Search NASA⌕ Search

SEARCH · Search NASA

Results for “evaluation datasets”

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 343 records · Page 19

Mitigate: An Adaptive Network Data Anonymization Tool Using Condensation-Based Differential Privacy

Modern network devices collect a large amount of data that can be analyzed to identify bottlenecks, anomalies, cyber-attacks, etc. Therefore, there is often a need to analyze such collections of network data quite often by an external expert or by the research community. However, these collections of data contain sensitive, proprietary information. In order for the network data to be shared, it must first be anonymized. The overall objective of this project is to develop an innovative privacy management tool to anonymize network data and achieve sufficient privacy, acceptable data utility, and efficient data analysis at the same time. No existing anonymization methods can achieve all of these at the same time. The core of this technology is a differential private clustering algorithm that provides strong privacy protection, preserves data properties important for subsequent analysis, and allows the party receiving the anonymized data to conduct analysis directly on anonymized data without the need of decryption or any extra processing. The research carried out was to design, implement and verify a solution to this problem by completing the following tasks: 1) developing the core technology; 2) developing a context based method that automatically recommends fields that must be anonymized; 3) conducted experiments showing superior results using our approach compared to existing tools, and 4) developed an intuitive but basic user interface. The research that was conducted generated novel algorithmic techniques that utilize state-of-the-art methods such as condensation, differential privacy preservation, clustering, automated tuning based on contextual awareness, and recommendation techniques to specify columns to users for anonymization leading to optimal privacy that allows research analysis on the dataset. Experiments were conducted to evaluate the efficacy of these novel algorithmic techniques by performing analysis on original non-anonymized datasets, then conducting analysis on the same yet anonymized datasets and comparing the results of the analyses. Overall, the anonymized analysis results were within 1% of the original results, verifying that the generated technology not only guarantees a high level of privacy but also enables research analysis as if it were conducted on the original dataset. Potential applications of this technology include anonymization of any type of structured network datasets that contain sensitive identifiers, such as IP addresses, that can be used in multiple applications. For example, to create an AI or machine learning model for cyber security, e.g., to detect attacks, or for performance analysis, e.g., identify bottlenecks or predict performance. In addition, a market analysis that was conducted for potential applications of this technology identified a broader range of applications of our anonymization technology beyond the network sector that includes healthcare, banking, insurance, securities, finance (FISB), data brokering, cloud services, ad sales, and government.

