Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Image Segmentation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 577 records · Page 32

Fully convolutional neural network for removing background in noisy images of uranium bearing particles

A fully convolutional neural network (FCN) was developed to supersede automatic or manual thresholding algorithms used for tabulating SIMS particle search data. The FCN was designed to perform a binary classification of pixels in each image belonging to a particle or not, thereby effectively removing background signal without manually or automatically determining an intensity threshold. Using 8,000 images from 28 different particle screening analyses, the FCN was trained to accurately predict pixels belonging to a particle with near 99% accuracy. Background eliminated images were then segmented using a watershed technique in order to determine isotopic ratios of particles. A comparison of the isotopic distributions of an independent data set segmented using the neural network, compared to a commercially available automated particle measurement (APM) program developed by CAMECA, highlighted the necessity for effective background removal to ensure that resulting particle identification is not only accurate, but preserves valuable signal that could be lost due to improper segmentation. The FCN approach improves the robustness of current state-of-the-art particle searching algorithms by reducing user input biases, resulting in an improved absolute signal per particle and decreased uncertainty of the determined isotope ratios.

Tarolli, Jay G. (ORCID:0000000307369057)↗

Improving reproducibility in synchrotron tomography using implementation-adapted filters

For reconstructing large tomographic datasets fast, filtered backprojection-type or Fourier-based algorithms are still the method of choice, as they have been for decades. These robust and computationally efficient algorithms have been integrated in a broad range of software packages. The continuous mathematical formulas used for image reconstruction in such algorithms are unambiguous. However, variations in discretization and interpolation result in quantitative differences between reconstructed images, and corresponding segmentations, obtained from different software. This hinders reproducibility of experimental results, making it difficult to ensure that results and conclusions from experiments can be reproduced at different facilities or using different software. In this paper, a way to reduce such differences by optimizing the filter used in analytical algorithms is proposed. These filters can be computed using a wrapper routine around a black-box implementation of a reconstruction algorithm, and lead to quantitatively similar reconstructions. Use cases for this approach are demonstrated by computing implementation-adapted filters for several open-source implementations and applying them to simulated phantoms and real-world data acquired at the synchrotron. Our contribution to a reproducible reconstruction step forms a building block towards a fully reproducible synchrotron tomography data processing pipeline.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Exploring Classification of Topological Priors With Machine Learning for Feature Extraction

In many scientific endeavors, increasingly abstract representations of data allow for new interpretive methodologies and conceptualization of phenomena. For example, moving from raw imaged pixels to segmented and reconstructed objects allows researchers new insights and means to direct their studies toward relevant areas. Thus, the development of new and improved methods for segmentation remains an active area of research. With advances in machine learning and neural networks, scientists have been focused on employing deep neural networks such as U-Net to obtain pixel-level segmentations, namely, defining associations between pixels and corresponding/referent objects and gathering those objects afterward. Topological analysis, such as the use of the Morse-Smale complex to encode regions of uniform gradient flow behavior, offers an alternative approach: first, create geometric priors, and then apply machine learning to classify. This approach is empirically motivated since phenomena of interest often appear as subsets of topological priors in many applications. Using topological elements not only reduces the learning space but also introduces the ability to use learnable geometries and connectivity to aid the classification of the segmentation target. Here, in this article, we describe an approach to creating learnable topological elements, explore the application of ML techniques to classification tasks in a number of areas, and demonstrate this approach as a viable alternative to pixel-level classification, with similar accuracy, improved execution time, and requiring marginal training data.

