Low Matching Rate Research Articles

Using citizen science in the photo-identification of adult individuals of an amphibian based on two facial skin features.

Among amphibians, adults have traditionally been identified in capture-mark-recapture studies using invasive marking techniques with associated ethical, cost and logistical considerations. However, species in this group may be strong candidates for photo-identification based on natural skin features that removes many of these concerns, with this technique opening up opportunities for citizen scientists to be involved in animal monitoring programs. We investigated the feasibility of using citizen science to distinguish between individuals of an Australian anuran (the sandpaper frog, Lechriodus fletcheri) based on a visual analysis of their natural skin features. We collected photographs of marked individuals in the field over three breeding seasons using a smartphone device. This photo-database was used to create an online survey to determine how easily members of the general public could photo-match individuals by a comparison of two facial skin features; black banding that runs horizontally above the tympanum and a background array of tubercles present in this region. Survey participants were provided with 30 closed, multiple choice questions in which they were asked to match separate images of a query frog from small image pools of potential candidate matches. Participants were consistently able to match individuals with a low matching error rate (mean ± SD of 26 ± 5) despite the relatively low quality of photographs taken from a smartphone device in the field, with most query frogs being matched by a majority of participants (mean ± SD of 86.02 ± 9.52%). These features were found to be unique and stable among adult males and females. Thus, photo-identification is likely to be a valid, non-invasive method for capture-mark-recapture for L. fletcheri, and likely many anurans that display similar facial skin features. This may become an important alternative to artificial marking techniques, with the challenges of manual photo-matching reduced by spreading workloads among members of the public that can be recruited online.

Map Merging with Suppositional Box for Multi-Robot Indoor Mapping

For the map building of unknown indoor environment, compared with single robot, multi-robot collaborative mapping has higher efficiency. Map merging is one of the fundamental problems in multi-robot collaborative mapping. However, in the process of grid map merging, image processing methods such as feature matching, as a basic method, are challenged by low feature matching rate. Driven by this challenge, a novel map merging method based on suppositional box that is constructed by right-angled points and vertical lines is proposed. The paper firstly extracts right-angled points of suppositional box selected from the vertical point which is the intersection of the vertical line. Secondly, based on the common edge characteristics between the right-angled points, suppositional box in the map is constructed. Then the transformation matrix is obtained according to the matching pair of suppositional boxes. Finally, for matching errors based on the length of pairs, Kalman filter is used to optimize the transformation matrix. Experimental results show that this method can effectively merge map in different scenes and the successful matching rate is greater than that of other features.

A novel high precision mosaic method for sonar video sequence

The mosaic of sonar images is more difficult than the mosaic of traditional optical images due to their poor quality and the difficulty in extracting feature points. The existing mosaic methods of sonar images have a series of problems, such as low correct matching rate, large cumulative errors and high requirements for the quality of collected sonar images. In this paper, we proposed a high precision method to implement the mosaic of the underwater sonar video image sequence. Firstly, Accelerated Unsharp Masking (AUSM) algorithm is proposed to preprocess the original image. Then we extract KAZE feature points from preprocessed sonar images. A matching method combining multiple matching strategies with Progressive Sample Consensus (PROSAC) algorithm is followed to complete the image registration. Weighted fusion method and a region of interest (ROI) acquisition method based on the slope of right border is used to optimize the mosaicked image. Finally, we can obtain a high-quality panoramic image of underwater sonar video by a global mosaic strategy. Mosaic experiments on the sonar video image sequence collected by multi-beam sonar demonstrate that the proposed method in this paper can increase the number of feature points by about 70% and make the correct matching rate higher than 70%. The proposed method also has good robustness and the cumulative error during multi-image mosaic is less.