97 MATHEMATICS AND COMPUTING↗

Xanthos-Lake Dataset

The Xanthos-Lake v1.0 dataset provides the input data, trained machine-learning models, and simulation outputs needed to characterize lake water balance, snow and ice conditions, and mixing-layer temperature within the Xanthos global hydrological modeling framework. The dataset supports lake representation across a wide range of lake sizes and hydroclimatic conditions by combining xLSIM, a basin-specific machine-learning emulator of lake snow, ice, ice-cover fraction, and mixing-layer temperature, with the Xanthos-Lake water-balance model. The archive contains NetCDF datasets used to train and evaluate xLSIM, trained model weights, processed meteorological and lake-property inputs, and basin- and lake-category-specific simulation outputs. These materials are organized into four primary data groups, described below. Snowice_model_inputs: Contains the NetCDF input data used to train xLSIM. The xLSIM machine-learning framework uses three lake-based datasets. The meteorological forcing dataset provides monthly relative humidity, specific humidity, surface wind speed, maximum and minimum air temperature, downward longwave and shortwave radiation, snowfall, surface air pressure, and total precipitation. Lake surface area is included as an additional static predictor. The target-state dataset provides lake ice thickness, snow depth, snow cover, and lake mixing-layer temperature, while a companion lake-surface dataset provides the lake ice-cover fraction. Before training, ice thickness and snow depth are converted from meters to centimeters, mixing-layer temperature is converted from kelvin to degrees Celsius and constrained to nonnegative values, and ice-cover fraction is converted from a fraction to a percentage. The predictor variables are normalized using statistics calculated across the selected lakes and time steps. Snowice_model_outputs: Contains the NetCDF outputs generated by xLSIM. For each basin, xLSIM produces a file containing observed and predicted lake-state variables for the training, validation, and testing periods. The modeled variables include lake ice thickness, snow depth, snow cover, mixing-layer temperature, and lake ice-cover fraction. For basins without a sufficiently persistent snow-and-ice signal, the emulator predicts only mixing-layer temperature. The outputs also include training and validation loss histories, the selected model configuration, identifiers of the lakes used in training, and SHAP-based feature-importance information at the global, lake, and seasonal-regime levels. The trained machine-learning model weights are provided separately within the dataset archive. Together, these files support model evaluation and subsequent coupling with the Xanthos-Lake water-balance framework. XanthosLAKES: Contains the NetCDF input data used by the Xanthos-Lake framework. Monthly meteorological inputs include relative and specific humidity, downward shortwave and longwave radiation, mean, maximum, and minimum air temperature, wind speed, precipitation, snowfall, and surface air pressure. Static lake-property datasets provide lake identifiers, geographic locations, surface area, volume, mean depth, elevation, drainage area, fetch, outlet-routing information, and associated Xanthos grid-cell attributes. Separate bathymetric datasets provide the coefficients of the area–depth and volume–depth relationships for each aggregated lake unit. GLEV-based records provide observed lake surface area and evaporation data used to initialize lake states, define reference conditions, and calibrate and evaluate the model. Xanthos-Lake Outputs: Contains the basin- and lake-category-specific NetCDF outputs generated by Xanthos-Lake. Monthly variables include lake surface area, storage volume, outlet discharge, evaporation rate, evaporation volume, lake–groundwater exchange, lake inflow, ice thickness, snow depth, snow-cover fraction, ice-cover fraction, and mixing-layer temperature. The files also contain lake-specific calibration and validation statistics, including normalized root-mean-square error, mean absolute error, Nash–Sutcliffe efficiency, Kling–Gupta efficiency, and percent bias. Stored calibrated and derived parameters include the weir discharge coefficient, fractional freeboard, groundwater exchange coefficient, reference water level, corresponding reference surface area and storage volume, weir-width adjustment factor, and the fraction of routed inflow entering the lake. Basin identifiers, lake category, simulation period, calibration and validation periods, and parameter-schema information are retained as NetCDF metadata.

Abeshu, Guta [Pacific Northwest National Laborator↗

The impact of seasonality on multi-scale feature extraction techniques

A key component in constructing a broad-scale, gridded population dataset is fine resolution geospatial data accurately depicting the extent of human activity. Analogous datasets are often developed using a wide range of methods and classification techniques, including the use of spatial features, spectral features, or the coupling of both to identify the presence of man-made structures from high-resolution satellite imagery. By using spatial and textural-based descriptors to generate high-resolution settlement layers for two dissimilar regions at the peak of seasonal disparity, this study attempts to quantify the influence of seasonality on the accuracy of a supervised, multi-scale, feature extraction framework for automated delineation of human settlement. Results generated by numerous models are evaluated against a reference dataset allowing for assessment of seasonal and feature differences in the context of accuracy. Global or regional mapping of human settlement requires the assemblage of high-resolution satellite images with variegated acquisition characteristics (season, sun elevation, off-nadir, etc.) to produce a cloud-free composite image from which features are extracted. Results of this study suggest an emphasis on imagery criteria, in particular acquisition date, could improve classification accuracy when mapping human settlement at scale.

