widely used in non-destructive testing; however, it exhibits inherent limitations when scanning objects with a large aspect ratio. Computed laminography (CL) provides an alternative solution; nevertheless, due to its distinctive scanning geometry, analytical CL reconstruction methods, such as the filtered back projection (FBP) algorithm, are prone to laminographic and superposition artifacts, thereby degrading image quality. To address this issue, a perspective transformation-based data conversion method (PT-CL-FDK) for cone-beam CL is proposed. Firstly, a virtual detector is introduced, and the perspective transformation relationship between the real and virtual detectors is derived. Secondly, the acquired CL projection data are transformed onto the virtual detector. Finally, the converted projection data are reconstructed using a CL-FDK algorithm. Simulation and real printed circuit board (PCB) experimental results demonstrate that, compared with CL-FBP, the proposed method effectively suppresses superposition artifacts and preserves object boundaries, while significantly enhancing the contrast and resolution of void defects in PCBs. Full article

Discrete element method (DEM) has become an essential tool for analyzing particle-scale behavior and complex interactions in agricultural materials and machinery. This review synthesizes recent advances in DEM-based modeling and applications in agricultural engineering. It systematically examines particle modeling strategies, contact model development, and parameter calibration frameworks for representative materials such as soil, seeds, fertilizers, stalks, and roots. The evolution from simple spherical approximations to multi-sphere, bonded, and 3D scan-based reconstructions is highlighted, along with the shift from classical Hertz–Mindlin models to cohesive, elasto-plastic, and fracture-capable formulations. Calibration methods are traced from empirical assignment to systematic frameworks that integrate design of experiments, response surface methodology, and machine-learning-assisted inverse analysis. These advances in fidelity and accuracy have enabled extensive DEM applications in tillage, seeding, fertilization, and harvesting, with emphasis on equipment optimization, mechanism analysis, and performance prediction. Despite the progress, challenges remain in model standardization, parameter transferability, computational cost, and multiphysics coupling. Future work points to unified modeling frameworks, standardized databases, real-time simulation, and integration with artificial intelligence to support digital agriculture and intelligent machinery. This review aims to serve as a reference for high-fidelity DEM modeling and for advancing the digital transformation of agricultural engineering. Full article

Access to reliable medical images is essential for clinical training. To address this need, this paper presents an extended version of MIRAGE, a multimodal retrieval and generation system that utilizes a shared latent space to process medical queries by retrieving real images from the ROCO dataset, generating synthetic scans, and providing LLM-based clinical descriptions alongside dual-concept visual comparisons. To overcome previous computational limits and the lack of clinical validation, we introduce three core enhancements: first, an Auto-$\alpha$ module to dynamically weight visual and textual similarities; second, the integration of LCM-LoRA to accelerate synthetic image generation; and third, an automated clinical auditor based on Gemini 2.5 Flash. Experimental results demonstrate that Auto-$\alpha$ improves retrieval accuracy for heterogeneous queries, reaching $38.83$% Top-1 Recall over a 65,419-image gallery and outperforming nine fusion baselines evaluated under a unified configuration, with a controlled ablation attributing most of this gain to learning the weight rather than merely making it query-adaptive, while the LCM-LoRA module reduces computational costs by a factor of 12.5× in CPU environments, with a blinded radiologist evaluation confirming only a small drop in clinical quality. Furthermore, the clinical auditor achieves a 0.805 Pearson correlation against an expert radiologist, effectively correcting the systematic overestimation of traditional CLIP scores. Finally, the optimized platform is publicly deployed on Hugging Face. Full article

Background/Objectives: ATAC-seq footprinting can infer transcription-factor (TF) occupancy across the genome at near-base-pair resolution. However, its broad adoption is limited by high computational demands, complex command-line workflows, and fragmented support for bulk data with biological replicates and single-cell data. We developed fp-tools, a Python package that extends the TOBIAS framework into an integrated and reproducible platform for TF footprinting and motif discovery. Methods: fp-tools provides command-line and graphical workflows for Tn5 bias correction, footprint scoring, motif scanning, de novo motif discovery, replicate-aware differential analysis, scaled motif aggregation, and pseudobulk processing of single-cell ATAC-seq data. It compares TF occupancy across conditions using corrected cut-site profiles and motif-centered footprint scores. The package also produces interactive HTML reports, editable figures, and reusable YAML configurations. Results: In analyses of ENCODE ATAC-seq replicates from seven cancer cell lines, fp-tools recovered expected cell-type-associated TF programs, including erythroid and hepatocyte-lineage regulators. Validation against matched ChIP-seq data from four lines yielded a median area under the receiver operating characteristic curve (AUROC) of 0.765. In a public single-cell peripheral blood mononuclear cell (PBMC) ATAC-seq dataset, the pseudobulk workflow identified cell-type-specific footprint signatures across immune cell populations. fp-tools also discovered de novo motifs from candidate footprints that did not match known motif databases. In runtime benchmarking, fp-tools completed analyses faster and used less peak memory than the matched TOBIAS workflow tested in this study. Conclusions: fp-tools makes TOBIAS-style ATAC-seq footprinting more accessible and computationally efficient for bulk and single-cell studies. Its command-line tools, graphical interface, and interactive reports support reproducible analysis of TF occupancy in public and user-generated ATAC-seq datasets. Source code, examples, and documentation are available through the project’s GitHub repository and documentation website. Full article