A hybrid classical techniques and optimal decision model for iris recognition under variable image quality conditions

One of the best biometrics used for human verification and identification is iris recognition. The contrast of its unique characteristics differs from one candidate to another in which the iris pattern has numerous well-known features like uniqueness texture, stability and compactness representation for human identification. Among these facts, several approaches in these areas are localized, but there is still an abundant problems such as the low match rate of the score level and low accuracy. Therefore, a decision model should be essential for iris recognition frameworks. This paper proposes a iris recognition based on a decision model using a unified framework based on the integration of three detection schemes due to variation occurred in shading and position change using Smallest Univalue Segment Assimilating Nucleus (SUSAN), Generalized Hough Transform (GHT) and Viola–Jones (SUSANGHT-VJ) for eye detection, enhancing the dimmer and darker areas using fuzzy retinex method and normalizing the iris boundary using daugman’s Rubber Sheet Model for segmentation. Also, the iris corner points is extracted using Gabor Wavelet Transform (GWT), the vector properties of the blurred texture features are quantized using Local Phase Quantization (LPQ) and the optimal decision model based on Atom Search Optimization (ASO) and Feed Forward Counter propagation Neural Network (FFCNN) for matching score level and classification task. Furthermore, the current framework prevents false matches and inappropriate iris input, thus making the iris match score framework more reliable. The evaluation of the proposed approach is trained and tested with the employed eye template of iris and face datasets. Therefore, the results depict that the proposed technique gives a high recognition rate of 99.9% on different datasets compared to existing methods.

Binary sequence matching system based on cross-phase modulation and four-wave mixing in highly nonlinear fibers

Photonic firewalls that can directly conduct the data discrimination and intrusion detection at the optical layer are an important security technique for optical networks. In the photonic firewall, binary sequence matching module is the core part and is implemented by multiple all-optical logic gates such as all-optical AND gate, all-optical XOR gate, etc. However, the semiconductor optical amplifier (SOA)-based all-optical logic gates have the disadvantages of low operation rate which leads to low matching rate. We design a binary sequence matching module that is implemented by the cross-phase modulation effect (XPM) and the four-wave mixing effect (FWM) of the highly nonlinear fiber (HNLF). The nonlinear effect of HNLF has a very short response time, thus it can realize the logic operation at a very high optical signal rate. The developed binary sequence matching module is verified through VPI Transmission Maker 8.5. The simulation results show that it can achieve the matching of 8-bit data in 64-bit binary data at 160-Gbps signal rate.

The Path to U.S. Neurosurgical Residency for Foreign Medical Graduates: Trends from a Decade 2007–2017

The increasing competitiveness of the neurosurgical residency match has made it progressively difficult for foreign medical graduates (FMGs) to match in neurosurgery. We compared FMG to U.S. medical graduate (USMG) match rates in neurosurgery and identified factors associated with match outcomes for FMGs in neurosurgery. Retrospective review of American Association of Neurological Surgeons membership data and Association of American Medical Colleges Charting the Outcomes match reports (2007-2017). Across 1857 neurosurgical residents (USMG: 91.1%, FMG: 8.9%), average FMG match rates were 24% (range, 15%-35%) versus 83% (range, 75%-94%; P < 0.001) for USMG. FMGs were more male (89.5% vs. 82.0%, P= 0.016), older (33.9 vs. 31.8 years, P= 0.008), and more likely to take research year(s) before matching (95.8% vs. 78.5%, P < 0.001). FMGs had greater publications (5 vs. 2, P < 0.001) and H-indices (3 vs. 1, P < 0.001). The number of matched USMGs increased by 3.3 annually, whereas that of matched FMGs remained unchanged (β= 0.07). Compared with USMGs, FMGs were less likely to match to National Institutes of Health (NIH) Top 40 (32.7% vs. 47.5%, P < 0.001) and Doximity Top 20 (20.0% vs. 29.0%, P= 0.014) programs. FMGs with prior U.S. neurosurgery program affiliation were more likely to match at NIH and Doximity Top 20 programs (P < 0.05). For NIH programs, FMGs were older (35.3 vs. 32.0, P= 0.011), had higher H-indices (5 vs. 2, P < 0.001), publications (7 vs. 2, P < 0.001), and were more likely to take research year(s) (94.4% vs. 76.0%, P= 0.002) than USMGs. FMGs had similar patterns for matching into Doximity Top 20 programs. Although FMGs have lower match rates into U.S. neurosurgery residencies than USMGs, several demographic, professional, and academic factors could increase the chances of successful FMG neurosurgical match.

Challenges and future of precision medicine strategies for breast cancer based on a database on drug reactions.