58 GEOSCIENCES↗

Securing Federated Learning Against Active Reconstruction Attacks

Federated Learning (FL) has amassed notable attention for its ability to preserve user privacy while emphasizing the retainment of model training efficiency. Due to this potential, FL has been integrated in many domains, such as healthcare, finance, law, and industrial engineering, where data cannot be easily exchanged due to sensitive information and strict privacy laws. However, current research has indicated that FL protocols are easily compromised by active data reconstruction attacks employed by actively dishonest servers. The malicious modification of global model parameters allows an actively dishonest server to obtain a direct copy of users’ private data via gradient inversion. Here, this class of attacks is highly underexplored and continues to be a major challenge due to the intense threat model. In this paper, we propose OASIS as a scalable and modality-agnostic defense based on data augmentation that counteracts active data reconstruction attacks while preserving model performance. To generalize our defense, we uncover the intuition behind gradient inversion that enables these attacks and theoretically establish the conditions by which the defense can be considered robust regardless of attack design. From this, we formulate our defense with data augmentation that illustrates its ability to undermine the attack principle. We evaluate OASIS on five real-world datasets–two image-based (ImageNet and CIFAR100) and three text-based (Wikitext, Stack Overflow, and Shakespeare)–which span diverse uses cases such as vision tasks and language modeling. Comprehensive evaluations on these datasets exhibit the efficacy of OASIS and highlight its feasibility as a solution.

97 MATHEMATICS AND COMPUTING↗

Implementation of a realistic artificial data generator for crash data generation

In this paper, a framework is outlined to generate realistic artificial data (RAD) as a tool for comparing different models developed for safety analysis. The primary focus of transportation safety analysis is on identifying and quantifying the influence of factors contributing to traffic crash occurrence and its consequences. The current framework of comparing model structures using only observed data has limitations. With observed data, it is not possible to know how well the models mimic the true relationship between the dependent and independent variables. Further, real datasets do not allow researchers to evaluate the model performance for different levels of complexity of the dataset. RAD offers an innovative framework to address these limitations. Hence, we propose a RAD generation framework embedded with heterogeneous causal structures that generates crash data by considering crash occurrence as a trip level event impacted by trip level factors, demographics, roadway and vehicle attributes. Within our RAD generator we employ three specific modules: (a) disaggregate trip information generation, (b) crash data generation and (c) crash data aggregation. For disaggregate trip information generation, we employ a daily activity-travel realization for an urban region generated from an established activity-based model for the Chicago region. We use this data of more than 2 million daily trips to generate a subset of trips with crash data. For trips with crashes crash location, crash type, driver/vehicle characteristics, and crash severity. The daily RAD generation process is repeated for generating crash records at yearly or multi-year resolution. In conclusion, the crash databases generated can be employed to compare frequency models, severity models, crash type and various other dimensions by facility type - possibly establishing a universal benchmarking system for alternative model frameworks in safety literature.

42 ENGINEERING↗

Characterizing uncertainty in Community Land Model version 5 hydrological applications in the United States

Abstract Land surface models such as the Community Land Model Version 5 (CLM5) are essential tools for simulating the behavior of the terrestrial system. Despite the extensive application of CLM5, limited attention has been paid to the underlying uncertainties associated with its hydrological parameters and how these uncertainties affect water resource applications. To address this long-standing issue, we use five meteorological datasets to conduct a comprehensive hydrological parameter uncertainty characterization of CLM5 over the hydroclimatic gradients of the conterminous United States. Key datasets produced from the uncertainty characterization experiment include: a benchmark dataset of CLM5 default hydrological performance, parameter sensitivities for 28 hydrological metrics, and large-ensemble outputs for CLM5 hydrological predictions. The presented datasets will assist CLM5 calibration and support broad applications, such as evaluating drought and flood vulnerabilities. The datasets can be used to identify the hydroclimatological conditions under which parametric uncertainties demonstrate substantial effects on hydrological predictions and clarify where further investigations are needed to understand how hydrological prediction uncertainties interact with other Earth system processes.

54 ENVIRONMENTAL SCIENCES↗

Towards robust surrogate models: Benchmarking machine learning approaches to expediting phase field simulations of brittle fracture

