Search

Image restoration tasks have witnessed great performance improvement in recent years by developing large deep models. Despite the outstanding performance, the heavy computation demanded by the deep models has restricted the application of image restoration. To lift the restriction, it is required to reduce the size of the networks while maintaining accuracy. Recently, N:M structured pruning has appeared as one of the effective and practical pruning approaches for making the model efficient with the accuracy constraint. However, it fails to account for different computational complexities and performance requirements for different layers of an image restoration network. To further optimize the trade-off between the efficiency and the restoration accuracy, we propose a novel pruning method that determines the pruning ratio for N:M structured sparsity at each layer. Extensive experimental results on super-resolution and deblurring tasks demonstrate the efficacy of our method which outperforms previous pruning methods significantly. PyTorch implementation for the proposed methods will be publicly available at https://github.com/JungHunOh/SLS_CVPR2022., Accepted to CVPR 2022

Retinal artery&ndash, vein (AV) classification is a prerequisite for quantitative analysis of retinal vessels, which provides a biomarker for neurologic, cardiac, and systemic diseases, as well as ocular diseases. Although convolutional neural networks have presented remarkable performance on AV classification, it often comes with a topological error, like an abrupt class flipping on the same vessel segment or a weakness for thin vessels due to their indistinct appearances. In this paper, we present a new method for AV classification where the underlying vessel topology is estimated to give consistent prediction along the actual vessel structure. We cast the vessel topology estimation as iterative vascular connectivity prediction, which is implemented as deep-learning-based pairwise classification. In consequence, a whole vessel graph is separated into sub-trees, and each of them is classified as an artery or vein in whole via a voting scheme. The effectiveness and efficiency of the proposed method is validated by conducting experiments on two retinal image datasets acquired using different imaging techniques called DRIVE and IOSTAR.

3D shape representation and its processing have substantial effects on 3D shape recognition. The polygon mesh as a 3D shape representation has many advantages in computer graphics and geometry processing. However, there are still some challenges for the existing deep neural network (DNN)-based methods on polygon mesh representation, such as handling the variations in the degree and permutations of the vertices and their pairwise distances. To overcome these challenges, we propose a DNN-based method (PolyNet) and a specific polygon mesh representation (PolyShape) with a multi-resolution structure. PolyNet contains two operations; (1) a polynomial convolution (PolyConv) operation with learnable coefficients, which learns continuous distributions as the convolutional filters to share the weights across different vertices, and (2) a polygonal pooling (PolyPool) procedure by utilizing the multi-resolution structure of PolyShape to aggregate the features in a much lower dimension. Our experiments demonstrate the strength and the advantages of PolyNet on both 3D shape classification and retrieval tasks compared to existing polygon mesh-based methods and its superiority in classifying graph representations of images. The code is publicly available from https://myavartanoo.github.io/polynet/.

Most image super-resolution (SR) methods are developed on synthetic low-resolution (LR) and high-resolution (HR) image pairs that are constructed by a predetermined operation, e.g., bicubic downsampling. As existing methods typically learn an inverse mapping of the specific function, they produce blurry results when applied to real-world images whose exact formulation is different and unknown. Therefore, several methods attempt to synthesize much more diverse LR samples or learn a realistic downsampling model. However, due to restrictive assumptions on the downsampling process, they are still biased and less generalizable. This study proposes a novel method to simulate an unknown downsampling process without imposing restrictive prior knowledge. We propose a generalizable low-frequency loss (LFL) in the adversarial training framework to imitate the distribution of target LR images without using any paired examples. Furthermore, we design an adaptive data loss (ADL) for the downsampler, which can be adaptively learned and updated from the data during the training loops. Extensive experiments validate that our downsampling model can facilitate existing SR methods to perform more accurate reconstructions on various synthetic and real-world examples than the conventional approaches., Accepted at TPAMI

3D Shape representation has substantial effects on 3D shape reconstruction. Primitive-based representations approximate a 3D shape mainly by a set of simple implicit primitives, but the low geometrical complexity of the primitives limits the shape resolution. Moreover, setting a sufficient number of primitives for an arbitrary shape is challenging. To overcome these issues, we propose a constrained implicit algebraic surface as the primitive with few learnable coefficients and higher geometrical complexities and a deep neural network to produce these primitives. Our experiments demonstrate the superiorities of our method in terms of representation power compared to the state-of-the-art methods in single RGB image 3D shape reconstruction. Furthermore, we show that our method can semantically learn segments of 3D shapes in an unsupervised manner. The code is publicly available from https://myavartanoo.github.io/3dias/ ., Published at ICCV 2021

