Search NASA⌕ Search

SEARCH · Search NASA

Results for “Mask”

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 937 records · Page 52

Effects of experimentally added salmon subsidies on resident fishes via direct and indirect pathways

Abstract Artificial additions of nutrients of differing forms such as salmon carcasses and analog pellets (i.e. pasteurized fishmeal) have been proposed as a means of stimulating aquatic productivity and enhancing populations of anadromous and resident fishes. Nutrient mitigation to enhance fish production in stream ecosystems assumes that the central pathway by which effects occur is bottom‐up, through aquatic primary and secondary production, with little consideration of reciprocal aquatic‐terrestrial pathways. The net outcome (i.e. bottom‐up vs. top‐down) of adding salmon‐derived materials to streams depend on whether or not these subsidies indirectly intensify predation on in situ prey via increases in a shared predator or alleviate such predation pressure. We conducted a 3‐year experiment across nine tributaries of the N. Fork Boise River, Idaho, USA , consisting of 500‐m stream reaches treated with salmon carcasses (n = 3), salmon carcass analog (n = 3), and untreated control reaches (n = 3). We observed 2–8 fold increases in streambed biofilms in the 2–6 weeks following additions of both salmon subsidy treatments in years 1 and 2 and a 1.5‐fold increase in standing crop biomass of aquatic invertebrates to carcass additions in the second year of our experiment. The consumption of benthic invertebrates by stream fishes increased 110–140% and 44–66% in carcass and analog streams in the same time frame, which may have masked invertebrate standing crop responses in years 3 and 4. Resident trout directly consumed 10.0–24.0 g·m −2 ·yr −1 of salmon carcass and <1–11.0 g·m −2 ·yr −1 of analog material, which resulted in 1.2–2.9 g·m −2 ·yr −1 and 0.03–1.4 g·m −2 ·yr −1 of tissue produced. In addition, a feedback flux of terrestrial maggots to streams contributed 0.0–2.0 g·m −2 ·yr −1 to trout production. Overall, treatments increased annual trout production by 2–3 fold, though density and biomass were unaffected. Our results indicate the strength of bottom‐up and top‐down responses to subsidy additions was asymmetrical, with top‐down forces masking bottom‐up effects that required multiple years to manifest. The findings also highlight the need for nutrient mitigation programs to consider multiple pathways of energy and nutrient flow to account for the complex effects of salmon subsidies in stream‐riparian ecosystems.

Collins, Scott F.↗

Security enrichment of optical image cryptosystem based on superposition technique using fractional Hartley and gyrator transform domains deploying equal modulus decomposition

In this paper, a symmetric cryptosystem has been suggested to increase security of double random phase encoding using superposition technique in fractional Hartley transform (FrHT) and gyrator transform (GT) domains. It uses structured phase mask (SPM) and hybrid mask (HM) deploying equal modulus decomposition (EMD). Related to original phase truncated Fourier transform (PTFT) based cryptosystem, the security of encoding and decoding is upgraded by using the superposition technique. With the usage of SPM we can increase the key space while encoding and can also overcome the axis alignment difficulty. Decoding is just the reversal of the encoding method. The proposed scheme has robustness against common attacks like: noise and occlusion attacks. With the help of numerical simulation results we also show that our cryptosystem is vulnerable to various attacks. Performance of the suggested method is calculated in terms of Mean Square Error (MSE) between original and decoded image and also the sensitivity of the SPM parameters and rotational angles of GT is investigated. The competence of the suggested method includes analysis on EMD, histogram and correlation coefficient. The suggested cryptosystem is also open to the occlusion attack and noise attack to analyze its performance and evenness.

42 ENGINEERING↗

Instance Segmentation for Direct Measurements of Satellites in Metal Powders and Automated Microstructural Characterization from Image Data

In this work, we propose instance segmentation as a useful tool for image analysis in materials science. Instance segmentation is an advanced technique in computer vision which generates individual segmentation masks for every object of interest that is recognized in an image. Using an out-of-the-box implementation of Mask R-CNN, instance segmentation is applied to images of metal powder particles produced through gas atomization. Leveraging transfer learning allows for the analysis to be conducted with a very small training set of labeled images. As well as providing another method for measuring the particle size distribution, we demonstrate the first direct measurements of the satellite content in powder samples. After analyzing the results for the labeled data dataset, the trained model was used to generate measurements for a much larger set of unlabeled images. The resulting particle size measurements showed reasonable agreement with laser scattering measurements. The satellite measurements were self-consistent and showed good agreement with the expected trends for different samples. Finally, we present a small case study showing how instance segmentation can be used to measure spheroidite content in the UltraHigh Carbon Steel DataBase, demonstrating the flexibility of the technique.

