Search Results (330)

Introduction: Accurate nuclear segmentation is essential for the reliable diagnostic interpretation of fluorescence in situ hybridization (FISH) results. However, automated 2D digital algorithms often fail in samples with dense or overlapping nuclei, such as lymphomas, due to the loss of spatial depth information. Here, we tested if AI-based 3D nuclear segmentation can improve the accuracy, reproducibility, and diagnostic reliability of FISH reading in critical situations. Materials and Methods: Formalin-fixed follicular lymphoma sections were FISH-labeled for BCL2 gene rearrangements and digitally scanned in multilayer Z-stacks. The analytic performance in nuclear segmentation of the adaptive thresholding-based FISHQuant, and the freely accessible AI-based NucleAIzer, StarDist, and Cellpose algorithms, were compared to the eye control-based traditional FISH testing, primarily focusing on nuclear segmentation. Results: We revealed that the Cellpose algorithm showed limited sensitivity to low-intensity signals and the adaptive thresholding 2D segmentation, and FISHQuant struggled to resolve densely packed nuclei, occasionally underestimating their counts. In contrast, 3D segmentation across focal planes significantly improved the nuclear separation and signal localization. AI-driven 3D models, especially NucleAIzer and StarDist, showed improved precision, lower variance (VP/VS ≈ 0.96), and improved gene spot correlation (r > 0.82) across multiple focal planes. The similar average number of gene spots per cell nuclei in the AI-based analyses as the eye control counting, despite the elevated number of cell nuclei found with AI, validated the AI nuclear segmentation results. Conclusions: Inaccurate segmentation limits automated diagnostic FISH signal evaluation. Deep learning 3D approaches, particularly NucleAIzer and StarDist, may overcome thresholding and 2D constraints and improve the consistency of nuclear detection, resulting in better classification of pathogenetic gene aberrations with automated workflows in digital pathology. Full article

Visual information shapes spatial perception and body representation in human augmentation. However, the perceptual consequences of viewpoint-height changes produced by sensor–display geometry are not well understood. To address this gap, we developed an interface that maps a waist-mounted stereo fisheye camera to an eye-level viewpoint on a head-mounted display in real time. Geometric and timing calibration kept latency low enough to preserve a sense of agency and enable stable untethered walking. In a within-subject study comparing head- and waist-level viewpoints, participants approached adjustable gaps, rated passability confidence (1–7), and attempted passage when confident. We also recorded walking speed and assessed post-task body representation using a questionnaire. High gaps were judged passable and low gaps were not, irrespective of viewpoint. At the middle gap, confidence decreased with a head-level viewpoint and increased with a waist-level viewpoint, and walking speed decreased when a waist-level viewpoint was combined with a chest-height gap, consistent with added caution near the decision boundary. Body image reports most often indicated a lowered head position relative to the torso, consistent with visually driven rescaling rather than morphological change. These findings show that a waist-mounted interface for mobile viewpoint-height transformation can reliably shift spatial perception. Full article

The article addresses the issue of panoramic photogrammetry for the reconstruction of interior spaces. Such environments often present challenges, including poor lighting conditions and surfaces with variable texture for photogrammetric scanning. In this case study, we reconstruct the interior spaces of the historical house of Samuel Mikovíni, which represents these unfavorable conditions. The 3D reconstruction of interior spaces is performed using the Ricoh Theta Z1 spherical camera (Ricoh Company, Ltd.; Tokyo, Japan) in six variants, each employing a different number of images and different camera networks. Scale is introduced into the reconstructions based on significant dimensions measured with a measuring tape. A comparison is carried out using a point cloud obtained from terrestrial laser scanning and difference point clouds are generated for each variant. Based on the results, reconstructions produced from a reduced number of spherical images can serve as a basic source for simple documentation with accuracy up to 0.15 m. When the number of spherical images is increased and images from different height levels are included, the reconstruction accuracy improves markedly, achieving positional accuracy of up to 0.05 m, even in areas affected by poor lighting conditions or low-texture surfaces. The results confirm that for interior reconstruction, a higher number of images not only increases the density of the reconstructed point cloud but also enhances its positional accuracy. Full article