97 MATHEMATICS AND COMPUTING↗

An algorithm for efficient metal artifact reductions in permanent seed implants

Purpose: In permanent seed implants, 60 to more than 100 small metal capsules are inserted in the prostate, creating artifacts in x-ray computed tomography (CT) imaging. The goal of this work is to develop an automatic method for metal artifact reduction (MAR) from small objects such as brachytherapy seeds for clinical applications. Methods: The approach for MAR is based on the interpolation of missing projections by directly using raw helical CT data (sinogram). First, an initial image is reconstructed from the raw CT data. Then, the metal objects segmented from the reconstructed image are reprojected back into the sinogram space to produce a metal-only sinogram. The Steger method is used to determine precisely the position and edges of the seed traces in the raw CT data. By combining the use of Steger detection and reprojections, the missing projections are detected and replaced by interpolation of non-missing neighboring projections. Results: In both phantom experiments and patient studies, the missing projections have been detected successfully and the artifacts caused by metallic objects have been substantially reduced. The performance of the algorithm has been quantified by comparing the uniformity between the uncorrected and the corrected phantom images. The results of the artifact reduction algorithm are indistinguishable from the true background value. Conclusions: An efficient algorithm for MAR in seed brachytherapy was developed. The test results obtained using raw helical CT data for both phantom and clinical cases have demonstrated that the proposed MAR method is capable of accurately detecting and correcting artifacts caused by a large number of very small metal objects (seeds) in sinogram space. This should enable a more accurate use of advanced brachytherapy dose calculations, such as Monte Carlo simulations.

60 APPLIED LIFE SCIENCES↗

Evaluation of S-values and dose distributions for {sup 90}Y, {sup 131}I, {sup 166}Ho, and {sup 188}Re in seven lobes of the rat liver

Purpose: Rats have been widely used in radionuclide therapy research for the treatment of hepatocellular carcinoma (HCC). This has created the need to assess rat liver absorbed radiation dose. In most dose estimation studies, the rat liver is considered as a homogeneous integrated target organ with a tissue composition assumed to be similar to that of human liver tissue. However, the rat liver is composed of several lobes having different anatomical and chemical characteristics. To assess the overall impact on rat liver dose calculation, the authors use a new voxel-based rat model with identified suborgan regions of the liver. Methods: The liver in the original cryosectional color images was manually segmented into seven individual lobes and subsequently integrated into a voxel-based computational rat model. Photon and electron particle transport was simulated using the MCNPX Monte Carlo code to calculate absorbed fractions and S-values for {sup 90}Y, {sup 131}I, {sup 166}Ho, and {sup 188}Re for the seven liver lobes. The effect of chemical composition on organ-specific absorbed dose was investigated by changing the chemical composition of the voxel filling liver material. Radionuclide-specific absorbed doses at the voxel level were further assessed for a small spherical hepatic tumor. Results: The self-absorbed dose for different liver lobes varied depending on their respective masses. A maximum difference of 3.5% was observed for the liver self-absorbed fraction between rat and human tissues for photon energies below 100 keV. {sup 166}Ho and {sup 188}Re produce a uniformly distributed high dose in the tumor and relatively low absorbed dose for surrounding tissues. Conclusions: The authors evaluated rat liver radiation doses from various radionuclides used in HCC treatments using a realistic computational rat model. This work contributes to a better understanding of all aspects influencing radiation transport in organ-specific radiation dose evaluation for preclinical therapy studies, from tissue composition to organ morphology and activity distribution.

61 RADIATION PROTECTION AND DOSIMETRY↗

SU-F-T-65: AutomaticTreatment Planning for High-Dose Rate (HDR) Brachytherapy with a VaginalCylinder Applicator

Purpose: High dose rate (HDR) brachytherapy treatment planning is conventionally performed in a manual fashion. Yet it is highly desirable to perform computerized automated planning to improve treatment planning efficiency, eliminate human errors, and reduce plan quality variation. The goal of this research is to develop an automatic treatment planning tool for HDR brachytherapy with a cylinder applicator for vaginal cancer. Methods: After inserting the cylinder applicator into the patient, a CT scan was acquired and was loaded to an in-house developed treatment planning software. The cylinder applicator was automatically segmented using image-processing techniques. CTV was generated based on user-specified treatment depth and length. Locations of relevant points (apex point, prescription point, and vaginal surface point), central applicator channel coordinates, and dwell positions were determined according to their geometric relations with the applicator. Dwell time was computed through an inverse optimization process. The planning information was written into DICOM-RT plan and structure files to transfer the automatically generated plan to a commercial treatment planning system for plan verification and delivery. Results: We have tested the system retrospectively in nine patients treated with vaginal cylinder applicator. These cases were selected with different treatment prescriptions, lengths, depths, and cylinder diameters to represent a large patient population. Our system was able to generate treatment plans for these cases with clinically acceptable quality. Computation time varied from 3–6 min. Conclusion: We have developed a system to perform automated treatment planning for HDR brachytherapy with a cylinder applicator. Such a novel system has greatly improved treatment planning efficiency and reduced plan quality variation. It also served as a testbed to demonstrate the feasibility of automatic HDR treatment planning for more complicated cases.