36 MATERIALS SCIENCE↗

Phosphatidylserine directly and positively regulates fusion of myoblasts into myotubes

Highlights: {yields} PS broadly and persistently trans-locates to the outer leaflet of plasma membrane during myoblast fusion into myotubes. {yields} Robust myotubes are formed when PS liposomes are added exogenously. {yields} PS increases the width of de novo myotubes and the numbers of myonuclei, but not the myotube length. {yields} Annexin V or PS antibody inhibits myotube formation by masking exposed PS. -- Abstract: Cell membrane consists of various lipids such as phosphatidylserine (PS), phosphatidylcholine (PC), and phosphatidylethanolamine (PE). Among them, PS is a molecular marker of apoptosis, because it is located to the inner leaflet of plasma membrane generally but it is moved to the outer leaflet during programmed cell death. The process of apoptosis has been implicated in the fusion of muscle progenitor cells, myoblasts, into myotubes. However, it remained unclear whether PS regulates muscle cell differentiation directly. In this paper, localization of PS to the outer leaflet of plasma membrane in proliferating primary myoblasts and during fusion of these myoblasts into myotubes is validated using Annexin V. Moreover, we show the presence of PS clusters at the cell-cell contact points, suggesting the importance of membrane ruffling and PS exposure for the myogenic cell fusion. Confirming this conclusion, experimentally constructed PS, but not PC liposomes dramatically enhance the formation of myotubes from myoblasts, thus demonstrating a direct positive effect of PS on the muscle cell fusion. In contrast, myoblasts exposed to PC liposomes produce long myotubes with low numbers of myonuclei. Moreover, pharmacological masking of PS on the myoblast surface inhibits fusion of these cells into myotubes in a dose-dependent manner.

60 APPLIED LIFE SCIENCES↗

Trichostatin A suppresses lung adenocarcinoma development in Grg1 overexpressing transgenic mice

Trichostatin A (TSA) is a histone deacetylase inhibitor and a potential therapeutic for various malignancies. The in vivo effect of TSA, however, has not been investigated in a transgenic lung cancer model. Previously, we generated transgenic mice with overexpression of Groucho-related-gene 1 (Grg1) and these mice all developed mucinous lung adenocarcinoma. Grg1 is a transcriptional co-repressor protein, the function of which is thought to depend on HDAC activity. However, functions outside the nucleus have also been proposed. We tested the supposition that Grg1-induced tumorigenesis is HDAC-dependent by assaying the therapeutic effect of TSA in the Grg1 transgenic mouse model. We found that TSA significantly inhibited lung tumorigenesis in Grg1 transgenic mice (p < 0.01). TSA did not affect overall Grg1 protein levels, but instead reduced ErbB1 and ErbB2 expression, which are upregulated by Grg1 in the absence of TSA. We confirmed this effect in A549 cells. Furthermore, lapatinib, an inhibitor of both ErbB1 and ErbB2, effectively masked the effect of TSA on the inhibition of A549 cell proliferation and migration, suggesting TSA does work, at least in part, by downregulating ErbB receptors. We additionally found that TSA reduced the expression of VEGF and VEGFR2, but not basic FGF and FGFR1. Our findings indicate that TSA effectively inhibits Grg1-induced lung tumorigenesis through the down-regulation of ErbB1 and ErbB2, as well as reduced VEGF signaling. This suggests TSA and other HDAC inhibitors could have therapeutic value in the treatment of lung cancers with Grg1 overexpression. - Highlights: • TSA suppresses lung tumorigenesis in Grg1 overexpressing transgenic mice. • TSA does not affect overall Grg1 protein levels in the mice and in A549 cells. • TSA reduces ErbB1 and ErbB2 expression in the mice and in A549 cells. • Lapatinib masks TSA-induced inhibition of A549 cell proliferation and migration. • TSA inhibits VEGF signaling, but not basic FGF signaling.