Background: We aimed to report the prevalence and correlates of high-risk alcohol consumption and types of alcoholic beverages. Methods: The baseline data of the Health Examinees-Gem (HEXA-G) study participants, including 43,927 men and 85,897 women enrolled from 2005 through 2013, were used for analysis. Joinpoint regression was performed to estimate trends in the age-standardized prevalence of alcohol consumption. Associations of demographic and behavioral factors, perceived health-related effects, social relationships, and the diagnostic history of diseases with alcohol consumption were assessed using multinomial logistic regression. Results: The prevalence of alcohol consumption remained higher in men during the study period and increased in women. The amount of alcohol consumed has increased in women, especially that from beer and makgeolli, a traditional Korean fermented rice wine. Older participants were less likely to be high-risk drinkers (men and women who drink more than 40 or 20 g/day of alcohol, respectively) and drink Soju, a distilled liquor, and beer, and more likely to drink makgeolli. Educational level was negatively associated with high-risk drinking. However, it was positively associated with the consumption of strong spirits and wine. Smoking was associated with high-risk drinking and the consumption of soju and strong spirits. Engaging in regular exercise and having stress were associated with drinking all types of beverages except for soju. Conclusions: Sex-specific trends in alcohol consumption were influenced by demographic, behavioral, and perceived health-related factors. The findings will help improve the understanding of alcohol-related problems and provide evidence for establishing country-specific policies and campaigns in Korea.

Deep CNNs have achieved significant successes in image processing and its applications, including single image super-resolution (SR). However, conventional methods still resort to some predetermined integer scaling factors, e.g., x2 or x4. Thus, they are difficult to be applied when arbitrary target resolutions are required. Recent approaches extend the scope to real-valued upsampling factors, even with varying aspect ratios to handle the limitation. In this paper, we propose the SRWarp framework to further generalize the SR tasks toward an arbitrary image transformation. We interpret the traditional image warping task, specifically when the input is enlarged, as a spatially-varying SR problem. We also propose several novel formulations, including the adaptive warping layer and multiscale blending, to reconstruct visually favorable results in the transformation process. Compared with previous methods, we do not constrain the SR model on a regular grid but allow numerous possible deformations for flexible and diverse image editing. Extensive experiments and ablation studies justify the necessity and demonstrate the advantage of the proposed SRWarp method under various transformations., Accepted to CVPR 2021

Whole-body 3D human mesh estimation aims to reconstruct the 3D human body, hands, and face simultaneously. Although several methods have been proposed, accurate prediction of 3D hands, which consist of 3D wrist and fingers, still remains challenging due to two reasons. First, the human kinematic chain has not been carefully considered when predicting the 3D wrists. Second, previous works utilize body features for the 3D fingers, where the body feature barely contains finger information. To resolve the limitations, we present Hand4Whole, which has two strong points over previous works. First, we design Pose2Pose, a module that utilizes joint features for 3D joint rotations. Using Pose2Pose, Hand4Whole utilizes hand MCP joint features to predict 3D wrists as MCP joints largely contribute to 3D wrist rotations in the human kinematic chain. Second, Hand4Whole discards the body feature when predicting 3D finger rotations. Our Hand4Whole is trained in an end-to-end manner and produces much better 3D hand results than previous whole-body 3D human mesh estimation methods. The codes are available here at https://github.com/mks0601/Hand4Whole_RELEASE., Published at CVPRW 2022

Catastrophic forgetting is a critical challenge in training deep neural networks. Although continual learning has been investigated as a countermeasure to the problem, it often suffers from the requirements of additional network components and the limited scalability to a large number of tasks. We propose a novel approach to continual learning by approximating a true loss function using an asymmetric quadratic function with one of its sides overestimated. Our algorithm is motivated by the empirical observation that the network parameter updates affect the target loss functions asymmetrically. In the proposed continual learning framework, we estimate an asymmetric loss function for the tasks considered in the past through a proper overestimation of its unobserved sides in training new tasks, while deriving the accurate model parameter for the observable sides. In contrast to existing approaches, our method is free from the side effects and achieves the state-of-the-art accuracy that is even close to the upper-bound performance on several challenging benchmark datasets., ICCV 2019