60 APPLIED LIFE SCIENCES↗

Train Like a (Var)Pro: Efficient Training of Neural Networks with Variable Projection

Deep neural networks (DNNs) have achieved state-of-the-art performance across a variety of traditional machine learning tasks, e.g., speech recognition, image classification, and segmentation. The ability of DNNs to efficiently approximate high-dimensional functions has also motivated their use in scientific applications, e.g., to solve partial differential equations and to generate surrogate models. In this paper, we consider the supervised training of DNNs, which arises in many of the above applications. We focus on the central problem of optimizing the weights of the given DNN such that it accurately approximates the relation between observed input and target data. Devising effective solvers for this optimization problem is notoriously challenging due to the large number of weights, nonconvexity, data sparsity, and nontrivial choice of hyperparameters. To solve the optimization problem more efficiently, we propose the use of variable projection (VarPro), a method originally designed for separable nonlinear least-squares problems. Our main contribution is the Gauss--Newton VarPro method (GNvpro) that extends the reach of the VarPro idea to nonquadratic objective functions, most notably cross-entropy loss functions arising in classification. These extensions make GNvpro applicable to all training problems that involve a DNN whose last layer is an affine mapping, which is common in many state-of-the-art architectures. In our four numerical experiments from surrogate modeling, segmentation, and classification, GNvpro solves the optimization problem more efficiently than commonly used stochastic gradient descent (SGD) schemes. Finally, GNvpro finds solutions that generalize well, and in all but one example better than well-tuned SGD methods, to unseen data points.

97 MATHEMATICS AND COMPUTING↗

OReole-FM: successes and challenges toward billion-parameter foundation models for high-resolution satellite imagery

While the pretraining of Foundation Models (FMs) for remote sensing (RS) imagery is on the rise, models remain restricted to a few hundred million parameters. Scaling models to billions of parameters has been shown to yield unprecedented benefits including emergent abilities, but requires data scaling and computing resources typically not available outside industry R&D labs. In this work, we pair high-performance computing resources including Frontier supercomputer, America's first exascale system, and high-resolution optical RS data to pretrain billion-scale FMs. Our study assesses performance of different pretrained variants of vision Transformers across image classification, semantic segmentation and object detection benchmarks, which highlight the importance of data scaling for effective model scaling. Moreover, we discuss construction of a novel TIU pretraining dataset, model initialization, with data and pretrained models intended for public release. By discussing technical challenges and details often lacking in the related literature, this work is intended to offer best practices to the geospatial community toward efficient training and benchmarking of larger FMs.

Ambrozio Dias, Philipe↗

VoroClust

SAND2025-11465O VoroClust, also known as Voronoi Clustering, is a fast, density-based unsupervised clustering algorithm applicable to high-resolution and high-dimensional data. It operates as quickly as distance-based clustering methods while effectively capturing complex regional geometries, matching the performance of current density-based methods. VoroClust employs a data-centered sphere cover to reduce computational demands while preserving data topology. It propagates clusters outward from local density peaks. Although supervised machine learning is powerful for applications like image classification and segmentation, it requires comprehensive, consistent datasets, which many applications lack. Unsupervised clustering algorithms analyze the structure of each dataset rather than relying on similarities with other examples, making them well-suited for practical applications with insufficient or inappropriate data for supervised learning. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Ebeida, Mohamed [Sandia National Lab. (SNL-CA), Li↗