Background: This study aimed to describe the long-term clinical trajectory, stone burden changes, and secondary endourological reintervention outcomes in patients presenting with residual stone fragments (RSFs) > 4.0 mm confined to the lower pole following percutaneous nephrolithotomy (PCNL). Methods: We retrospectively reviewed a single-center cohort of 29 consecutive patients who underwent index-PCNL for isolated lower pole calculi and presented with RSFs > 4.0 mm on routine non-contrast computed tomography (NCCT) at 3 months postoperatively (defined as study baseline and time zero), followed for a minimum of 12 months. Demographic data, baseline and terminal stone burden, stone-related clinical events, procedural complications, and secondary endourological interventions were analyzed. Results: Twenty-nine patients (20 males, 9 females; mean age: 50.1 ± 12.0 years) were analyzed. Preoperative median two-dimensional (2D) stone burden was 332.0 mm² (IQR: 282.0–386.0 mm²). At the 3-month baseline NCCT, median residual 2D stone burden was 24.0 mm² (IQR: 13.0–33.0 mm²). Over a median follow-up of 36.0 months (IQR: 18.0–48.0 months), four patients (13.8%, 95% CI: 3.9–31.7%) achieved spontaneous stone passage, 15 patients (51.7%, 95% CI: 32.5–70.6%) required secondary reintervention due to interval enlargement or symptoms, and 10 patients (34.5%, 95% CI: 17.9–54.3%) remained on active surveillance. Across the cohort, median stone burden on the last positive scan prior to clinical events or the final surveillance visit increased to 57.0 mm² (IQR: 38.0–104.0 mm²). Conclusions: In this selected single-center cohort of 29 patients with lower-pole residual fragments > 4.0 mm at 3-month NCCT after PCNL, 15 patients underwent secondary intervention and four experienced documented spontaneous passage during variable follow-up. Because interval monitoring and intervention criteria were individualized, these findings are descriptive and do not establish the optimal timing of treatment. Larger prospective studies with standardized imaging protocols are required to better define management strategies. Full article

Medical images such as Magnetic Resonance Imaging (MRI) scans consist of various confidential data that can hamper the privacy of the patient if leaked. Organizations have to abide by various international standards in order to ensure the safe storage and transmission of medical data. Strong encryption is a solution; however, conventional full-image encryption which treats every pixel as equally sensitive results in high computational costs. The resulting energy and carbon emissions pose an environmental challenge. Adaptive selective encryption secures images based on the sensitivity of different portions, resulting in lower carbon emissions. However, these claims have rarely been validated in existing studies covering both real adaptive detection performance and fair energy accounting. To alleviate these issues, this study presents DNA-AdCrypt, a carbon-aware deoxyribonucleic acid (DNA)-inspired tiered encryption system to secure MRI scans. The approach is simulated on brain tumors based on the Ultralytics MRI scan dataset. Fine-tuned You Only Look Once (YOLO) detectors are used to segment the scans into three domains, based on which the strength of the cryptosystem changes. YOLOv8n outperforms other variants, with the highest mAP50 (0.538) and mAP50-95 (0.396). In addition, DNA-Adcrypt reduces the encryption time by up to ≈49% and remains competitive in per-image energy against different simulated ciphers. This study puts forward a design intended to generalize for different modalities in order to promote a greener deployable medical security standard, with brain tumor MRI as an initial validated case. Full article

Background/Objectives: Detecting panoramic curves is a critical step in dental imaging, as it serves as the foundation for generating high-quality panoramic radiographs from Cone Beam Computed Tomography (CBCT) scans. These curves trace the dental arch, ensuring that key anatomical structures, such as teeth, alveolar ridges, and jaws, are accurately represented in a single 2D image. This study aimed to evaluate the performance of the PAN_CURVE system across diverse clinical scenarios. Methods: The system was evaluated in a preliminary external validation performed within a single imaging platform on a dataset of 50 CBCT scans acquired with the same CBCT device and including diverse dental conditions, such as edentulous zones and metal elements like implants and orthodontic devices. Performance was assessed in terms of root mean square error (RMSE), mean absolute error (MAE), and processing time. Results: The system achieved an average RMSE of 1.55 ± 2.91 mm and a MAE of 1.28 ± 2.43 mm. Its processing time averaged 4.70 ± 4.33 s per scan, demonstrating efficiency while meeting usability requirements. Conclusions: These preliminary findings support the potential suitability of the PAN_CURVE system for integration into clinical visualization and planning software for dental and maxillofacial applications; confirmation on larger, multi-device datasets is required before generalized conclusions can be drawn. Full article