Breast cancer (BC) is a malignancy with the highest incidence in women. Great progress has been made in research related to traditional precision medicine for BC. However, many reports have suggested that patients with BC have not benefited a lot from such progress. Thus, we analyze traditional precision medicine strategies for BC, sum up their limitations and challenges, and preliminarily propose future orientations of precision medicine strategies based on a database on drug reaction of patients with BC. According to related research, traditional precision medicine strategies for BC, which are based on molecular subtypes, perform pertinent treatments, new drug research and development according to molecular typing results. Nevertheless, these strategies still have some deficiencies. First, there are very few patients with each molecular subtype, the match ratio of drugs is low. Second, these strategies can not solve the problem of poor drug sensitivity resulting from heterogeneity. The main strategy we put forward in the present paper is based on patients’ varying drug reactions. Focusing on treating existing patients and maximizing the utilization of existing drugs, it is expected to not have deficiencies of traditional precision medicine for BC, including low match rate and poor therapeutic efficacy arising from tumor heterogeneity of BC.

Cross-Matching VIIRS Boat Detections with Vessel Monitoring System Tracks in Indonesia

A methodology had been proposed for cross-matching visible infrared imaging radiometer suite (VIIRS) boat detections (VBD) with vessel monitoring system (VMS) tracks. The process involves predicting the probable location of VMS vessels at the time of each VIIRS data collection with an orbital model. Thirty-two months of Indonesian VMS data was segmented into fishing and transit activity types and then cross-matched with the VBD record. If a VBD record is found within 700 m and 5 s of the predicted location, it is marked as a match. The cross-matching indicates that 96% of the matches occur while the vessel is fishing. Small pelagic purse seiners account for 27% of the matches. Other gear types with high match rates include hand line tuna, squid dip net, squid jigging, and large pelagic purse seiners. Low match rates were found for gillnet, trawlers, and long line tuna. There is an indication that VMS vessels using submersible lights can be identified based on consistently low average radiances and match rates under 45%. Overall, VBD numbers exceed VMS vessel numbers in Indonesia by a nine to one ratio, indicating that VIIRS detects large numbers of fishing boats under the 30 Gross Tonnage (GT) level set for the VMS requirement. The cross-matching could be used to identify “dark” vessels that lack automatic identification system (AIS) or VMS.

Axon Registry® data validation: Accuracy assessment of data extraction and measure specification.

To conduct a data validation study encompassing an accuracy assessment of the data extraction process for the Axon Registry®. Data elements were abstracted from electronic health records (EHRs) by an external auditor (IQVIA) using virtual site visits at participating sites. IQVIA independently calculated Axon Registry quality measure performance rates based on American Academy of Neurology measure specifications and logic using Axon Registry data. Agreement between Axon Registry and IQVIA data elements and measure performance rates was calculated. Discordance was investigated to elucidate underlying systemic or idiosyncratic reasons for disagreement. Nine sites (n = 720 patients; n = 80 patients per site) with diversity among EHR vendor, practice settings, size, locations, and data transfer method were included. There was variable concordance between the data elements in the Axon Registry and those abstracted independently by IQVIA; high match rates (≥92%) were observed for discrete elements (e.g., demographics); lower match rates (<44%) were observed for elements with free text (e.g., plan of care). Across all measures, there was a 76% patient-level measure performance agreement between Axon Registry and IQVIA (κ = 0.53, p < 0.001). There was a range of concordance between data elements and quality measures in the Axon Registry and those independently abstracted and calculated by an independent vendor. Validation of data and processes is important for the Axon Registry as a clinical quality data registry that utilizes automated data extraction methods from the EHR. Implementation of remediation strategies to improve data accuracy will support the ability of the Axon Registry to perform accurate quality reporting.

Abstract WP316: Assessing the Accuracy of a Linkage Between the Michigan Emergency Medical Services Information System and the Michigan Coverdell Acute Stroke Registry