Quantifying Mesoscale Neuroanatomy Using X-Ray Microtomography

Methods for resolving the three-dimensional (3D) microstructure of the brain typically start by thinly slicing and staining the brain, followed by imaging numerous individual sections with visible light photons or electrons. In contrast, X-rays can be used to image thick samples, providing a rapid approach for producing large 3D brain maps without sectioning. Here we demonstrate the use of synchrotron X-ray microtomography (µCT) for producing mesoscale (~1 µm 3 resolution) brain maps from millimeter-scale volumes of mouse brain. We introduce a pipeline for µCT-based brain mapping that develops and integrates methods for sample preparation, imaging, and automated segmentation of cells, blood vessels, and myelinated axons, in addition to statistical analyses of these brain structures. Our results demonstrate that X-ray tomography achieves rapid quantification of large brain volumes, complementing other brain mapping and connectomics efforts.

60 APPLIED LIFE SCIENCES↗

Sister Rod Destructive Examinations (FY23) Appendix E2: Four-Point Bend Fractured Specimen Images

This report documents work performed under the Spent Fuel and Waste Disposition’s Spent Fuel and Waste Science and Technology program for the US Department of Energy (DOE) Office of Nuclear Energy (NE). This work was performed to fulfill Level 2 Milestone M2SF-24OR010201024, “FY23 ORNL Testing on Sibling Pins,” within work package SF-24OR01020102 and is an update to the work reported in M2SF-23OR010201024, M2SF-22OR010201047, M2SF-21OR010201032, M2SF-19ORO010201026, and M2SF-19OR010201028. As a part of the DOE NE High Burnup Spent Fuel Data Project, Oak Ridge National Laboratory (ORNL) is performing destructive examinations (DEs) of high burnup (HBU) (>45 GWd/MTU) spent nuclear fuel (SNF) rods from the North Anna Nuclear Power Station operated by Dominion Energy. The SNF rods, called sister rods or sibling rods, are all HBU and include four different kinds of fuel rod cladding: standard Zircaloy-4 (Zirc-4), low-tin Zirc-4, ZIRLO, and M5. The DEs are being conducted to obtain a baseline of the HBU rods’ condition before they are sent to dry storage and are focused on gaining an understanding of overall SNF rod strength and durability. Composite fuel and defueled cladding will be tested to derive material properties. Although the data generated can be used for multiple purposes, one primary goal for obtaining the post-irradiation examination data and the associated measured mechanical properties is to support SNF dry storage licensing and relicensing activities by (1) addressing identified knowledge gaps and (2) enhancing the technical basis for post-storage transportation, handling, and subsequent disposition. This report documents the status of the ORNL Phase 1 DE activities related to the mechanical testing of selected sister rods in Phase 1 of the sister rod test program. This appendix provides images of the segments broken in four-point bending (4PB).

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Effect of Off-Target Binding on 18 F-Flortaucipir Variability in Healthy Controls Across the Life Span