Semantic segmentation of crowdsourced street-level imagery plays a critical role in urban analytics by enabling pixel-wise understanding of urban scenes for applications such as walkability scoring, environmental comfort evaluation, and urban planning, where robustness to geometric transformations and projection-induced symmetry variations is essential. This study presents a comparative evaluation of two primary families of semantic segmentation models: transformer-based models (SegFormer and Mask2Former) and prompt-based models (CLIPSeg, LangSAM, and SAM+CLIP). The evaluation is conducted on images with varying geometric properties, including normal perspective, fisheye distortion, and panoramic format, representing different forms of projection symmetry and symmetry-breaking transformations, using data from Google Street View and Mapillary. Each model is evaluated on a unified benchmark with pixel-level annotations for key urban classes, including road, building, sky, vegetation, and additional elements grouped under the “Other” class. Segmentation performance is assessed through metric-based, statistical, and visual evaluations, with mean Intersection over Union (mIoU) and pixel accuracy serving as the primary metrics. Results show that LangSAM demonstrates strong robustness across different image formats, with mIoU scores of 64.48% on fisheye images, 85.78% on normal perspective images, and 96.07% on panoramic images, indicating strong semantic consistency under projection-induced symmetry variations. Among transformer-based models, SegFormer proves to be the most reliable, attains higher accuracy on fisheye and normal perspective images among all models, with mean IoU scores of 72.21%, 94.92%, and 75.13% on fisheye, normal, and panoramic imagery, respectively. LangSAM not only demonstrates robustness across different projection geometries but also delivers the lowest segmentation error, consistently identifying the correct class for corresponding objects. In contrast, CLIPSeg remains the weakest prompt-based model, with mIoU scores of 77.60% on normal images, 59.33% on panoramic images, and a substantial drop to 59.33% on fisheye imagery, reflecting sensitivity to projection-related symmetry distortions. Full article

Since the 1960s, Grassy Narrows First Nation, Canada, has been exposed to methylmercury (MeHg) from fish consumption following Hg discharge from a chloralkali plant. Prenatal exposure to MeHg is known to affect the neurodevelopment of fetuses and the retina is sensitive to neurodevelopmental damage. The multidisciplinary, cross-sectional Niibin study, developed with Grassy Narrows First Nations, included visual examinations with retinal evaluation using optical coherence tomography (OCT). The present analyses focused on the 59 participants (116 eyes) with umbilical cord Hg measurements, sampled between 1971 and 1992. Associations between cord blood Hg and retinal thickness layers surrounding the optic nerve head (RNFL) and inner macula (GC-IPL) were examined using mixed-effect models. Higher cord blood Hg was significantly associated with reduced thickness of GC-IPL layers across all macular sectors; less pronounced associations were observed for RNFL. A qualitative clinical assessment of the OCT results showed that persons with cord blood Hg concentrations ≥ 5.8 µg/L were more likely to present bilateral abnormal retinal thinning (OR = 3.51; [95% CI: 1.06–11.53]). These findings suggest that, in this Indigenous community, prenatal MeHg exposure may have enduring effects on retinal thickness and underline the importance of OCT technology in providing tailored eye care. Full article

