Search NASA⌕ Search

SEARCH · Search NASA

Results for “human performance data”

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 271 records · Page 15

Automated Classification of Vehicle Movements at Signalized Intersections Using Vehicle Trajectories

Accurate vehicle movement classification through signalized intersections is of paramount importance to the analysis of intersection performance and the optimization of traffic control strategies. Conventional techniques for tracking vehicle turning movements depend on infrastructure-based strategies like human counts, loop detectors, and video analytics, all of which are costly, prone to errors, and spatially constrained. High-frequency trajectory data can be utilized to determine vehicle movement patterns in a scalable and infrastructure-independent method due to the adoption of connected vehicles (CVs). In recent years, several studies have utilized connected vehicle data to generate performance measures. Most of the trajectory-based performance measures approaches, however, require map matching-i.e., extracting geospatial references from maps to identify the movements that individual vehicles make at a signalized intersection. These approaches are often time-consuming and hinder scalability since geographic features need to be provided for an analysis to be conducted. Map matching methods are prone to errors as different map versions change these geographic features. This research presents a novel automatic classification pipeline that uses CV trajectory data to classify vehicle movements at signalized crossings, specifically pass-through left-turn and right-turn maneuvers. The process starts by filtering trips that cross a spatial bounding box that has been defined at the target intersection. Approach and departure headings for each trajectory crossing the boundary are computed and are clustered together to identify dominant movements. The proposed algorithm is used to classify the movement of vehicles at 10 intersections in the state of California, and the results indicate that the algorithm can classify movements at these intersections with varying traffic volumes and road network configurations, all in a map-less framework with no need for conflation of vehicle trajectories to a digital base map.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Closure Report for Corrective Action Unit 117: Area 26 Pluto Disassembly Facility, Nevada Test Site, Nevada, Revision 0

This Closure Report (CR) presents information supporting the closure of Corrective Action Unit (CAU) 117: Area 26 Pluto Disassembly Facility, Nevada Test Site, Nevada. This CR complies with the requirements of the Federal Facility Agreement and Consent Order that was agreed to by the State of Nevada; U.S. Department of Energy (DOE), Environmental Management; U.S. Department of Defense; and DOE, Legacy Management. Corrective Action Unit 117 comprises Corrective Action Site (CAS) 26-41-01, Pluto Disassembly Facility, located in Area 26 of the Nevada Test Site. The purpose of this CR is to provide documentation supporting the completed corrective actions and provide data confirming that the closure objectives for CAU 117 were met. To achieve this, the following actions were performed: • Review the current site conditions, including the concentration and extent of contamination. • Implement any corrective actions necessary to protect human health and the environment. • Properly dispose of corrective action and investigation wastes. • Document Notice of Completion and closure of CAU 117 issued by the Nevada Division of Environmental Protection. From May 2008 through February 2009, closure activities were performed as set forth in the Streamlined Approach for Environmental Restoration Plan for Corrective Action Unit 117, Area 26 Pluto Disassembly Facility, Nevada Test Site, Nevada. The purpose of the activities as defined during the data quality objectives process were: • Determine whether contaminants of concern (COCs) are present. • If COCs are present, determine their nature and extent, implement appropriate corrective actions, and properly dispose of wastes. Analytes detected during the closure activities were evaluated against final action levels to determine COCs for CAU 117. Assessment of the data generated from closure activities indicated that the final action levels were exceeded for polychlorinated biphenyls (PCBs) reported as total Aroclor and radium-226. A corrective action was implemented to remove approximately 50 cubic yards of PCB-contaminated soil, approximately 1 cubic foot of radium-226 contaminated soil (and scabbled asphalt), and a high-efficiency particulate air filter that was determined to meet the criteria of a potential source material (PSM). Electrical and lighting components (i.e., PCB-containing ballasts and capacitors) and other materials (e.g., mercury-containing thermostats and switches, lead plugs and bricks) assumed to be PSM were also removed from Building 2201, as practical, without the need for sampling. Because the COC contamination and PSMs have been removed, clean closure of CAS 26-41-01 is recommended, and no use restrictions are required to be placed on this CAU. No further action is necessary because no other contaminants of potential concern were found above preliminary action levels. The physical end state for Building 2201 is expected to be eventual demolition to slab. The DOE, National Nuclear Security Administration Nevada Site Office provides the following recommendations: • Clean closure is the recommended corrective action for CAS 26-41-01 in CAU 117. • A Notice of Completion to the DOE, National Nuclear Security Administration Nevada Site Office is requested from the Nevada Division of Environmental Protection for closure of CAU 117. • Corrective Action Unit 117 should be moved from Appendix III to Appendix IV of the Federal Facility Agreement and Consent Order.