Measuring early tau accumulation is important in studying aging and Alzheimer disease and is only as accurate as the signal-to-noise ratio of the tracer. Along with aggregated tau in the form of neurofibrillary tangles, 18 F-flortaucipir has been reported to bind to neuromelanin, monoamine oxidase, calcifications, iron, leptomeningeal melanocytes, and microhemorrages. Although 18 F-flortaucipir successfully differentiates healthy controls (HCs) from subjects with Alzheimer disease, variability exists in the cortical signal in amyloid-negative HCs. We aimed to explore the relationship between off-target binding signal and variability in the cortical signal in HCs. Methods: Subjects ($n$ = 139) received 11 C-Pittsburgh compound B (PIB) and 18 F-flortaucipir PET scans and a magnetization-prepared rapid gradient echo MRI scan. PET frames were realigned and coregistered to the MR images, which were segmented using FreeSurfer. In amyloid-negative HCs ($n$ = 90; age range, 21-94 y), 7 nonspecific or off-target binding regions were considered: caudate, pallidum, putamen, thalamus, cerebellar white matter, hemispheric white matter, and choroid plexus. These regions of interest were assigned to 3 similarly behaving groups using principle components analysis, exploratory factor analysis, and Pearson correlations for caudate, putamen, and pallidum (also correlated with age); thalamus and white matter; and choroid plexus. In amyloid-negative HCs with 11 C-PIB and 18 F-flortaucipir scans, correlations were calculated between white and gray matter before and after partial-volume correction. Results: The correlation between white and gray matter disappeared after partial-volume correction in 11 C-PIB (r 2 = 0) but persisted for 18 F-flortaucipir (r 2 = 0.27), demonstrating that the correlation between white and gray matter signal in 18 F-flortaucipir is not solely due to partial-volume effects. A linear regression showed that off-target signal from putamen and thalamus together explained 64% of the variability in the cortical signal in amyloid-negative HCs (not seen in amyloid-positive HCs). Variability in amyloid-negative HCs but not amyloid-positive HCs correlated with white matter signal (unrelated to partial-volume effects) and age-related off-target signal (possibly related to iron load). Conclusion: The noise in the 18 F-flortaucipir measurement could pose challenges when studying early tau accumulation.

18F-flortaucipir PET↗

Image registration method for medical image sequences

Image registration of low contrast image sequences is provided. In one aspect, a desired region of an image is automatically segmented and only the desired region is registered. Active contours and adaptive thresholding of intensity or edge information may be used to segment the desired regions. A transform function is defined to register the segmented region, and sub-pixel information may be determined using one or more interpolation methods.

60 APPLIED LIFE SCIENCES↗

Image registration method for medical image sequences

Image registration of low contrast image sequences is provided. In one aspect, a desired region of an image is automatically segmented and only the desired region is registered. Active contours and adaptive thresholding of intensity or edge information may be used to segment the desired regions. A transform function is defined to register the segmented region, and sub-pixel information may be determined using one or more interpolation methods.

97 MATHEMATICS AND COMPUTING↗

MORPHOMETRIC SUBTYPING FOR A PANEL OF BREAST CANCER CELL LINES

A panel of cell lines of diverse molecular background offers an improved model system for high-content screening, comparative analysis, and cell systems biology. A computational pipeline has been developed to collect images from cell-based assays, segment individual cells and colonies, represent segmented objects in a multidimensional space, and cluster them for identifying distinct subpopulations. While each segmentation strategy can vary for different imaging assays, representation and subpopulation analysis share a common thread. Application of this pipeline to a library of 41 breast cancer cell lines is demonstrated. These cell lines are grown in 2D and imaged through immunofluorescence microscopy. Subpopulations in this panel are identified and shown to correlate with previous subtyping literature that was derived from transcript data.

60 APPLIED LIFE SCIENCES↗

Nonlocal Means Denoising of Self-Gated and k-Space Sorted 4-Dimensional Magnetic Resonance Imaging Using Block-Matching and 3-Dimensional Filtering: Implications for Pancreatic Tumor Registration and Segmentation