Few-shot learning is a challenging problem where the goal is to achieve generalization from only few examples. Model-agnostic meta-learning (MAML) tackles the problem by formulating prior knowledge as a common initialization across tasks, which is then used to quickly adapt to unseen tasks. However, forcibly sharing an initialization can lead to conflicts among tasks and the compromised (undesired by tasks) location on optimization landscape, thereby hindering the task adaptation. Further, we observe that the degree of conflict differs among not only tasks but also layers of a neural network. Thus, we propose task-and-layer-wise attenuation on the compromised initialization to reduce its influence. As the attenuation dynamically controls (or selectively forgets) the influence of prior knowledge for a given task and each layer, we name our method as L2F (Learn to Forget). The experimental results demonstrate that the proposed method provides faster adaptation and greatly improves the performance. Furthermore, L2F can be easily applied and improve other state-of-the-art MAML-based frameworks, illustrating its simplicity and generalizability., CVPR 2020. Code at https://github.com/baiksung/L2F

Multi-person pose estimation from a 2D image is an essential technique for human behavior understanding. In this paper, we propose a human pose refinement network that estimates a refined pose from a tuple of an input image and input pose. The pose refinement was performed mainly through an end-to-end trainable multi-stage architecture in previous methods. However, they are highly dependent on pose estimation models and require careful model design. By contrast, we propose a model-agnostic pose refinement method. According to a recent study, state-of-the-art 2D human pose estimation methods have similar error distributions. We use this error statistics as prior information to generate synthetic poses and use the synthesized poses to train our model. In the testing stage, pose estimation results of any other methods can be input to the proposed method. Moreover, the proposed model does not require code or knowledge about other methods, which allows it to be easily used in the post-processing step. We show that the proposed approach achieves better performance than the conventional multi-stage refinement models and consistently improves the performance of various state-of-the-art pose estimation methods on the commonly used benchmark. The code is available in this https URL\footnote{\url{https://github.com/mks0601/PoseFix_RELEASE}}., Published at CVPR 2019

Most of the existing deep learning-based methods for 3D hand and human pose estimation from a single depth map are based on a common framework that takes a 2D depth map and directly regresses the 3D coordinates of keypoints, such as hand or human body joints, via 2D convolutional neural networks (CNNs). The first weakness of this approach is the presence of perspective distortion in the 2D depth map. While the depth map is intrinsically 3D data, many previous methods treat depth maps as 2D images that can distort the shape of the actual object through projection from 3D to 2D space. This compels the network to perform perspective distortion-invariant estimation. The second weakness of the conventional approach is that directly regressing 3D coordinates from a 2D image is a highly non-linear mapping, which causes difficulty in the learning procedure. To overcome these weaknesses, we firstly cast the 3D hand and human pose estimation problem from a single depth map into a voxel-to-voxel prediction that uses a 3D voxelized grid and estimates the per-voxel likelihood for each keypoint. We design our model as a 3D CNN that provides accurate estimates while running in real-time. Our system outperforms previous methods in almost all publicly available 3D hand and human pose estimation datasets and placed first in the HANDS 2017 frame-based 3D hand pose estimation challenge. The code is available in https://github.com/mks0601/V2V-PoseNet_RELEASE., HANDS 2017 Challenge Frame-based 3D Hand Pose Estimation Winner (ICCV 2017), Published at CVPR 2018

The objective of this study was to verify a change in the longitudinal trend of blood lead levels for the Korean population, before and after the regulation of leaded gasoline- which occurred between 1987 and 1993 in Korea. A total of 77 reports on blood lead levels among general Korean population between 1981 and 2014 were selected, and the results were summarized to have the variables of year, number of subjects, the subjects' range in age, gender, and blood lead concentrations (arithmetic mean). The annual average atmospheric lead levels for four major cities (i.e., Seoul, Busan, Daegu and Gwangju) were collected from the Air Pollution Monitoring Database from 1991, and pilot studies from 1985 to 1990 before the national air quality monitoring system was launched in 1991. Blood lead levels were visualized in a bubble plot in which the size of each bubble represented the sample size of each study, and the annual average concentrations in ambient air were depicted on line graphs. Blood lead levels in the Korean population tended to gradually increase from the early 1980s (approximately 15-20 μg/dL) until 1990-1992 (20-25 μg/dL). Blood lead levels then began to rapidly decrease until 2014 (