60 APPLIED LIFE SCIENCES↗

NeRVI: Compressive neural representation of visualization images for communicating volume visualization results

We present NeRVI, a new deep-learning approach that compresses a large collection of visualization images generated from time-varying data for communicating volume visualization results. Based on an image-based implicit neural representation, our approach represents tens of thousands of high-resolution rendering images parametrized by different parameters via a hybrid model of multilayer perceptrons and convolutional neural networks. Here, our model predicts images and corresponding masks, and the masks are utilized for loss computation and network training to capture fine structural details and small components. In conjunction with model quantization and weight encoding, NeRVI yields highly compact compressive neural representations while preserving the image fidelity well. We demonstrate the effectiveness of NeRVI with isosurface rendering and direct volume rendering images generated from multiple data sets and compare NeRVI with other state-of-the-art deep learning-based (InSituNet, SIREN, NeRF, and NeRV) methods. Quantitative and qualitative results show that NeRVI provides an alternative solution that augments domain scientists' ability to manage, represent, and communicate scientific visualization output.

97 MATHEMATICS AND COMPUTING↗

Mechanisms of NH3 and NO2 detection in carbon-nanotube-based sensors: An ab initio investigation

The mechanisms of NH3 and NO2 detection by single-walled carbon nanotube-based devices are investigated by ab initio calculations and the non-equilibrium Greens function (NEGF) methodology. While both NH3 and NO2 can physisorb to a pristine carbon nanotube, we show that their adsorption only results in small current changes through the device. For a carbon nanotube (CNT) attached to gold nanowire leads, the most sensitive detection site is at the CNT near the CNT-Au contact, where chemisorption occurs. The resulting change in electron transmission and low-bias current can lead to over 30% sensitivity. While both NH3 and NO2 can also chemisorb at the Au electrodes, their adsorption results in only a small change in the plurality of the conducting levels of the gold layers, and thus a small effect on current. In order to enhance the detection sensitivity, it is thus beneficial to mask the electrodes to prevent chemisorption. Furthermore, the length of the pure CNT segment does not strongly affect the relative sensitivity. Our results suggest that a short-CNT device with exposed contact regions and masked electrodes would have the greatest sensitivity.

Chemistry↗

Modeling of Cd adsorption to goethite-bacteria composites

The accurate modeling of heavy metal adsorption in complex systems is fundamental for risk assess-ments in soils and associated environments. Bacteria-iron (hydr)oxide associations in soils and sedi-ments play a critical role in heavy metal immobilization. The reduced adsorption of heavy metals on these composites have been widely reported using the component additivity (CA) method. However, there is a lack of a mechanism model to account for these deviations. In this study, we established models for Cd adsorption on goethite-Pseudomonas putida composites at 1:1 and 5:1 mass ratios. Cadmium adsorption on the 5:1 composite was consistent with the additivity method. However, the CA method over predicted Cd adsorption by approximately 8% on the 1:1 composite at high Cd concentration. The deviation was corrected by adding the site blockage reactions between P. putida and goethite. Both CA and “CA-site masking” models for Cd adsorption onto the composites were in line with the ITC data. These results indicate that CA method in simulating Cd adsorption on bacteria-iron oxides composites is limited to low bacterial and Cd concentrations. Therefore the interfacial complexation reactions that occur between iron (hydr)oxides and bacteria should be taken into account when high concentrations of bacteria and heavy metals are present.

Qu, Chenchen↗

Overcoming small minirhizotron datasets using transfer learning

Minirhizotron technology is widely used to study root growth and development. Yet, standard approaches for tracing roots in minirhiztron imagery is extremely tedious and time consuming. Machine learning approaches can help to automate this task. However, lack of enough annotated training data is a major limitation for the application of machine learning methods. Transfer learning is a useful technique to help with training when available datasets are limited. In this paper, we investigated the effect of pre-trained features from the massives-cale, irrelevant ImageNet dataset and a relatively moderate-scale, but relevant peanut root dataset on switchgrass root imagery segmentation applications. We compiled two minirhizotron image datasets to accomplish this study: one with 17,550 peanut root images and another with 28 switchgrass root images. Both datasets were paired with manually labeled ground truth masks. Deep neural networks based on the U-net architecture were used with different pre-trained features as initialization for automated, precise pixel-wise root segmentation in minirhizotron imagery. We observed that features pre-trained on a closely related but relatively moderate size dataset like our peanut dataset were more effective than features pre-trained on the large but unrelated ImageNet dataset. Here, we achieved high quality segmentation on peanut root dataset with 99.04% accuracy at the pixel-level and overcame errors in human-labeled ground truth masks. By applying transfer learning technique on limited switchgrass dataset with features pre-trained on peanut dataset, we obtained 99% segmentation accuracy in switchgrass imagery using only 21 images for training (fine tuning). Furthermore, the peanut pre-trained features can help the model converge faster and have much more stable performance.