54 ENVIRONMENTAL SCIENCES↗

Brain Connectomics: Opportunities for High-Performance Computing (Final Report)

The human brain is estimated to contain approximately 100 trillion neural connections. This complex map of connectivity (human connectome) underlies cognitive processes and informs disease states, but until recently, studying the human connectome has been hampered by the difficulty of collecting and processing suitable data for analysis. In this context, an initiative denominated The Human Connectome Project (HCP) was created to acquire and collect the largest, most cohesive set of brain data to accelerate neuroscience research. Recently the HCP released a magnetic resonance (MR) dataset of brain imagery with unprecedented scale and organization, together with high quality spatial and temporal resolution; however, fully taking advantage of this high-resolution brain imagery to estimate neurological connectivity networks requires high-performance computing (HPC). In this project, we explored the feasibility of using HPC resources at Lawrence Livermore to effectively estimate brain connectivity networks from this large set of imagery. Specifically, we addressed one of the main current challenges in brain connectomics, which is to advance the computation of individual structural and functional connectomes.

42 ENGINEERING↗

cTULIP: application of a human-based RNA-seq primary tumor classification tool for cross-species primary tumor classification in canine

The domestic dog, Canis familiaris, is quickly gaining traction as an advantageous model for use in the study of cancer, one of the leading causes of death worldwide. Naturally occurring canine cancers share clinical, histological, and molecular characteristics with the corresponding human diseases. In this study, we take a deep-learning approach to test how similar the gene expression profile of canine glioma and bladder cancer (BLCA) tumors are to the corresponding human tumors. We likewise develop a tool for identifying misclassified or outlier samples in large canine oncological datasets, analogous to that which was developed for human datasets. We test a number of machine learning algorithms and found that a convolutional neural network outperformed logistic regression and random forest approaches. We use a recently developed RNA-seq-based convolutional neural network, TULIP, to test the robustness of a human-data-trained primary tumor classification tool on cross-species primary tumor prediction. Our study ultimately highlights the molecular similarities between canine and human BLCA and glioma tumors, showing that protein-coding one-to-one homologs shared between humans and canines, are sufficient to distinguish between BLCA and gliomas. The results of this study indicate that using protein-coding one-to-one homologs as the features in the input layer of TULIP performs good primary tumor prediction in both humans and canines. Furthermore, our analysis shows that our selected features also contain the majority of features with known clinical relevance in BLCA and gliomas. Our success in using a human-data-trained model for cross-species primary tumor prediction also sheds light on the conservation of oncological pathways in humans and canines, further underscoring the importance of the canine model system in the study of human disease.

60 APPLIED LIFE SCIENCES↗

Video performance for high security applications.

The complexity of physical protection systems has increased to address modern threats to national security and emerging commercial technologies. A key element of modern physical protection systems is the data presented to the human operator used for rapid determination of the cause of an alarm, whether false (e.g., caused by an animal, debris, etc.) or real (e.g., a human adversary). Alarm assessment, the human validation of a sensor alarm, primarily relies on imaging technologies and video systems. Developing measures of effectiveness (MOE) that drive the design or evaluation of a video system or technology becomes a challenge, given the subjectivity of the application (e.g., alarm assessment). Sandia National Laboratories has conducted empirical analysis using field test data and mathematical models such as binomial distribution and Johnson target transfer functions to develop MOEs for video system technologies. Depending on the technology, the task of the security operator and the distance to the target, the Probability of Assessment (PAs) can be determined as a function of a variety of conditions or assumptions. PAs used as an MOE allows the systems engineer to conduct trade studies, make informed design decisions, or evaluate new higher-risk technologies. This paper outlines general video system design trade-offs, discusses ways video can be used to increase system performance and lists MOEs for video systems used in subjective applications such as alarm assessment.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

SEPATH: benchmarking the search for pathogens in human tissue whole genome sequence data leads to template pipelines