Background/Objectives: The ultra-high-resolution (UHR) mode of photon-counting detector computed tomography (PCD-CT) is the most dose-intensive series of a multiphase pancreatic protocol in relation to its scan length, and the pancreas cannot be localised on a topogram. We assessed whether an ultra-low-dose (ULD) unenhanced planning series reduces its scan range and dose without compromising complete depiction of the pancreas. Methods: In this prospective single-centre study, 61 pancreatic protocol examinations in 56 patients were performed on a first-generation dual-source PCD-CT. Following a protocol change, the UHR scan range was planned from the topogram in group A (n = 30) and from a ULD unenhanced series in group B (n = 31). Dose indices, scan length, margins to the pancreas and completeness of depiction were recorded. Results: The UHR scan range was 27% shorter in group B (114 vs. 160 mm; p < 0.001), driven by the cranial margin (−58%; p < 0.001) rather than the caudal margin (p = 0.076). Neither the volume CT dose index nor the size-specific dose estimate of the UHR series differed (p = 0.74 and p = 0.35), yet its dose–length product fell by 25% (p = 0.003) and its share of the dose excluding the mandatory venous phase fell from 62.0% to 50.7% (p < 0.001). The total dose was 7.5% lower without reaching significance (p = 0.37). Incomplete depiction occurred in 5 of 30 vs. 5 of 31 examinations (p = 1.00). The body habitus was closely matched. Conclusions: A ULD planning series was associated with a more tightly targeted UHR pancreatic phase, with a 27% shorter scan range and a 25% lower dose–length product for that series, and with no detected increase in incomplete depiction of the pancreas. With topogram-based planning alone, over-scanning was most pronounced at the cranial border. The reduction was confined to the ultra-high-resolution series and was statistically significant there; the total examination dose was lower with ultra-low-dose planning, but this difference was not statistically significant and no conclusion about total dose should be drawn from it. Full article

Background/Objectives: Severe posterior mandibular atrophy presents anatomical challenges for implant rehabilitation because of reduced residual bone volume and the close relationship between the mental foramen, its anatomical variations, and the surrounding alveolar bone. Among these variations, the anterior loop may further influence the available lingual bone adjacent to the mental foramen. Despite the clinical importance of this region, quantitative three-dimensional anatomical information regarding lingual bone availability remains limited. Therefore, the aim of this retrospective cone-beam computed tomography (CBCT)-based study was to quantitatively evaluate the availability of lingual bone adjacent to the mental foramen in posterior atrophied mandibles and to assess the influence of the anterior loop on these anatomical dimensions, thereby providing quantitative anatomical information about this region. Methods: In a retrospective CBCT-based anatomical study, 110 CBCT scans of adult patients with atrophied mandibles were analyzed using standardized cross-sectional views perpendicular to the dental arch; linear measurements were obtained from the lingual cortical plate to the mesial (Point A), middle (Point B), and distal (Point C) borders of the mental foramen. The presence of the anterior loop was recorded. Descriptive statistics and independent-samples t-tests were performed to compare measurements between loop and no-loop groups (α = 0.05). Results: The mean measurements of the lingual bone were 7.22 ± 1.65 mm at Point A, 6.18 ± 1.55 mm at Point B, and 4.82 ± 1.38 mm at Point C, showing a gradual decrease from mesial to distal. An anterior loop was found in 26.4% of the cases. Significantly thinner dimensions of the lingual bone were observed with an anterior loop at all three measurement points (p < 0.01). Conclusions: The present CBCT-based anatomical study measured the amount of lingual bone available adjacent to the mental foramen in posterior atrophied mandibles and observed a significant correlation between the presence of an anterior loop and reduced lingual bone dimensions. These findings add to the knowledge of the anatomy of this area and could serve as the basis for future clinical studies to investigate the potential for a mental foramen bypass concept. Full article