59 BASIC BIOLOGICAL SCIENCES↗

The evens and odds of CMB anomalies

The lack of power of large-angle CMB anisotropies is known to increase its statistical significance at higher Galactic latitudes, where a string-inspired pre-inflationary scale can also be detected. Considering the Planck 2015 data, and relying largely on a Bayesian approach, we show that the effect is mostly driven by the even - harmonic multipoles with , which appear sizably suppressed in a way that is robust with respect to Galactic masking, along with the corresponding detections of . On the other hand, the first odd - multipoles are only suppressed at high Galactic latitudes. We investigate this behavior in different sky masks, constraining through even and odd multipoles, and we elaborate on possible implications. We include low- polarization data which, despite being noise-limited, help in attaining confidence levels of about 3 in the detection of . We also show by direct forecasts that a future all-sky -mode cosmic-variance-limited polarization survey may push the constraining power for beyond 5 .

79 ASTRONOMY AND ASTROPHYSICS↗

Clinical Experiences With Onboard Imager KV Images for Linear Accelerator-Based Stereotactic Radiosurgery and Radiotherapy Setup

Purpose: To report our clinical experiences with on-board imager (OBI) kV image verification for cranial stereotactic radiosurgery (SRS) and radiotherapy (SRT) treatments. Methods and Materials: Between January 2007 and May 2008, 42 patients (57 lesions) were treated with SRS with head frame immobilization and 13 patients (14 lesions) were treated with SRT with face mask immobilization at our institution. No margin was added to the gross tumor for SRS patients, and a 3-mm three-dimensional margin was added to the gross tumor to create the planning target volume for SRT patients. After localizing the patient with stereotactic target positioner (TaPo), orthogonal kV images using OBI were taken and fused to planning digital reconstructed radiographs. Suggested couch shifts in vertical, longitudinal, and lateral directions were recorded. kV images were also taken immediately after treatment for 21 SRS patients and on a weekly basis for 6 SRT patients to assess any intrafraction changes. Results: For SRS patients, 57 pretreatment kV images were evaluated and the suggested shifts were all within 1 mm in any direction (i.e., within the accuracy of image fusion). For SRT patients, the suggested shifts were out of the 3-mm tolerance for 31 of 309 setups. Intrafraction motions were detected in 3 SRT patients. Conclusions: kV imaging provided a useful tool for SRS or SRT setups. For SRS setup with head frame, it provides radiographic confirmation of localization using the stereotactic target positioner. For SRT with mask, a 3-mm margin is adequate and feasible for routine setup when TaPo is combined with kV imaging.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Quality Assessment of Frameless Fractionated Stereotactic Radiotherapy Using Cone Beam Computed Tomography

Purpose: A quality assessment of intracranial stereotactic radiotherapy was performed using cone beam computed tomography (CBCT). Setup errors were analyzed for two groups of patients: (1) those who were positioned using a frameless SonArray (FSA) system and immobilized with a bite plate and thermoplastic (TP) mask (the bFSA group); and (2) those who were positioned by room laser and immobilized using a TP mask (the mLAS group). Methods and Materials: A quality assurance phantom was used to study the system differences between FSA and CBCT. The quality assessment was performed using an Elekta Synergy imager (XVI) (Elekta Oncology Systems, Norcross, GA) and an On-Board Imager (OBI) (Varian Medical Systems, Palo Alto, CA) for 25 patients. For the first three fractions, and weekly thereafter, the FSA system was used for patient positioning, after which CBCT was performed to obtain setup errors. Results: (1) Phantom tests: The mean differences in the isocenter displacements for the two systems was 1.2 {+-} 0.7 mm. No significant variances were seen between the XVI and OBI units (p{approx}0.208). (2)Patient tests: The mean of the displacements between FSA and CBCT were independent of the CBCT system used; mean setup errors for the bFSA group were smaller (1.2 mm) than those of the mLAS group (3.2 mm) (p < 0.005). For the mLAS patients, the 90th percentile and the maximum rotational displacements were 3{sup o} and 5{sup o}, respectively. A 4-mm drift in setup accuracy occurred over the treatment course for 1 bFSA patient. Conclusions: System differences of less than 1 mm between CBCT and FSA were seen. Error regression was observed for the bFSA patients, using CBCT (up to 4 mm) during the treatment course. For the mLAS group, daily CBCT imaging was needed to obtain acceptable setup accuracies.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Setup Accuracy of the Novalis ExacTrac 6DOF System for Frameless Radiosurgery