Background: Michigan’s Ongoing Stroke Registry to Accelerate Improvement of Care (MOSAIC) supports expansion of statewide stroke systems of care. The Michigan EMS Information System (Mi-EMSIS) is a repository for EMS data developed to standardize collection, storage, sharing and analysis of data. To provide feedback to EMS providers concerning stroke patient identification, care, and outcomes, development of statewide linkage between hospital and EMS registry data is needed. Objectives: Develop data linkage between MOSAIC and MiEMSIS stroke data and assess completeness and accuracy of match results. Method: All patients transported by EMS in 2017 with an impression of stroke (i.e. suspected stroke) were linked to patients who were discharged with confirmed stroke or TIA from 33 participating MOSAIC hospitals with successful linkage defined as suspected stroke transport having confirmed stroke upon discharge. Records were deterministically matched using four variables: receiving hospital code, hospital admission date, date of birth, and sex. Additionally, performance on EMS stroke quality measures and hospital outcomes were calculated. Results: 197 EMS agencies transported 4,580 suspected stroke patients to MOSAIC hospitals. Of the 11,352 confirmed stroke cases discharged from MOSAIC hospitals, 4,924 (43.4%) were documented to have arrived via EMS. However, only 1,287 (26.1%) cases were successfully linked to the MiEMSIS data. Of the 4,580 suspected stroke transports 1,287 (28.1%) were confirmed as acute stroke cases at discharge. Conclusions: A deterministic linkage between Coverdell stroke cases and EMS suspected strokes from a state EMS registry resulted in low match rates, which may imply low rates of EMS stroke recognition or high rates of match failure. Alternative linkage methods such as probabilistic matching and inclusion of all EMS transports should be explored.

A resource sharing (sharing platform) scheme on online taxi services

This paper will review the match between single driver and single rider in online taxi services through a resource sharing (sharing platform) for the operators with the objectives to maximize the profit for drivers (operators) and minimize waiting time for passengers so that the matching rate is higher. A low matching rate between rider and driver can cause the consumer to drop the services. The matching between single driver and single rider in online taxi services through a sharing platform scheme is formulated in maximum weighted bipartite matching problem. To solve the proposed model, we use Kuhn Munkres Algorithm, while to solve the shortest path for the driver to pick up the passenger and the shortest path of passenger's origin destination, modified Dijkstra with adaptive algorithm based on Wei Peng et.al (2012) is used. Based on illustrative example with several cases, we found a resource sharing scenario can optimize the matching between driver and rider and moreover can solve the surge pricing problem which is deemed as less transparant to customer

A Hardware-Accelerated Solution for Hierarchical Index-Based Merge-Join

Hardware acceleration through field programmable gate arrays (FPGAs) has recently become a technique of growing interest for many data-intensive applications. Join query is one of the most fundamental database query types useful in relational database management systems. However, the available solutions so far have been beset by higher costs in comparison to other query types. In this paper, we develop a novel solution to accelerate the processing of sort-merge join queries with low match rates. Specifically, our solution makes use of hierarchical indexes to identify result-yielding regions in the solution space in order to take advantage of result sparseness. Further, in addition to one-dimensional equi-join query processing, our solution supports processing of multidimensional similarity join queries. Experimental results show that our solution is superior to the best existing method in a low match rate setting; the method achieves a speedup factor of 4.8 for join queries with a match rate of 5 percent.

SKA-PS: Secure key agreement protocol using physiological signals

In this paper, we propose SKA-PS, a novel Secure Key Agreement protocol using Physiological Signals, for Body Area Networks (BANs). Our protocol generates symmetric cryptographic keys using the physiological parameters derived from the physiological signals of the users, such as electrocardiogram, photoplethysmogram and blood pressure. In our construction, we reduce the problem of secure key agreement into the problem of set reconciliation by representing the physiological parameter sequences generated from the physiological signals of the BAN users with appropriate sets. When properly selected parameters are applied, biosensors of the same BAN user can agree on symmetric cryptographic keys with remarkably high true match and low false match rates (as much as 100% and 0.46% for pairwise execution, and 100% and 0.26% for group execution, respectively), and low communication, computational and storage costs. We implemented our model in an embedded system, thus the results show real implementation outcomes. Moreover, we comparatively analyze the performance of SKA-PS with two existing bio-cryptographic key agreement protocols and show that SKA-PS outperforms both in all performance metrics.

Characteristics of schools of pharmacy associated with a successful PGY1 residency match

IntroductionThe literature is limited related to school of pharmacy (SOP) characteristics that may impact a student's residency match success. Given the paucity of information, the aim of our study was to examine the association between multiple school specific characteristics and Post-Graduate Year 1 (PGY1) residency match rate. MethodsPGY1 residency match data and SOP specific characteristics were obtained. The median aggregate match rate was used to categorize programs as high or low match rate. Univariate and multivariate analyses were conducted to assess the impact of SOP characteristics on match rate. ResultsMedian match rates were 64% (2013), 63.25% (2014), and 64.25% (2015). Schools with >95% North American Pharmacist Licensure Examination® (NAPLEX®) pass rate and >15 funded faculty were more likely to have a high match rate. Private schools were less likely to have a high match rate when compared to public programs. ConclusionSOP characteristics associated with a high likelihood of a successful PGY1 residency match were greater than 95% of NAPLEX® pass rate, more than 15 funded faculty members, and public school status.