Background: Human tissue is increasingly being whole genome sequenced as we transition into an era of genomic medicine. With this arises the potential to detect sequences originating from microorganisms, including pathogens amid the plethora of human sequencing reads. In cancer research, the tumorigenic ability of pathogens is being recognized, for example, Helicobacter pylori and human papillomavirus in the cases of gastric non-cardia and cervical carcinomas, respectively. As of yet, no benchmark has been carried out on the performance of computational approaches for bacterial and viral detection within host-dominated sequence data. Results: We present the results of benchmarking over 70 distinct combinations of tools and parameters on 100 simulated cancer datasets spiked with realistic proportions of bacteria. mOTUs2 and Kraken are the highest performing individual tools achieving median genus-level F1 scores of 0.90 and 0.91, respectively. mOTUs2 demonstrates a high performance in estimating bacterial proportions. Employing Kraken on unassembled sequencing reads produces a good but variable performance depending on post-classification filtering parameters. These approaches are investigated on a selection of cervical and gastric cancer whole genome sequences where Alphapapillomavirus and Helicobacter are detected in addition to a variety of other interesting genera. Conclusions: We provide the top-performing pipelines from this benchmark in a unifying tool called SEPATH, which is amenable to high throughput sequencing studies across a range of high-performance computing clusters. SEPATH provides a benchmarked and convenient approach to detect pathogens in tissue sequence data helping to determine the relationship between metagenomics and disease.

Viral↗

Deep Learning for Automated Identification of Eels in Sonar Data

Freshwater eels, such as the American eel (Anguilla rostrata) present numerous challenges related to safe downstream fish passage at hydroelectric facilities. One of those challenges is effective monitoring of their abundance, movements, and behavior to facilitate design and operation of eel protection and passage facilities. A previous EPRI study documented the ability of human analysts to reliably identify American eels in data obtained with a 1100/1800 kHz, multibeam sonar. This report describes a project to develop deep learning (a subset of artificial intelligence) tools to automate the time-consuming, subjective process of eel identification in multibeam sonar data. The project exploited new data collected in the laboratory and the existing data from the prior EPRI field study to develop and test deep learning and other data analytic tools, including wavelet filtering, differencing for static object removal, and convolutional neural network analysis. The analysis of the laboratory data demonstrated feasibility of the approach, revealed object characteristics observed with the sonar that distinguish eels from similarly sized and shaped acoustic targets, and provided additional data for algorithm selection and training. Deep learning algorithms trained and tested on the laboratory data alone achieved accuracy rates of greater than 98% when classifying acoustic images of eels and similar-sized neutrally buoyant sticks. The algorithm trained and tested on the pre-existing field data alone, and yielded classification accuracy of 9.3% false positives and 13.3% false negatives when distinguishing between eels and sticks/PVC pipes based on video clips (i.e., multiple, consecutive images). This performance is comparable to the classification accuracy achieved by human analysts in the prior study. The deep learning algorithm trained on a combination of video clips obtained in the laboratory and the field and tested on video clips from the field, was able to distinguish eels from sticks and PVC pipes (a river debris analog) of similar size with 100% accuracy. Outreach to the hardware, software, and end-user communities early in the project helped to identify needs and specify the application space. Outreach to those communities at the end of the project communicated project results and opportunities for further development. The project achieved proof of concept for automated identification of eel in multibeam sonar data. Future work should focus on acquisition of additional data for more robust algorithm training and testing; modification of the software tools to accommodate multiple acoustic targets in the acoustic field at a given time; identification of additional object classes; incorporation of motion in the object identification and classification algorithms; operationalizing the software tools, including integration with other existing sonar data analysis tools; and partnering with hardware and software providers for distribution of the software tools with their commercial products.

13 HYDRO ENERGY↗

Estimating the Contributions to Human Error Probability from the Convolution of the Distribution of Time Available and Time Required