Purpose: Stereotactic radiosurgery using frame-based positioning is a well-established technique for the treatment of benign and malignant lesions. By contrast, a new trend toward frameless systems using image-guided positioning techniques is gaining mainstream acceptance. This study was designed to measure the detection and positioning accuracy of the ExacTrac/Novalis Body (ET/NB) for rotations and to compare the accuracy of the frameless with the frame-based radiosurgery technique. Methods and Materials: A program was developed in house to rotate reference computed tomography images. The angles measured by the system were compared with the known rotations. The accuracy of ET/NB was evaluated with a head phantom with seven lead beads inserted, mounted on a treatment couch equipped with a robotic tilt module, and was measured with a digital water level and portal films. Multiple hidden target tests (HTT) were performed to measure the overall accuracy of the different positioning techniques for radiosurgery (i.e., frameless and frame-based with relocatable mask or invasive ring, respectively). Results: The ET/NB system can detect rotational setup errors with an average accuracy of 0.09 Degree-Sign (standard deviation [SD] 0.06 Degree-Sign ), 0.02 Degree-Sign (SD 0.07 Degree-Sign ), and 0.06 Degree-Sign (SD 0.14 Degree-Sign ) for longitudinal, lateral, and vertical rotations, respectively. The average positioning accuracy was 0.06 Degree-Sign (SD 0.04 Degree-Sign ), 0.08 Degree-Sign (SD 0.06 Degree-Sign ), and 0.08 Degree-Sign (SD 0.07 Degree-Sign ) for longitudinal, lateral and vertical rotations, respectively. The results of the HTT showed an overall three-dimensional accuracy of 0.76 mm (SD 0.46 mm) for the frameless technique, 0.87 mm (SD 0.44 mm) for the relocatable mask, and 1.19 mm (SD 0.45 mm) for the frame-based technique. Conclusions: The study showed high detection accuracy and a subdegree positioning accuracy. On the basis of phantom studies, the frameless technique showed comparable accuracy to the frame-based approach.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Clinical Evaluation of a Robotic 6-Degree of Freedom Treatment Couch for Frameless Radiosurgery