A series of high-cycle rotating-bending fatigue tests was conducted on H13 steel produced by electroslag remelting (ESR) and by vacuum induction melting followed by vacuum arc remelting (VIM+VAR). At 10⁷ cycles, the fatigue strength of VIM+VAR steel was 1040 MPa, which is greater than the 967 MPa of ESR steel. A metallographic analysis was conducted to compare the structure and grain size of the two steels. The results indicated that while the two steels were similar, ESR steel contained a greater number of larger inclusions and carbides. The mean inclusion size in VIM+VAR steel was approximately 55% of that in ESR steel, and the maximum inclusion size was around 44%. Notwithstanding this finding, the fatigue strength of VIM+VAR steel was found to be approximately 7.5% higher. Scanning electron microscopy of fracture surfaces revealed that the primary cause of crack initiation was predominantly oxides or oxide-sulfide composites. The measurements obtained for inclusion size, fisheye diameter, and crack propagation length indicated that the fatigue life of the material is governed primarily by the applied stress and the size of the inclusion. The presence of larger inclusions has been demonstrated to reduce the crack-propagation stage and decrease the steel’s tolerance to defects, thereby reducing fatigue life and endurance limit. The researchers derived formulae relating inclusion size to stress intensity factor and fatigue life by utilizing the Paris law. These equations ·the fatigue-fracture mechanism and provided a basis for predicting the rotating-bending fatigue life of H13 steel. Full article

This review provides an in-depth look at the key process limitations and (structural) defects encountered in the production of polymer films via film blowing extrusion technology. Film blowing is the most widely used method for producing plastic films across various industries, with its increasing demand driven by flexible packaging needs. Overcoming the challenges of this complex production process is essential for ensuring high quality and meeting the growing demand for modern applications, taking into account polymer circularity. In the first part of this paper, the focus is on conventional films, generally polyolefin single-layer films. Common defects such as bubble instability, gauge variations, wrinkles, melt fractures, optical defects, blocking, and surface imperfections like fish eyes are discussed. The most important causes behind these issues are elaborated on, including various molecular and processing parameters, with this paper also offering practical mitigating strategies. In the second part, the specific process limitations and defect types associated with emerging sustainable film technology are focused on, covering films made from recycled materials, biodegradable polymers, polymer blends, and multilayer and machine-direction oriented (MDO) films. While these innovative films offer significant advantages in terms of sustainability and property enhancement, they also present additional points of attention. Also, effective mitigation strategies for addressing these technical issues are incorporated. Overall, this study provides a comprehensive review of film blowing defects, contributing to improved process control, reduced waste, and the production of high-quality films that meet modern requirements. By identifying the root causes of common defects and discussing viable solutions, this review plays a key role in advancing the efficiency, consistency, and sustainability of film blowing technology by presenting a combined experimental and modelling approach that can be used in future work. Full article

Accurate sky–obstacle segmentation in hemispherical fisheye imagery is essential for solar irradiance forecasting, photovoltaic system design, and environmental monitoring. However, existing methods often rely on expensive all-sky imagers and region-specific training data, produce coarse sky–obstacle boundaries, and ignore the optical properties of fisheye lenses. We propose a low-cost segmentation framework designed for fisheye imagery that combines synthetic data generation, lens-aware augmentation, and a hybrid deep-learning pipeline. Synthetic fisheye training images are created from publicly available street-view panoramas to cover diverse environments without dedicated hardware, and lens-aware augmentations model fisheye projection and photometric effects to improve robustness across devices. On this dataset, we train a convolutional neural network (CNN) and refine its output with gradient-boosted decision trees (GBDT) to sharpen sky–obstacle boundaries. The method is evaluated on real fisheye images captured with smartphones and low-cost clip-on lenses across multiple sites, achieving an Intersection over Union (IoU) of 96.63% and an F1 score of 98.29%, along with high boundary accuracy. An additional evaluation on an external panoramic baseline dataset confirms strong cross-dataset generalization. Together, these results show that the proposed framework enables accurate, low-cost, and widely deployable hemispherical sky segmentation for practical solar and environmental imaging applications. Full article