Data-driven approaches have the potential to make modeling complex, nonlinear physical phenomena significantly more computationally tractable. For example, computational modeling of fracture is a core challenge where machine learning techniques have the potential to provide a much needed speedup that would enable progress in areas such as multi-scale modeling and uncertainty quantification. Currently, phase field modeling (PFM) of fracture is one such approach that offers a convenient variational formulation to model crack nucleation, branching and propagation. To date, machine learning techniques have shown promise in approximating PFM simulations. While standard fracture benchmarks represent realistic scenarios frequently observed in practice, they typically do not provide sufficiently challenging tests for data-driven methods. Here, to address this gap, we introduce a challenging dataset based on PFM simulations designed to benchmark and advance ML methods for fracture modeling. This dataset includes three energy decomposition methods, two boundary conditions, and 1000 random initial crack configurations for a total of 6000 simulations. Each sample contains 100 time steps capturing the temporal evolution of the crack field. Alongside this dataset, we also implement and evaluate Physics Informed Neural Networks (PINN), Fourier Neural Operators (FNO), and UNet models as baselines, and explore the impact of ensembling strategies on prediction accuracy. With this combination of our dataset and baseline models drawn from the literature we aim to provide a standardized and challenging benchmark for evaluating machine learning approaches to solid mechanics. Our results highlight both the promise and limitations of popular current models, and demonstrate the utility of this dataset as a testbed for advancing machine learning in fracture mechanics research.

Benchmark dataset↗

Macromolecular crystallography with a large format CMOS detector

Recent advances in CMOS technology have allowed the production of large surface area detectors suitable for macromolecular crystallography experiments [1]. The Molecular Biology Consortium (MBC) Beamline 4.2.2 at the Advanced Light Source in Berkeley, CA, has installed a 2952 x 2820 mm RDI CMOS-8M detector with funds from NIH grant S10OD012073. The detector has a 20nsec dead pixel time and performs well with shutterless data collection strategies. The sensor obtains sharp point response and minimal optical distortion by use of a thin fiber-optic plate between the phosphor and sensor module. Shutterless data collections produce high-quality redundant datasets that can be obtained in minutes. The fine-sliced data are suitable for processing in standard crystallographic software packages (XDS, HKL2000, D*TREK, MOSFLM). Faster collection times relative to the previous CCD detector have resulted in a record number of datasets collected in a calendar year and de novo phasing experiments have resulted in publications in both Science and Nature [2,3]. The faster collections are due to a combination of the decreased overhead requirements of shutterless collections combined with exposure times that have decreased by over a factor of 2 for images with comparable signal to noise of the NOIR-1 detector. The overall increased productivity has allowed the development of new beamline capabilities and data collection strategies.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Distributed caching for processing raw arrays

As applications continue to generate multi-dimensional data at exponentially increasing rates, fast analytics to extract meaningful results is becoming extremely important. The database community has developed array databases that alleviate this problem through a series of techniques. In-situ mechanisms provide direct access to raw data in the original format---without loading and partitioning. Parallel processing scales to the largest datasets. In-memory caching reduces latency when the same data are accessed across a workload of queries. However, we are not aware of any work on distributed caching of multi-dimensional raw arrays. In this paper, we introduce a distributed framework for cost-based caching of multi-dimensional arrays in native format. Given a set of files that contain portions of an array and an online query workload, the framework computes an effective caching plan in two stages. First, the plan identifies the cells to be cached locally from each of the input files by continuously refining an evolving R-tree index. In the second stage, an optimal assignment of cells to nodes that collocates dependent cells in order to minimize the overall data transfer is determined. We design cache eviction and placement heuristic algorithms that consider the historical query workload. A thorough experimental evaluation over two real datasets in three file formats confirms the superiority -- by as much as two orders of magnitude -- of the proposed framework over existing techniques in terms of cache overhead and workload execution time.

Zhao, Weijie↗

What does fault tolerant Deep Learning need from MPI?