As part of their duties, Human Reliability Analysis must often evaluate if crews in nuclear power plants (NPPs) can complete tasks associated with a human-failure event within time limits. For example, the time required in NPP scenarios is determined by systematic and structured walkthroughs, feasibility studies, recorded times from training exercises, and interviews with experienced operators and experts. Typically, a point estimate is derived for the estimate (mean, maximum, or 95th percentile of time required). Using point-estimate values can mask the risk associated with variability among crews, plant conditions and set-up, environmental conditions, and other impact factors under which these actions are executed. While point estimates for time required and time available have served the industry well, without considering the uncertainty they could lead to biased understanding about the risk. The Integrated Human Event Analysis System - General Methodology (IDHEAS-G) model (developed by the US Nuclear Regulatory Commission, NRC) for human error probability calculates human error probability by summing two probabilities: insufficient time and cognitive error. As such, the model takes a more holistic approach by considering the full distributions for time required and time available to calculate the human error probability because the time available to complete the task is insufficient. In this study, we expand on the work of the NRC and discuss methods for estimating these time considerations. For example, for the time required, the impact of Performance Influencing Factors (PIFs) on the distribution was divided into impacts that are aleatory in nature, such as crew-to-crew variability, and those that are epistemic (i.e., the PIFs). Starting with the factors that introduce aleatory uncertainty, a first-order distribution was developed from a large set of time required (i.e., NPP task completion times) data for the range of operator actions that occur in the NPP control room under simulated accident conditions. The first-order distribution can then be adjusted to account for epistemic uncertainty using research associated with the impact of applicable PIFs on the time required. We also develop guidance for analysts to address the probability distributions for the time available. The guidance we developed on how to estimate time required and time available distributions is based on the identification of pertinent research and data, data analyses, and expert knowledge elicitation.

human error probability, human performance, time e↗

The Iterative Processing Framework: A New Paradigm for Automatic Event Building

Abstract In a traditional data-processing pipeline, waveforms are acquired, a detector makes the signal detections (i.e., arrival times, slownesses, and azimuths) and passes them to an associator. The associator then links the detections to the fitting-event hypotheses to generate an event bulletin. Most of the time, this traditional pipeline requires substantial human-analyst involvement to improve the quality of the resulting event bulletin. For the year 2017, for instance, International Data Center (IDC) analysts rejected about 40% of the events in the automatic bulletin and manually built 30% of the legitimate events. Here, we introduce an iterative processing framework (IPF) that includes a new data-processing module that incorporates automatic analyst behaviors (auto analyst [AA]) into the event-building pipeline. In the proposed framework, through an iterative process, the AA takes over many of the tasks traditionally performed by human analysts. These tasks can be grouped into two major processes: (1) evaluating small events with a low number of location-defining arrival phases to improve their formation; and (2) scanning for and exploiting unassociated arrivals to form potential events missed by previous association runs. To test the proposed framework, we processed a two-week period (15–28 May 2010) of the signal-detections dataset from the IDC. Comparison with an expert analyst-reviewed bulletin for the same time period suggests that IPF performs better than the traditional pipelines (IDC and baseline pipelines). The majority of the additional events built by the AA are low-magnitude events that were missed by these traditional pipelines. The AA also adds additional signal detections to existing events, which saves analyst time, even if the event locations are not significantly affected.

58 GEOSCIENCES↗

MetaboTools: A comprehensive toolbox for analysis of genome-scale metabolic models

Metabolomic data sets provide a direct read-out of cellular phenotypes and are increasingly generated to study biological questions. Previous work, by us and others, revealed the potential of analyzing extracellular metabolomic data in the context of the metabolic model using constraint-based modeling. With the MetaboTools, we make our methods available to the broader scientific community. The MetaboTools consist of a protocol, a toolbox, and tutorials of two use cases. The protocol describes, in a step-wise manner, the workflow of data integration, and computational analysis. The MetaboTools comprise the Matlab code required to complete the workflow described in the protocol. Tutorials explain the computational steps for integration of two different data sets and demonstrate a comprehensive set of methods for the computational analysis of metabolic models and stratification thereof into different phenotypes. The presented workflow supports integrative analysis of multiple omics data sets. Importantly, all analysis tools can be applied to metabolic models without performing the entire workflow. Taken together, the MetaboTools constitute a comprehensive guide to the intra-model analysis of extracellular metabolomic data from microbial, plant, or human cells. In conclusion, this computational modeling resource offers a broad set of computational analysis tools for a wide biomedical and non-biomedical research community.

59 BASIC BIOLOGICAL SCIENCES↗

Analysis in Motion Initiative – Human Machine Intelligence

As computers and machines become more pervasive in our everyday lives, we are looking for ways for humans and machines to work more intelligently together. How can we help machines understand their users so the team can do smarter things together? The Analysis in Motion Initiative is advancing the science of human machine intelligence — creating human-machine teams that work better together to make correct, useful, and timely interpretations of data.