Iris Recognition With Off-the-Shelf CNN Features: A Deep Learning Perspective

Iris recognition refers to the automated process of recognizing individuals based on their iris patterns. The seemingly stochastic nature of the iris stroma makes it a distinctive cue for biometric recognition. The textural nuances of an individual’s iris pattern can be effectively extracted and encoded by projecting them onto Gabor wavelets and transforming the ensuing phasor response into a binary code - a technique pioneered by Daugman. This textural descriptor has been observed to be a robust feature descriptor with very low false match rates and low computational complexity. However, recent advancements in deep learning and computer vision indicate that generic descriptors extracted using convolutional neural networks (CNNs) are able to represent complex image characteristics. Given the superior performance of CNNs on the ImageNet large scale visual recognition challenge and a large number of other computer vision tasks, in this paper, we explore the performance of state-of-the-art pre-trained CNNs on iris recognition. We show that the off-the-shelf CNN features, while originally trained for classifying generic objects, are also extremely good at representing iris images, effectively extracting discriminative visual features and achieving promising recognition results on two iris datasets: ND-CrossSensor-2013 and CASIA-Iris-Thousand. We also discuss the challenges and future research directions in leveraging deep learning methods for the problem of iris recognition.

Selection of cassava accessions with multiple resistance to pathogens associated with root rot disease

Cassava root rot disease (CRRD) severely affects productivity in several countries. The objective of this study was to estimate genetic parameters and to identify multiple resistance sources against pathogens associated with CRRD. The symptoms caused by Fusarium spp., Phytophthora spp., and Botryosphaeriaceae species in peel and pulp from the roots were evaluated in 277 accessions using a whole tuberous root inoculation method. The resistance data were obtained by REML/BLUP (restricted maximum likelihood/best linear unbiased predictor). The classic selection index (CI), multiplicative (MI), and sum of ranks (SRI) were used to identify the accessions with multiple resistance. For all pathogens, the environmental variance ( $$\sigma_{e}^{2}$$ ) was the most important component. Individual heritability $$\left( {h_{g}^{2} } \right)$$ was of low magnitude for resistance to most pathogens (0.16 ± 0.02—peel and 0.31 ± 0.03—pulp for Fusarium spp.; 0.26 ± 0.03—peel and 0.30 ± 0.03—pulp for Phytophthora spp.; and 0.28 ± 0.03—peel and 0.27 ± 0.03—pulp for Botryosphaereacea species). The distribution of CRRD symptoms indicated the presence of quantitative inheritance. The direct selection of the 15 more resistant accessions based on the genotypic predicted values result in high reductions of disease (>50%). However, there was a low matching rate of the most resistant accessions for each pathogen and the different parts of the tuberous roots (peel and pulp). The CI and MI were the most promising compared to the SRI to ensure high and balanced resistance for each pathogen. Understanding the genetic basis of resistance to CRRD and the identification of sources with multiple resistance may be useful in various management strategies to control the disease.

Recommendation system with Automated Web Usage data mining using K-Nearest Neighbor(KNN) classification

The major problem of many on-line web sites is the presentation of many choices to the various clients at a time. This usually results into time consumingtask in finding outthe right product or information on the site. The user’s current interest depends upon the navigational behavior which helps the organizations to guideusers in their browsing activities and obtain some relevant information in a short span of time. Since, the resulting patterns which are obtained through data mining techniques did not perform well in the prediction of future browsing patterns because of the low matching rate of resulting rules and of user’s browsing behavior. This paper focuses on the study of the automatic web usage data mining and recommendation system which is based on current user behavior through his/her click stream data. The K-Nearest-Neighbor (KNN) classification method has been trained to be used in real-time and on-line to identify clients and visitors click stream data, matching it to a particular user group and recommends a tailored browsing option that meet the needs of the specific user at particular time.