Traditional morphometric analysis of Large Yellow Croaker (Larimichthys crocea) relies heavily on manual dissection, which is time-consuming, labor-intensive, and prone to subjectivity. To address these limitations, we propose an automated quantitative approach based on deep-learning–driven instance segmentation. A dataset comprising 160 X-ray images of L. crocea was established, encompassing five anatomical categories: whole fish, air bladder, spine, eyes, and otoliths. Building upon the baseline YOLOv11-Seg model, we integrated a lightweight Convolutional Block Attention Module (CBAM) to construct an improved YOLOv11-CBAM network, thereby enhancing segmentation accuracy for complex backgrounds and fine-grained targets. Experimental results demonstrated that the modified model achieved superior performance in both mAP50 and mAP50–95 compared with the baseline, with particularly notable improvements in the segmentation of small-scale structures such as the air bladder and spine. By introducing coin-based calibration, pixel counts were converted into absolute areas and relative proportions. The measured area ratios of the air bladder, otoliths, eyes, and spine were 7.72%, 0.59%, 2.20%, and 8.48%, respectively, with standard deviations remaining within acceptable ranges, thus validating the robustness of the proposed method. Collectively, this study establishes a standardized, efficient, and non-destructive workflow for X-ray image-based morphometric analysis, providing practical applications for aquaculture management, germplasm conservation, and fundamental biological research. Full article

High-yielding lactating cows generate considerable internal heat, making thermoregulation challenging in warm conditions. Traditionally, sprinkler systems have cooled dairy cows by spraying water droplets onto their skin to aid heat dissipation, especially when used with fans. This study explores the benefits of AI-assisted monitoring of water usage for cooling dairy cows, aiming to optimize water consumption and enhance sustainability. An object detection model, trained with 200 random images from a fisheye security camera installed above pens of dairy cows in a dairy farm, was used to detect the presence or absence of cows in headgate sections to guide water sprinkler activity. According to the object detection model, the implementation of AI-assisted detection of cows’ presence or absence in headgates with an accuracy of 0.924 has the potential to save up to 75 percent of water annually for cooling cows. Additionally, the model can detect cows’ behavior patterns regarding location in the pens depending on the occurrence of heat stress. The implementation of AI-powered detection systems in dairy farms has been proven to enhance sustainability and significantly reduce expenses by curbing the excessive use of water. Full article

Under climate change, the importance of ecosystem monitoring has been repeatedly emphasized over the past decades. Leaf Area Index (LAI), a key ecosystem variable linking the atmosphere and rhizosphere, has been widely studied through various LAI measurement methods. As satellite-based LAI products continue to advance, the demand for extensive and periodic in situ LAI observations has also increased. In this study, we evaluated the combinations of binarization techniques and temporal filtering to reduce variability in an automatic in situ LAI observation network using fisheye lens imagery, which was established by the National Institute of Forest Science (NIFoS). Compared to the widely used methods such as Otsu thresholding (Otsu) and K-means clustering (K-means), the deep learning (DL) method showed more stable LAI time series under field conditions. Under different illumination conditions, mean LAI values fluctuated significantly—from 0.89 to 3.15—depending on image acquisition time. Furthermore, sixteen temporal filtering methods were tested to identify a reasonable range of LAI values, with optimal post-processing strategies suggested: seven-day moving average for maximum LAI (LAI different range among filtering methods −6.1~−1.5) and a three-day moving average excluding rainy days for minimum LAI (LAI different range among filtering methods 0~0.9). This study highlights uncertainties in canopy classification methods, the effects of acquisition timing and lighting, and the necessity of outlier filtering in automatic LAI networks. Despite these challenges, the need for automated LAI observation system is growing, particularly in complex and fragmented forests such as those found in South Korea. Full article

Calibrating cameras accurately requires the identification of projection and distortion models that effectively account for lens-specific deviations. Conventional formulations, like the pinhole model or radial–tangential corrections, often struggle to represent the asymmetric and nonlinear distortions encountered in complex environments such as autonomous navigation, robotics, and immersive imaging. Although neural methods offer greater adaptability, they demand extensive training data, are computationally intensive, and often lack transparency. This work introduces a symbolic model discovery framework guided by physical knowledge, where symbolic regression and genetic programming (GP) are used in tandem to identify calibration models tailored to specific optical behaviors. The approach incorporates a broad class of known distortion models, including Brown–Conrady, Mei–Rives, Kannala–Brandt, and double-sphere, as modular components, while remaining extensible to any predefined or domain-specific formulation. Embedding these models directly into the symbolic search process constrains the solution space, enabling efficient parameter fitting and robust model selection without overfitting. Through empirical evaluation across a variety of lens types, including fisheye, omnidirectional, catadioptric, and traditional cameras, we show that our method produces results on par with or surpassing those of established calibration techniques. The outcome is a flexible, interpretable, and resource-efficient alternative suitable for deployment scenarios where calibration data are scarce or computational resources are constrained. Full article