97 MATHEMATICS AND COMPUTING↗

Efficient DNA sequence compression with neural networks

Abstract Background The increasing production of genomic data has led to an intensified need for models that can cope efficiently with the lossless compression of DNA sequences. Important applications include long-term storage and compression-based data analysis. In the literature, only a few recent articles propose the use of neural networks for DNA sequence compression. However, they fall short when compared with specific DNA compression tools, such as GeCo2. This limitation is due to the absence of models specifically designed for DNA sequences. In this work, we combine the power of neural networks with specific DNA models. For this purpose, we created GeCo3, a new genomic sequence compressor that uses neural networks for mixing multiple context and substitution-tolerant context models. Findings We benchmark GeCo3 as a reference-free DNA compressor in 5 datasets, including a balanced and comprehensive dataset of DNA sequences, the Y-chromosome and human mitogenome, 2 compilations of archaeal and virus genomes, 4 whole genomes, and 2 collections of FASTQ data of a human virome and ancient DNA. GeCo3 achieves a solid improvement in compression over the previous version (GeCo2) of $2.4\%$, $7.1\%$, $6.1\%$, $5.8\%$, and $6.0\%$, respectively. To test its performance as a reference-based DNA compressor, we benchmark GeCo3 in 4 datasets constituted by the pairwise compression of the chromosomes of the genomes of several primates. GeCo3 improves the compression in $12.4\%$, $11.7\%$, $10.8\%$, and $10.1\%$ over the state of the art. The cost of this compression improvement is some additional computational time (1.7–3 times slower than GeCo2). The RAM use is constant, and the tool scales efficiently, independently of the sequence size. Overall, these values outperform the state of the art. Conclusions GeCo3 is a genomic sequence compressor with a neural network mixing approach that provides additional gains over top specific genomic compressors. The proposed mixing method is portable, requiring only the probabilities of the models as inputs, providing easy adaptation to other data compressors or compression-based data analysis tools. GeCo3 is released under GPLv3 and is available for free download at https://github.com/cobilab/geco3.

Silva, Milton↗

Veridical data science

Building and expanding on principles of statistics, machine learning, and scientific inquiry, we propose the predictability, computability, and stability (PCS) framework for veridical data science. Our framework, composed of both a workflow and documentation, aims to provide responsible, reliable, reproducible, and transparent results across the data science life cycle. The PCS workflow uses predictability as a reality check and considers the importance of computation in data collection/storage and algorithm design. It augments predictability and computability with an overarching stability principle. Stability expands on statistical uncertainty considerations to assess how human judgment calls impact data results through data and model/algorithm perturbations. As part of the PCS workflow, we develop PCS inference procedures, namely PCS perturbation intervals and PCS hypothesis testing, to investigate the stability of data results relative to problem formulation, data cleaning, modeling decisions, and interpretations. We illustrate PCS inference through neuroscience and genomics projects of our own and others. Moreover, we demonstrate its favorable performance over existing methods in terms of receiver operating characteristic (ROC) curves in high-dimensional, sparse linear model simulations, including a wide range of misspecified models. Finally, we propose PCS documentation based on R Markdown or Jupyter Notebook, with publicly available, reproducible codes and narratives to back up human choices made throughout an analysis. The PCS workflow and documentation are demonstrated in a genomics case study available on Zenodo.

97 MATHEMATICS AND COMPUTING↗

Closure Report for Corrective Action Unit 408: Bomblet Target Area Tonopah Test Range (TTR), Nevada, Revision 0