Purpose: To denoise self-gated k-space sorted 4-dimensional magnetic resonance imaging (SG-KS-4D-MRI) by applying a nonlocal means denoising filter, block-matching and 3-dimensional filtering (BM3D), to test its impact on the accuracy of 4D image deformable registration and automated tumor segmentation for pancreatic cancer patients. Methods and Materials: Nine patients with pancreatic cancer and abdominal SG-KS-4D-MRI were included in the study. Block-matching and 3D filtering was adapted to search in the axial slices/frames adjacent to the reference image patch in the spatial and temporal domains. The patches with high similarity to the reference patch were used to collectively denoise the 4D-MRI image. The pancreas tumor was manually contoured on the first end-of-exhalation phase for both the raw and the denoised 4D-MRI. B-spline deformable registration was applied to the subsequent phases for contour propagation. The consistency of tumor volume defined by the standard deviation of gross tumor volumes from 10 breathing phases (σ-GTV), tumor motion trajectories in 3 cardinal motion planes, 4D-MRI imaging noise, and image contrast-to-noise ratio were compared between the raw and denoised groups. Results: Block-matching and 3D filtering visually and quantitatively reduced image noise by 52% and improved image contrast-to-noise ratio by 56%, without compromising soft tissue edge definitions. Automatic tumor segmentation is statistically more consistent on the denoised 4D-MRI (σ-GTV = 0.6 cm{sup 3}) than on the raw 4D-MRI (σ-GTV = 0.8 cm{sup 3}). Tumor end-of-exhalation location is also more reproducible on the denoised 4D-MRI than on the raw 4D-MRI in all 3 cardinal motion planes. Conclusions: Block-matching and 3D filtering can significantly reduce random image noise while maintaining structural features in the SG-KS-4D-MRI datasets. In this study of pancreatic tumor segmentation, automatic segmentation of GTV in the registered image sets is shown to be more consistent on the denoised 4D-MRI than on the raw 4D-MRI.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Prediction of Solar Irradiance and Photovoltaic Solar Energy Product Based on Cloud Coverage Estimation Using Machine Learning Methods

Cloud cover estimation from images taken by sky-facing cameras can be an important input for analyzing current weather conditions and estimating photovoltaic power generation. The constant change in position, shape, and density of clouds, however, makes the development of a robust computational method for cloud cover estimation challenging. Accurately determining the edge of clouds and hence the separation between clouds and clear sky is difficult and often impossible. Toward determining cloud cover for estimating photovoltaic output, we propose using machine learning methods for cloud segmentation. We compare several methods including a classical regression model, deep learning methods, and boosting methods that combine results from the other machine learning models. To train each of the machine learning models with various sky conditions, we supplemented the existing Singapore whole sky imaging segmentation database with hazy and overcast images collected by a camera-equipped Waggle sensor node. We found that the U-Net architecture, one of the deep neural networks we utilized, segmented cloud pixels most accurately. However, the accuracy of segmenting cloud pixels did not guarantee high accuracy of estimating solar irradiance. We confirmed that the cloud cover ratio is directly related to solar irradiance. Additionally, we confirmed that solar irradiance and solar power output are closely related; hence, by predicting solar irradiance, we can estimate solar power output. This study demonstrates that sky-facing cameras with machine learning methods can be used to estimate solar power output. This ground-based approach provides an inexpensive way to understand solar irradiance and estimate production from photovoltaic solar facilities.

14 SOLAR ENERGY↗

Segmentation method comparison for residual fiber length measurement across tiled microscopy images

Fiber length distribution (FLD), in part, governs mechanical properties in discontinuous fiber composites, yet manual measurement methods limit the high-throughput characterization needed for materials design optimization. This study compares deep learning segmentation approaches for automated FLD measurement in large-field microscopy, evaluating how method choice affects the microstructural descriptors used in structure-property-processing relationships. A critical challenge is that high-resolution microscopy images (10,000×10,000 pixels) must be tiled for deep learning analysis, fragmenting fibers at boundaries. We demonstrate that segmentation method proves crucial for measurement accuracy. For example, instance segmentation with Slicing Aided Hyper Inference (SAHI) preserves individual fiber integrity across tiles while semantic segmentation prioritizes speed. Comparing against manual measurement of extracted carbon fibers, YOLOv11-SAHI matched manual ground truth (238 μm weighted mean) with 40x speedup (4.5 vs 167 minutes per image). U-Net provides rapid quantification although it is at the cost of reduced accuracy due only reliably measuring stand-alone fibers. Our comparative analysis reveals that instance segmentation with SAHI better preserves length measurements while semantic segmentation prioritizes speed, providing empirical guidance for method selection. The characterization provides essential inputs for mechanical property prediction models and inverse design workflows, accelerating composite materials development cycles.

Additive manufacturing↗