The ciliary marginal zone (CMZ) is a region in the peripheral-most retina that displays ongoing retinogenesis during growth and expansion of the eye in adulthood. While there is evidence that this capacity also exists in birds and mammals, it is far more robust in fish and amphibians. The process of CMZ retinogenesis is essentially equivalent to that seen early in the central retina; however, its regulation is not fully understood. In a previous study, we attempted to uncover novel regulatory genes by using a forward genetics screen in zebrafish, looking for recessive CMZ mutants. One of the mutants found was called no marginal zone (nmz). The nmz mutant showed relatively normal central retina development, but a lack of cells in the CMZ by 5 days post fertilization (dpf). Mapping, genomic sequencing, and complementation analysis using a second mutant line (m863) isolated in another laboratory showed that a mutation in damage-specific DNA binding protein-1 (ddb-1) gene underlies the phenotype seen in nmz. BrdU labeling suggested that later expansion and differentiation of CMZ retinal progenitors is more affected by ddb-1 loss than the earlier process of stem cell asymmetric division. As was seen for the m863 mutant and in other studies with mice, one profound effect of ddb-1 loss in nmz was the upregulation in expression of tp53 and several of its downstream effectors. Several important genes important in CMZ retinogenesis are also downregulated in the nmz mutant. The change in gene expression would suggest that ddb-1 loss leads to increased cell cycle disruption and apoptosis at the expense of CMZ retinogenesis. While homozygosity is lethal, heterozygous fish appear to be completely normal in morphology, visual function, and behavior. Full article

The large-eye dentex (Dentex macrophthalmus) is a relatively small sparid fish with increasing potential as a supplementary fishery resource in the Mediterranean Sea, particularly as traditional stocks face overexploitation. Despite its widespread distribution, biological data on this species, especially from Greek waters, remain scarce. This study presents the first comprehensive biological assessment of D. macrophthalmus in the Pagasitikos Gulf, focusing on population structure, growth, mortality, and the application of machine learning (ML) for age prediction. A total of 305 individuals were collected, revealing a female-biased sex ratio and negative allometric growth in both somatic and otolith dimensions. The von Bertalanffy growth parameters indicated a slow growth rate (k = 0.16 year⁻¹), with an estimated asymptotic length (L∞) of 25.97 cm. The population was found to be underexploited (E = 0.41), suggesting resilience to current fishing pressure. Stepwise regression and ML models were employed to predict age from otolith morphometrics. A linear model identified otolith weight and aspect ratio as the most significant predictors of age (R² = 0.8). Among the ML algorithms tested, the Neural Network model achieved the highest performance (R² = 0.764, MAPE = 14.10%), demonstrating its potential for accurate and efficient age estimation. These findings provide crucial baseline data for the sustainable management of D. macrophthalmus and highlight the value of integrating advanced ML techniques into fisheries biology. Full article

This account is an exploration of concepts exploring the widespread damage to liquid metals caused by poor current liquid metal handling and casting technology. The defects introduced in the liquid state are suggested to affect many properties of our engineering metals, especially tensile elongation and Charpy toughness, but also time-dependent degradation processes, which can result in failure by fracture, and which can be significantly aided by hydrogen, leading to hydrogen embrittlement (HE), and invasive corrosion, leading to stress corrosion cracking (SCC). The new phenomenon of ‘precipitation cleavage’ is introduced, explaining the sensitization of alloys by certain heat treatments. Direct visual evidence for precipitation cleavage is provided by the previously unexplained phenomenon of ‘fisheyes’ observed frequently on the fracture surfaces of steels, and more recently also in light alloys. Full article