This Closure Report (CR) presents information supporting the closure of Corrective Action Unit (CAU) 408: Bomblet Target Area (TTR), Tonopah Test Range, Nevada. This CR complies with the requirements of the Federal Facility Agreement and Consent Order that was agreed to by the State of Nevada; U.S. Department of Energy (DOE), Environmental Management; U.S. Department of Defense; and DOE, Legacy Management. Corrective Action Unit 408 is located at the Tonopah Test Range, Nevada, and consists of Corrective Action Site (CAS) TA-55-002-TAB2, Bomblet Target Areas. This CAS includes the following seven target areas: • Mid Target • Flightline Bomblet Location • Strategic Air Command (SAC) Target Location 1 • SAC Target Location 2 • South Antelope Lake • Tomahawk Location 1 • Tomahawk Location 2 The purpose of this CR is to provide documentation supporting the completed corrective actions and data confirming that the closure objectives for the CAS within CAU 408 were met. To achieve this, the following actions were performed: • Review the current site conditions, including the concentration and extent of contamination. • Implement any corrective actions necessary to protect human health and the environment. • Properly dispose of corrective action and investigation wastes. • Document Notice of Completion and closure of CAU 408 issued by the Nevada Division of Environmental Protection. From July 2009 through August 2010, closure activities were performed as set forth in the Streamlined Approach for Environmental Restoration Plan for CAU 408: Bomblet Target Area, Tonopah Test Range (TTR), Nevada. The purposes of the activities as defined during the data quality objectives process were as follows: • Identify and remove munitions of explosive concern (MEC) associated with DOE activities. • Investigate potential disposal pit locations. • Remove depleted uranium-contaminated fragments and soil. • Determine whether contaminants of concern (COCs) are present. • If COCs are present, determine their nature and extent, implement appropriate corrective actions, and properly dispose of wastes. Analytes detected during the closure activities were evaluated against final action levels to determine COCs for CAU 408. Assessment of the data indicated COCs are not present at CAS TA-55-002-TAB2; therefore, no corrective action is necessary. No use restrictions are required to be placed on this CAU because the investigation showed no evidence of remaining soil contamination or remaining debris/waste upon completion of all investigation activities. The MEC was successfully removed and dispositioned as planned using current best available technologies. As MEC guidance and general MEC standards acknowledge that MEC response actions cannot determine with 100 percent certainty that all MEC and unexploded ordnance (UXO) are removed, the clean closure of CAU 408 will implement a best management practice of posting UXO hazard warning signs near the seven target areas. The signs will warn future land users of the potential for encountering residual UXO hazards. The DOE, National Nuclear Security Administration Nevada Site Office, provides the following recommendations: • A Notice of Completion to the DOE, National Nuclear Security Administration Nevada Site Office, is requested from the Nevada Division of Environmental Protection for closure of CAU 408. • Corrective Action Unit 408 should be moved from Appendix III to Appendix IV of the Federal Facility Agreement and Consent Order.

54 ENVIRONMENTAL SCIENCES↗

UAVs for Radiological Monitoring: Determining Ideal Operational Parameters From GEANT4 Simulations - 18025

Owing in no small part to the enormous advances in radiation detector technology (size, sensitivity and spectral-resolution), there has been an increased drive and requirement to accurately map both the spatial extent and magnitude of radiological contamination at the greatest possible spatial resolutions. To undertake this, a range of different methods have been devised and include the use of cars and human, but also planes (and more recently) unmanned aerial vehicles (UAVs). However, critical to the remote sensing using such UAVs in contrast to ground-based sensing, is (i) the selection of the ideal operational parameters to attain the best on-ground resolution in addition to, (ii) understanding the correct normalisation procedures and other influences on results that surveys at these reduced altitudes might present. Resulting from earlier works, through employing combined modelling and experimental data, the optimal height for a UAV survey using the more realistic 'area' radioactive source term (opposed to an infinitely-small 'point-source') is determined to be approximately 5 m - suitably maximizing cross detector counts. At greater altitudes than this, the influence of variations in neighboring topology / topography and nearby structures are observed to be negligible. However, at this and further-reduced altitudes (where optimal count-rates exists), small height-related variations are observed to result, and impact considerably on the complexity of the data normalisation. Whilst not occurring as a tool to entirely replace ground-based surveys, UAV systems are able to perform in challenging or dangerous environments, inadvisable or problematic to send humans. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Data-centric framework for crystal structure identification in atomistic simulations using machine learning

Atomic-level modeling performed at large scales enables the investigation of mesoscale materials properties with atom-by-atom resolution. The spatial complexity of such cross-scale simulations renders them unsuitable for simple human visual inspection. Instead, specialized structure characterization techniques are required to aid interpretation. These have historically been challenging to construct, requiring significant intuition and effort. Here we propose an alternative framework for a fundamental structural characterization task: classifying atoms according to the crystal structure to which they belong. Our approach is data-centric and favors the employment of Machine Learning over heuristic rules of classification. A group of data-science tools and simple local descriptors of atomic structure are employed together with an efficient synthetic training set. We also introduce the first standard and publicly available benchmark data set for evaluation of algorithms for crystal-structure classification. Further, it is demonstrated that our data-centric framework outperforms all of the most popular heuristic methods—especially at high temperatures when lattices are the most distorted—while introducing a systematic route for generalization to new crystal structures. Moreover, through the use of outlier detection algorithms our approach is capable of discerning between amorphous atomic motifs (i.e., noncrystalline phases) and unknown crystal structures, making it uniquely suited for exploratory materials synthesis simulations.