Background/Objectives: A recent feature in the Invisalign^® system allowed integration of cone-beam computed tomography (CBCT) with intraoral scans, thereby enabling visualization of the tooth roots and surrounding alveolar bone. Here, we aimed to compare differences between anatomical measurements obtained from integrated CBCT using Invisalign^® software and conventional CBCT images. Methods: In this cross-sectional study, we included 45 adult patients (mean age 29.53 ± 7.66 years). Alveolar bone width was measured at the crestal, mid-root, and apical levels and the root length, maxillary and mandibular outer and inner transverse widths, and symphyseal width were assessed. Intra-examiner reliability was assessed using intraclass correlation coefficients. Paired t-tests were used to compare the two methods, and Bland–Altman analysis was used to evaluate agreement. Results: Intra-examiner reliability demonstrated excellent agreement for all measurements (intraclass correlation coefficients > 0.90). Integrated CBCT models showed greater alveolar bone width, symphyseal width, and interbuccal transverse measurements than conventional CBCT, whereas the root length and interlingual transverse widths were underestimated. Most of the differences between the two methods were statistically significant (p < 0.05), although 5 of the 12 crestal measurements did not reach statistical significance (p > 0.05). For most measurements, the mean absolute differences were below 0.7 mm, except for root length, which showed larger discrepancies. Bland–Altman analysis revealed good agreement between the methods, with most observations lying within the 95% limits of agreement and no evident proportional bias. Conclusions: Most measured variables showed significant differences between integrated and conventional CBCT (p < 0.05), ranging from 0.7 mm to 1.3 mm; however, at the crestal level, 5/12 measurements showed no statistically significant difference between methods (p > 0.05). Full article

Bamboo served as an important writing support and raw material for tools and daily objects in ancient Asian societies, yet archaeological bamboo remains far less systematically reviewed than archaeological wood, paper, and parchment. Owing to its lignocellulosic composition and hierarchical vascular structure, archaeological bamboo is highly susceptible to hydrolysis, oxidation, microbial attack, and post-excavation dimensional instability. A structured search of the Web of Science Core Collection, completed in June 2026, identified 134 records for screening; after predefined eligibility assessment and exclusion of studies unrelated to archaeological or heritage bamboo, 75 studies were ultimately included and cited in this review. Unlike previous publications that address bamboo structure, degradation, or digital reconstruction separately, this review integrates archaeological context, graded anatomy, material deterioration, analytical characterization, treatment compatibility, inscription preservation, and computational restoration within a unified conservation-science framework. Evidence from Fourier transform infrared spectroscopy, Raman spectroscopy, scanning electron microscopy, X-ray diffraction, and thermal analysis is synthesized to link chemical loss with microstructural collapse and mechanical fragility. Excavation stabilization, dehydration, polymer consolidation, environmental control, artificial aging, digital reconstruction, and artificial intelligence-assisted fragment reassembly are critically evaluated. The available evidence indicates that bamboo-specific conservation research remains limited and that several treatments are still extrapolated from waterlogged wood or paper conservation. Major research gaps include standardized condition-assessment indicators, quantitative long-term treatment data, ink–substrate compatibility studies, and validated decision frameworks for archaeological bamboo conservation. Full article

Background and Objectives: Magnesium-based biomaterials are increasingly investigated due to their biodegradability, bone-like elastic modulus, and potential osteogenic properties. However, alloy composition may influence degradation behavior and local tissue response. This study aimed to assess and compare the local biocompatibility, degradation characteristics, and bone regenerative response induced by Mg-Zn-Mn-Ca and Mg-Nd-Y-Zr-Zn magnesium alloy powders. Materials and Methods: Two magnesium alloys were evaluated through physicochemical characterization, in vitro osteoblast cytocompatibility and antimicrobial assays, and in vivo testing in Sprague–Dawley rats. Standardized cylindrical defects were created in the medial femoral condyle and filled with either Mg-Nd-Y-Zr-Zn or Mg-Zn-Mn-Ca alloy powders, while untreated defects served as controls. Bone remodeling and defect healing were assessed by micro-computed tomography, allowing three-dimensional qualitative and quantitative evaluation of newly formed bone. Scanning electron microscopy was used to analyze surface morphology and degradation features, while histopathological examination assessed inflammation, osteogenesis, and tissue integration at the implant site. Results: Both magnesium alloy powders showed antimicrobial potency and elicited no cytotoxic effects in vitro, while animal studies revealed progressive biodegradation over time, associated with new bone formation within and around the defect area. Micro-CT analysis demonstrated active bone remodeling in experimental groups, with differences in bone distribution and defect-filling patterns between Mg-Nd-Y-Zr-Zn and Mg-Zn-Mn-Ca implants. SEM revealed alloy-specific degradation morphologies. Histological evaluation showed a moderate early inflammatory response that decreased at later time points, together with ongoing osteogenesis and favorable tissue integration. No severe local adverse reactions or persistent inflammation were observed. Conclusions: Both Mg-Nd-Y-Zr-Zn and Mg-Zn-Mn-Ca alloy powders were locally biocompatible and supported bone regeneration in a rat femoral condyle defect model. Differences in degradation behavior and tissue response emphasize the relevance of alloy composition in developing magnesium-based biomaterials for bone defect treatment. Full article