Modeling developmental growth stages in mango on the basis of absolute vs. relative growth rates

Plant development, the change in plant structure, and growth, the increase of organ size over time are often studied as two separate processes. The concept of developmental growth stages has been recently proposed in mango to integrate both morphological traits and growth measurements. These stages rely on a strong match between developmental (i.e., phenological) stages and growth stages deduced from absolute growth rate (AGR) sequences. In this study, we explore growth stages deduced from relative growth rate (RGR) sequences. For a given mango cultivar (Cogshall or Jose) and either growth units (GUs) or inflorescences, a multistage model based on RGR sequences was first built. Growth stages deduced from this model were then compared with developmental stages determined visually. For GUs, the match rates between RGR-based stages and developmental stages were similar to the match rates between AGR-based stages and developmental stages, because of the rich information provided by the four organs modeled (the axis and three selected leaves). The match rates were far lower for the inflorescences where only the main axis was modeled. This is related to the fact that, compared to AGRs, RGRs amplify the variations at the beginning of growth of an organ while damping the variations at the end of growth. This explains the low match rates at the end of inflorescence growth. In plant growth studies, RGRs are often preferred to AGRs with the objective to take account of sink strength and to abstract from the effect of organ size. This latter argument is not relevant in our case since our probabilistic multistage model intrinsically abstracts from the organ size and only relies on the overlap of growth rate distributions for successive stages. (Resume d'auteur)

Combining deterministic and probabilistic matching to reduce data linkage errors in hospital administrative data

ABSTRACTObjectivesData linkage algorithms are used to link together multiple episodes of care belonging to the same patient. For example, the HESID algorithm is used to generate Hospital Episode Statistics (HES) in England. HESID is a deterministic algorithm, requiring identifiers to agree or disagree at each step. Data linkage errors occur when episodes belonging to two patients are incorrectly linked (a false match) or when episodes belonging to the same patient are not linked (a missed match). This typically occurs because patient identifiers (e.g. NHS number, postcode) contain errors or have missing data. We previously showed that HESID has a low false match rate (0.2%) but a high missed match rate (4.1%) when applied to paediatric intensive care data. This biased the true readmission rate, particularly for some patient groups including ethnic minorities. The aim of our study was to evaluate whether an additional step involving probabilistic matching would lower the missed match rate in HES without increasing the false matched rate.&#x0D; ApproachWe simulated three datasets having the same characteristics as HES, for three age groups expected to have different levels of postcode stability (at age 0/1, 5/6 and 18/19). We compared the deterministic algorithm to a probabilistic algorithm, and then to a deterministic algorithm with an additional probabilistic step. In sensitivity analyses, we evaluated the algorithms under different data quality scenarios.&#x0D; ResultsResults show that deterministic followed by probabilistic matching is the best solution for reducing missed matches, particularly in scenarios where errors in patient identifiers are more common.&#x0D; ConclusionData linkage algorithms need to be evaluated against good quality reference standard data sets. For hospital data in England, the Personal Demographics Service (PDS) could be used to evaluate our approach, because it contains many of the same patient identifiers used in HES. Reducing data linkage error will improve monitoring of hospital activity in England.

Abstract CS1-2: CS1-2 Clinical utility of multi-gene testing in metastatic breast cancer

Abstract The goal of many ongoing precision medicine initiatives is to match patients to targeted treatment(s) based upon the identification of specific driver mutations in their tumors. Although large-scale research projects have elucidated the genomic landscape of breast cancer, they have not provided insight into the clinical utility of genomic testing. Institutional studies of small, targeted sequencing panel-based testing approaches involving patients with advanced, metastatic disease demonstrate that only a small proportion of patients ultimately receive genotype-matched treatment. The reasons for this low treatment matching rate include: institutional and/or geographic barriers to access to genotype-matched clinical trials; patient deterioration after testing results are reported; and the limited range of somatic mutations that are “druggable” with approved or investigational drug therapies. While there are promising signals of activity from early phase clinical trials with novel drug treatments targeting specific genomically defined subpopulations - including AKT1, ERBB2, ESR1 and PIK3CA mutations as well as FGFR1 amplification - and advances in circulating tumor DNA technology may allow for greater integration at the point of care, further evidence is needed before multi-gene testing can be recommended in routine clinical practice. Citation Format: Bedard P. CS1-2 Clinical utility of multi-gene testing in metastatic breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr CS1-2.