Purpose: To evaluate the added value of 6-degree of freedom (DOF) patient positioning with a robotic couch compared with 4DOF positioning for intracranial lesions and to estimate the immobilization characteristics of the BrainLAB frameless mask (BrainLAB AG, Feldkirchen, Germany), more specifically, the setup errors and intrafraction motion. Methods and Materials: We enrolled 40 patients with 66 brain metastases treated with frameless stereotactic radiosurgery and a 6DOF robotic couch. Patient positioning was performed with the BrainLAB ExacTrac stereoscopic X-ray system. Positioning results were collected before and after treatment to assess patient setup error and intrafraction motion. Existing treatment planning data were loaded and simulated for 4DOF positioning and compared with the 6DOF positioning. The clinical relevance was analyzed by means of the Paddick conformity index and the ratio of prescribed isodose volume covered with 4DOF to that obtained with the 6DOF positioning. Results: The mean three-dimensional setup error before 6DOF correction was 1.91 mm (SD, 1.25 mm). The rotational errors were larger in the longitudinal (mean, 0.23 Degree-Sign ; SD, 0.82 Degree-Sign ) direction compared with the lateral (mean, -0.09 Degree-Sign ; SD, 0.72 Degree-Sign ) and vertical (mean, -0.10 Degree-Sign ; SD, 1.03 Degree-Sign ) directions (p < 0.05). The mean three-dimensional intrafraction shift was 0.58 mm (SD, 0.42 mm). The mean intrafractional rotational errors were comparable for the vertical, longitudinal, and lateral directions: 0.01 Degree-Sign (SD, 0.35 Degree-Sign ), 0.03 Degree-Sign (SD, 0.31 Degree-Sign ), and -0.03 Degree-Sign (SD, 0.33 Degree-Sign ), respectively. The mean conformity index decreased from 0.68 (SD, 0.08) (6DOF) to 0.59 (SD, 0.12) (4DOF) (p < 0.05). A loss of prescribed isodose coverage of 5% (SD, 0.08) was found with the 4DOF positioning (p < 0.05). Half a degree for longitudinal and lateral rotations can be identified as a threshold for coverage loss. Conclusions: With a mask immobilization, patient setup error and intrafraction motions need to be evaluated and corrected for. The 6DOF patient positioning with a 6DOF robotic couch to correct translational and rotational setup errors improves target positioning with respect to treatment isocenter, which is in direct relation with the clinical outcome, compared with the 4DOF positioning.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Accelerating Dynamic Magnetic Resonance Imaging (MRI) for Lung Tumor Tracking Based on Low-Rank Decomposition in the Spatial–Temporal Domain: A Feasibility Study Based on Simulation and Preliminary Prospective Undersampled MRI

Purpose: To evaluate a low-rank decomposition method to reconstruct down-sampled k-space data for the purpose of tumor tracking. Methods and Materials: Seven retrospective lung cancer patients were included in the simulation study. The fully-sampled k-space data were first generated from existing 2-dimensional dynamic MR images and then down-sampled by 5 × -20 × before reconstruction using a Cartesian undersampling mask. Two methods, a low-rank decomposition method using combined dynamic MR images (k-t SLR based on sparsity and low-rank penalties) and a total variation (TV) method using individual dynamic MR frames, were used to reconstruct images. The tumor trajectories were derived on the basis of autosegmentation of the resultant images. To further test its feasibility, k-t SLR was used to reconstruct prospective data of a healthy subject. An undersampled balanced steady-state free precession sequence with the same undersampling mask was used to acquire the imaging data. Results: In the simulation study, higher imaging fidelity and low noise levels were achieved with the k-t SLR compared with TV. At 10 × undersampling, the k-t SLR method resulted in an average normalized mean square error <0.05, as opposed to 0.23 by using the TV reconstruction on individual frames. Less than 6% showed tracking errors >1 mm with 10 × down-sampling using k-t SLR, as opposed to 17% using TV. In the prospective study, k-t SLR substantially reduced reconstruction artifacts and retained anatomic details. Conclusions: Magnetic resonance reconstruction using k-t SLR on highly undersampled dynamic MR imaging data results in high image quality useful for tumor tracking. The k-t SLR was superior to TV by better exploiting the intrinsic anatomic coherence of the same patient. The feasibility of k-t SLR was demonstrated by prospective imaging acquisition and reconstruction.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Magnetic Resonance–Based Automatic Air Segmentation for Generation of Synthetic Computed Tomography Scans in the Head Region

Purpose: To incorporate a novel imaging sequence for robust air and tissue segmentation using ultrashort echo time (UTE) phase images and to implement an innovative synthetic CT (synCT) solution as a first step toward MR-only radiation therapy treatment planning for brain cancer. Methods and Materials: Ten brain cancer patients were scanned with a UTE/Dixon sequence and other clinical sequences on a 1.0 T open magnet with simulation capabilities. Bone-enhanced images were generated from a weighted combination of water/fat maps derived from Dixon images and inverted UTE images. Automated air segmentation was performed using unwrapped UTE phase maps. Segmentation accuracy was assessed by calculating segmentation errors (true-positive rate, false-positive rate, and Dice similarity indices using CT simulation (CT-SIM) as ground truth. The synCTs were generated using a voxel-based, weighted summation method incorporating T2, fluid attenuated inversion recovery (FLAIR), UTE1, and bone-enhanced images. Mean absolute error (MAE) characterized Hounsfield unit (HU) differences between synCT and CT-SIM. A dosimetry study was conducted, and differences were quantified using γ-analysis and dose-volume histogram analysis. Results: On average, true-positive rate and false-positive rate for the CT and MR-derived air masks were 80.8% ± 5.5% and 25.7% ± 6.9%, respectively. Dice similarity indices values were 0.78 ± 0.04 (range, 0.70-0.83). Full field of view MAE between synCT and CT-SIM was 147.5 ± 8.3 HU (range, 138.3-166.2 HU), with the largest errors occurring at bone–air interfaces (MAE 422.5 ± 33.4 HU for bone and 294.53 ± 90.56 HU for air). Gamma analysis revealed pass rates of 99.4% ± 0.04%, with acceptable treatment plan quality for the cohort. Conclusions: A hybrid MRI phase/magnitude UTE image processing technique was introduced that significantly improved bone and air contrast in MRI. Segmented air masks and bone-enhanced images were integrated into our synCT pipeline for brain, and results agreed well with clinical CTs, thereby supporting MR-only radiation therapy treatment planning in the brain.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