Deep Learning (DL) algorithms have become the {\em de facto} Machine Learning (ML) algorithm for large scale data analysis. DL algorithms are computationally expensive -- even distributed DL implementations which use MPI require days of training (model learning) time on commonly studied datasets. Long running DL applications become susceptible to faults -- requiring development of a fault tolerant system infrastructure, in addition to fault tolerant DL algorithms. This raises an important question: {\em What is needed from MPI for designing fault tolerant DL implementations?} In this paper, we address this problem for permanent faults. We motivate the need for a fault tolerant MPI specification by an in-depth consideration of recent innovations in DL algorithms and their properties, which drive the need for specific fault tolerance features. We present an in-depth discussion on the suitability of different parallelism types (model, data and hybrid); a need (or lack thereof) for check-pointing of any critical data structures; and most importantly, consideration for several fault tolerance proposals (user-level fault mitigation (ULFM), Reinit) in MPI and their applicability to fault tolerant DL implementations. We leverage a distributed memory implementation of Caffe, currently available under the Machine Learning Toolkit for Extreme Scale (MaTEx). We implement our approaches by extending MaTEx-Caffe for using ULFM-based implementation. Our evaluation using the ImageNet dataset and AlexNet neural network topology demonstrates the effectiveness of the proposed fault tolerant DL implementation using OpenMPI based ULFM.

Amatya, Vinay C.↗

Model Residuals as Shields: A Two-Level Formulation to Defend Smart Grids From Poisoning Attacks

The advancement of smart grids presents both vast opportunities and heightened cybersecurity risks. Data-driven defense mechanisms, though designed as a shield against these threats, can fall prey to poisoning attacks. We delve into regression settings, underscoring the imperative to fortify defenses against a spectrum of poison ratios, notably those above 0.5—an issue scarcely addressed in prior studies. Recognizing the susceptibilities of smart grids and their manipulable sensors, we exploit the very intent of poisoning attacks, compromising model accuracy, as our defense mechanism. Our proposed two-level optimization framework discerns between poisoned and authentic data based on model residuals, outperforming or matching existing methods in 72% to 77% of precision and 75% to 80% of recalls across various poisoning attacks, poison ratios, and datasets. Once the authentic data are identified, the trained model is adaptable for a variety of applications. Comprehensive evaluations on different smart grid datasets, pitted against myriad poisoning schemes, validate our methodology’s edge over existing methods. Here, we also shed light on the implications of model misspecification originating from temporal auto-correlation, a common feature in Internet of Things and smart grid data.

Adversarial machine learning (ML)↗

An algorithm to correct saturated mass spectrometry ion abundances for enhanced quantitation and mass accuracy in omic studies

The mass accuracy and peak intensity of ions detected by mass spectrometry (MS) measurements are essential to facilitate compound identification and quantitation. However, high concentration species can easily cause problems if their ion intensities reach beyond the limits of the detection system, leading to distorted and non-ideal detector response (e.g. saturation), and largely precluding the calculation of accurate m/z and intensity values. Here we present an open source computational method to correct peaks above a defined intensity (saturated) threshold determined by the MS instrumentation such as the analog-to-digital converters or time-to-digital converters used in conjunction with time-of-flight MS. In this method, the isotopic envelope for each observed ion above the saturation threshold is compared to its expected theoretical isotopic distribution. The most intense isotopic peak for which saturation does not occur is then utilized to re-calculate the precursor m/z and correct the intensity, resulting in both higher mass accuracy and greater dynamic range. The benefits of this approach were evaluated with proteomic and lipidomic datasets of varying complexities. After correcting the high concentration species, reduced mass errors and enhanced dynamic range were observed for both simple and complex omic samples. Specifically, the mass error dropped by more than 50% in most cases with highly saturated species and dynamic range increased by 1-2 orders of magnitude for peptides in a blood serum sample.

Environmental Molecular Sciences Laboratory↗

SU-F-I-12: Region-Specific Dictionary Learning for Low-Dose X-Ray CT Reconstruction

Purpose: Dictionary learning based method has attracted more and more attentions in low-dose CT due to the superior performance on suppressing noise and preserving structural details. Considering the structures and noise vary from region to region in one imaging object, we propose a region-specific dictionary learning method to improve the low-dose CT reconstruction. Methods: A set of normal-dose images was used for dictionary learning. Segmentations were performed on these images, so that the training patch sets corresponding to different regions can be extracted out. After that, region-specific dictionaries were learned from these training sets. For the low-dose CT reconstruction, a conventional reconstruction, such as filtered back-projection (FBP), was performed firstly, and then segmentation was followed to segment the image into different regions. Sparsity constraints of each region based on its dictionary were used as regularization terms. The regularization parameters were selected adaptively according to different regions. A low-dose human thorax dataset was used to evaluate the proposed method. The single dictionary based method was performed for comparison. Results: Since the lung region is very different from the other part of thorax, two dictionaries corresponding to lung region and the rest part of thorax respectively were learned to better express the structural details and avoid artifacts. With only one dictionary some artifact appeared in the body region caused by the spot atoms corresponding to the structures in the lung region. And also some structure in the lung regions cannot be recovered well by only one dictionary. The quantitative indices of the result by the proposed method were also improved a little compared to the single dictionary based method. Conclusion: Region-specific dictionary can make the dictionary more adaptive to different region characteristics, which is much desirable for enhancing the performance of dictionary learning based method.