36 MATERIALS SCIENCE↗

Bone-marrow densitometry: Assessment of marrow space of human vertebrae by single energy high resolution-quantitative computed tomography

Purpose: Accurate noninvasive assessment of vertebral bone marrow fat fraction is important for diagnostic assessment of a variety of disorders and therapies known to affect marrow composition. Moreover, it provides a means to correct fat-induced bias of single energy quantitative computed tomography (QCT) based bone mineral density (BMD) measurements. The authors developed new segmentation and calibration methods to obtain quantitative surrogate measures of marrow-fat density in the axial skeleton. Methods: The authors developed and tested two high resolution-QCT (HR-QCT) based methods which permit segmentation of bone voids in between trabeculae hypothesizing that they are representative of bone marrow space. The methods permit calculation of marrow content in units of mineral equivalent marrow density (MeMD). The first method is based on global thresholding and peeling (GTP) to define a volume of interest away from the transition between trabecular bone and marrow. The second method, morphological filtering (MF), uses spherical elements of different radii (0.1–1.2 mm) and automatically places them in between trabeculae to identify regions with large trabecular interspace, the bone-void space. To determine their performance, data were compared ex vivo to high-resolution peripheral CT (HR-pQCT) images as the gold-standard. The performance of the methods was tested on a set of excised human vertebrae with intact bone marrow tissue representative of an elderly population with low BMD. Results: 86% (GTP) and 87% (MF) of the voxels identified as true marrow space on HR-pQCT images were correctly identified on HR-QCT images and thus these volumes of interest can be considered to be representative of true marrow space. Within this volume, MeMD was estimated with residual errors of 4.8 mg/cm{sup 3} corresponding to accuracy errors in fat fraction on the order of 5% both for GTP and MF methods. Conclusions: The GTP and MF methods on HR-QCT images permit noninvasive localization and densitometric assessment of marrow fat with residual accuracy errors sufficient to study disorders and therapies known to affect bone marrow composition. Additionally, the methods can be used to correct BMD for fat induced bias. Application and testing in vivo and in longitudinal studies are warranted to determine the clinical performance and value of these methods.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Improvement of prostate cancer detection combining a computer-aided diagnostic system with TRUS-MRI targeted biopsy

Purpose: To validate a novel consensus method, called target-in-target, combining human analysis of mpMRI with automated CAD system analysis, with the aim to increasing the prostate cancer detection rate of targeted biopsies. Methods: A cohort of 420 patients was enrolled and 253 patients were rolled out, due to exclusion criteria. 167 patients, underwent diagnostic 3T MpMRI. Two expert radiologists evaluated the exams adopting PI-RADSv2 and CAD system. When a CAD target overlapped with a radiologic one, we performed the biopsy in the overlapping area which we defined as target-in-target. Targeted TRUS-MRI fusion biopsy was performed in 63 patients with a total of 212 targets. The MRI data of all targets were quantitatively analyzed, and diagnostic findings were compared to pathologist’s biopsy reports. Results: CAD system diagnostic performance exhibited sensitivity and specificity scores of 55.2% and 74.1% [AUC = 0.63 (0.54 ÷ 0.71)] , respectively. Human readers achieved an AUC value, in ROC analysis, of 0.71 (0.63 ÷ 0.79). The target-in-target method provided a detection rate per targeted biopsy core of 81.8 % vs. a detection rate per targeted biopsy core of 68.6 % for pure PI-RADS based on target definitions. The higher per-core detection rate of the target-in-target approach was achieved irrespective of the presence of technical flaws and artifacts. Conclusions: A novel consensus method combining human reader evaluation with automated CAD system analysis of mpMRI to define prostate biopsy targets was shown to improve the detection rate per biopsy core of TRUS-MRI fusion biopsies. Results suggest that the combination of CAD system analysis and human reader evaluation is a winning strategy to improve targeted biopsy efficiency.

62 RADIOLOGY AND NUCLEAR MEDICINE↗