EHR-BERT: A BERT-based model for effective anomaly detection in electronic health records

Objective: Physicians and clinicians rely on data contained in electronic health records (EHRs), as recorded by health information technology (HIT), to make informed decisions about their patients. The reliability of HIT systems in this regard is critical to patient safety. Consequently, better tools are needed to monitor the performance of HIT systems for potential hazards that could compromise the collected EHRs, which in turn could affect patient safety. In this paper, we propose a new framework for detecting anomalies in EHRs using sequence of clinical events. This new framework, EHR-Bidirectional Encoder Representations from Transformers (BERT), is motivated by the gaps in the existing deep-learning related methods, including high false negatives, sub-optimal accuracy, higher computational cost, and the risk of information loss. EHR-BERT is an innovative framework rooted in the BERT architecture, meticulously tailored to navigate the hurdles in the contemporary BERT method; thus, enhancing anomaly detection in EHRs for healthcare applications.Methods: The EHR-BERT framework was designed using the Sequential Masked Token Prediction (SMTP) method. This approach treats EHRs as natural language sentences and iteratively masks input tokens during both training and prediction stages. This method facilitates the learning of EHR sequence patterns in both directions for each event and identifies anomalies based on deviations from the normal execution models trained on EHR sequences.Results: Extensive experiments on large EHR datasets across various medical domains demonstrate that EHR-BERT markedly improves upon existing models. It significantly reduces the number of false positives and enhances the detection rate, thus bolstering the reliability of anomaly detection in electronic health records. This improvement is attributed to the model’s ability to minimize information loss and maximize data utilization effectively.Conclusion: EHR-BERT showcases immense potential in decreasing medical errors related to anomalous clinical events, positioning itself as an indispensable asset for enhancing patient safety and the overall standard of healthcare services. The framework effectively overcomes the drawbacks of earlier models, making it a promising solution for healthcare professionals to ensure the reliability and quality of health data.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

A massively-parallel, unstructured overset method for mesh connectivity

Here we present a method that dynamically and efficiently performs connectivity calculations between many O(10 5 ) moving, unstructured overset meshes in parallel. In order to connect overset meshes, elements exterior to the solution domain must be removed from the simulation. In regions with many overlapping meshes, elements must be selectively removed to reduce redundancy while maintaining a solution over the entire domain. Around masked regions interpolation partner pairing is required between meshes to provide boundary conditions. For general unstructured meshes, these steps involve challenging computational geometry calculations which must be efficient and automatic. For many moving meshes each step must be massively parallelized and scalable to large numbers of computational cores. To establish communication patterns a parallelized master/slave algorithm is used which minimizes global communication and storage. To remove elements a parallel ‘Forest Fire’ flood-fill algorithm is used to set a masking variable. For interpolation partner pairing, and other necessary searches, k-dimensional tree data structures (k-d trees) are extensively used. Often in a calculation, the connectivity between overset meshes remains largely the same between time steps. The temporal coherence of the various objects in the connectivity calculation is directly used to only update necessary information with time, resulting in substantial cost savings. Details of the different algorithms are presented. Resulting connectivity and timings are shown for complex geometries. Parallel scaling is demonstrated for 100,000 spherical particles within a channel up to 492,000 processors.

97 MATHEMATICS AND COMPUTING↗