60 APPLIED LIFE SCIENCES↗

zPerf: A Statistical Gray-Box Approach to Performance Modeling and Extrapolation for Scientific Lossy Compression

With the scaling up of simulation-based scientific discovery on high-performance computing systems, the disparity between compute and I/O has increased, forcing domain scientists to save only a small amount of simulation data to persistent storage. This can result in the loss of essential physics fields that are needed for data analysis. While error-bounded lossy compression has made tremendous progress in bridging the gap between compute and I/O, the lack of understanding of compression performance remains a key hurdle to its wide adoption. Here, in this work, we present zPerf, a statistical gray-box performance modeling approach for scientific lossy compression. Our contributions are threefold: 1) We develop zPerf to estimate the performance of lossy compression techniques, based on in-depth understanding and statistical modeling for data features and core compression metrics; 2) We demonstrate the in-detailed implementation of zPerf using two case studies, where we derive the performance modeling for SZ and ZFP, two leading lossy compressors; 3) We evaluate the effectiveness of zPerf on real-world datasets across various domains. Based on the evaluation, we demonstrate the efficacy of the zPerf performance model; 4) We further discuss three case studies where zPerf is applied to extrapolate the compression ratio of SZ and ZFP with alternative encoding schemes as well as ZFP with an alternative transform scheme. Through the case studies, we demonstrate the potential of zPerf for exploring the design space of lossy compression, which has hardly been studied in the literature.

97 MATHEMATICS AND COMPUTING↗

A kernel machine method for detecting higher order interactions in multimodal datasets: Application to schizophrenia

Technological advances are enabling us to collect multimodal datasets at an increasing depth and resolution while with decreasing labors. Understanding complex interactions among multimodal datasets, however, is challenging. Here in this study, we tested the interaction effect of multimodal datasets using a novel method called the kernel machine for detecting higher order interactions among biologically relevant multimodal data. Using a semiparametric method on a reproducing kernel Hilbert space, we formulated the proposed method as a standard mixed-effects linear model and derived a score-based variance component statistic to test higher order interactions between multimodal datasets. The method was evaluated using extensive numerical simulation and real data from the Mind Clinical Imaging Consortium with both schizophrenia patients and healthy controls. Our method identified 13-triplets that included 6 gene-derived SNPs, 10 ROIs, and 6 gene-specific DNA methylations that are correlated with the changes in hippocampal volume, suggesting that these triplets may be important for explaining schizophrenia-related neurodegeneration. The performance of the proposed method is compared with the following methods: test based on only first and first few principal components followed by multiple regression, and full principal component analysis regression, and the sequence kernel association test. With strong evidence (p-value ≤0.000001), the triplet (MAGI2, CRBLCrus1.L, FBXO28) is a significant biomarker for schizophrenia patients. This novel method can be applicable to the study of other disease processes, where multimodal data analysis is a common task.

59 BASIC BIOLOGICAL SCIENCES↗

DECONTAMINATION FACTORS AND FILTRATION FLUX IMPACT TO ARP AT REDUCED MST CONCENTRATION

Tank Farm and Closure Engineering is evaluating changes to the Actinide Removal Process facility operations to decrease the MST concentration from 0.4 g/L to 0.2 g/L and the contact time from 12 hours to between 6 and 8 hours. For this evaluation, SRNL reviewed previous datasets investigating the performance of MST at 0.2 g/L in salt solutions ranging from 4.5 to 7.5 M in sodium concentration. In general, reducing the MST concentration from 0.4 to 0.2 g/L and increasing the ionic strength from 4.5 to 7.5 M in sodium concentration will decrease the measured decontamination factors for plutonium, neptunium, uranium and strontium. The decontamination factors as well as single standard deviation values for each sorbate are reported. These values are applicable within the sorbate and sodium concentrations used in the experimental measurements. Decreasing the MST concentration in the ARP from 0.4 g/L to 0.2 g/L will produce an increase in the filter flux, and could lead to longer operating times between filter cleaning. The increase in flux is a function of a number of operating parameters, and is difficult to quantify. However, it is estimated that the reduction in MST could result in a reduction of filtration time of up to 20%.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Utilization and Outcomes of Breast Brachytherapy in Younger Women

Purpose: To directly compare (1) radiation treatment utilization patterns; (2) risks of subsequent mastectomy; and (3) costs of radiation treatment in patients treated with brachytherapy versus whole-breast irradiation (WBI), in a national, contemporary cohort of women with incident breast cancer, aged 64 years and younger. Methods and Materials: Using MarketScan health care claims data, we identified 45,884 invasive breast cancer patients (aged 18-64 years), treated from 2003 to 2010 with lumpectomy, followed by brachytherapy (n=3134) or whole-breast irradiation (n=42,750). We stratified patients into risk groups according to age (Age<50 vs Age≥50) and endocrine therapy status (Endocrine− vs Endocrine+). “Endocrine+” patients filled an endocrine therapy prescription within 1 year after lumpectomy. Pathologic hormone receptor status was not available in this dataset. In brachytherapy versus WBI patients, utilization trends and 5-year subsequent mastectomy risks were compared. Stratified, adjusted subsequent mastectomy risks were calculated using proportional hazards regression. Results: Brachytherapy utilization increased from 2003 to 2010: in patients Age<50, from 0.6% to 4.9%; patients Age≥50 from 2.2% to 11.3%; Endocrine− patients, 1.3% to 9.4%; Endocrine+ patients, 1.9% to 9.7%. Age influenced treatment selection more than endocrine status: 17% of brachytherapy patients were Age<50 versus 32% of WBI patients (P<.001); whereas 41% of brachytherapy patients were Endocrine–versus 44% of WBI patients (P=.003). Highest absolute 5-year subsequent mastectomy risks occurred in Endocrine−/Age<50 patients (24.4% after brachytherapy vs 9.0% after WBI (hazard ratio [HR] 2.18, 95% confidence interval [CI] 1.37-3.47); intermediate risks in Endocrine−/Age≥50 patients (8.6% vs 4.9%; HR 1.76, 95% CI 1.26-2.46); and lowest risks in Endocrine+ patients of any age: Endocrine+/Age<50 (5.5% vs 4.5%; HR 1.18, 95% CI 0.61-2.31); Endocrine+/Age≥50 (4.2% vs 2.4%; HR 1.71, 95% CI 1.16-2.51). Conclusion: In this younger cohort, endocrine status was a valuable discriminatory factor predicting subsequent mastectomy risk after brachytherapy versus WBI and therefore may be useful for selecting appropriate younger brachytherapy candidates.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

AUTOMATIC RETINA EXUDATES SEGMENTATION WITHOUT A MANUALLY LABELLED TRAINING SET

Diabetic macular edema (DME) is a common vision threatening complication of diabetic retinopathy which can be assessed by detecting exudates (a type of bright lesion) in fundus images. In this work, two new methods for the detection of exudates are presented which do not use a supervised learning step and therefore do not require ground-truthed lesion training sets which are time consuming to create, difficult to obtain, and prone to human error. We introduce a new dataset of fundus images from various ethnic groups and levels of DME which we have made publicly available. We evaluate our algorithm with this dataset and compare our results with two recent exudate segmentation algorithms. In all of our tests, our algorithms perform better or comparable with an order of magnitude reduction in computational time.

59 BASIC BIOLOGICAL SCIENCES↗