State-of-the-art video deblurring methods are capable of removing non-uniform blur caused by unwanted camera shake and/or object motion in dynamic scenes. However, most existing methods are based on batch processing and thus need access to all recorded frames, rendering them computationally demanding and time consuming and thus limiting their practical use. In contrast, we propose an online (sequential) video deblurring method based on a spatio-temporal recurrent network that allows for real-time performance. In particular, we introduce a novel architecture which extends the receptive field while keeping the overall size of the network small to enable fast execution. In doing so, our network is able to remove even large blur caused by strong camera shake and/or fast moving objects. Furthermore, we propose a novel network layer that enforces temporal consistency between consecutive frames by dynamic temporal blending which compares and adaptively (at test time) shares features obtained at different time steps. We show the superiority of the proposed method in an extensive experimental evaluation., 10 pages

Non-uniform blind deblurring for general dynamic scenes is a challenging computer vision problem as blurs arise not only from multiple object motions but also from camera shake, scene depth variation. To remove these complicated motion blurs, conventional energy optimization based methods rely on simple assumptions such that blur kernel is partially uniform or locally linear. Moreover, recent machine learning based methods also depend on synthetic blur datasets generated under these assumptions. This makes conventional deblurring methods fail to remove blurs where blur kernel is difficult to approximate or parameterize (e.g. object motion boundaries). In this work, we propose a multi-scale convolutional neural network that restores sharp images in an end-to-end manner where blur is caused by various sources. Together, we present multi-scale loss function that mimics conventional coarse-to-fine approaches. Furthermore, we propose a new large-scale dataset that provides pairs of realistic blurry image and the corresponding ground truth sharp image that are obtained by a high-speed camera. With the proposed model trained on this dataset, we demonstrate empirically that our method achieves the state-of-the-art performance in dynamic scene deblurring not only qualitatively, but also quantitatively.

We present a highly accurate single-image super-resolution (SR) method. Our method uses a very deep convolutional network inspired by VGG-net used for ImageNet classification \cite{simonyan2015very}. We find increasing our network depth shows a significant improvement in accuracy. Our final model uses 20 weight layers. By cascading small filters many times in a deep network structure, contextual information over large image regions is exploited in an efficient way. With very deep networks, however, convergence speed becomes a critical issue during training. We propose a simple yet effective training procedure. We learn residuals only and use extremely high learning rates ($10^4$ times higher than SRCNN \cite{dong2015image}) enabled by adjustable gradient clipping. Our proposed method performs better than existing methods in accuracy and visual improvements in our results are easily noticeable., CVPR 2016 Oral

We propose an image super-resolution method (SR) using a deeply-recursive convolutional network (DRCN). Our network has a very deep recursive layer (up to 16 recursions). Increasing recursion depth can improve performance without introducing new parameters for additional convolutions. Albeit advantages, learning a DRCN is very hard with a standard gradient descent method due to exploding/vanishing gradients. To ease the difficulty of training, we propose two extensions: recursive-supervision and skip-connection. Our method outperforms previous methods by a large margin., CVPR 2016 Oral

Several state-of-the-art video deblurring methods are based on a strong assumption that the captured scenes are static. These methods fail to deblur blurry videos in dynamic scenes. We propose a video deblurring method to deal with general blurs inherent in dynamic scenes, contrary to other methods. To handle locally varying and general blurs caused by various sources, such as camera shake, moving objects, and depth variation in a scene, we approximate pixel-wise kernel with bidirectional optical flows. Therefore, we propose a single energy model that simultaneously estimates optical flows and latent frames to solve our deblurring problem. We also provide a framework and efficient solvers to optimize the energy model. By minimizing the proposed energy function, we achieve significant improvements in removing blurs and estimating accurate optical flows in blurry frames. Extensive experimental results demonstrate the superiority of the proposed method in real and challenging videos that state-of-the-art methods fail in either deblurring or optical flow estimation., CVPR 2015 oral

Graph cuts-based algorithms have achieved great success in energy minimization for many computer vision applications. These algorithms provide approximated solutions for multi-label energy functions via move-making approach. This approach fuses the current solution with a proposal to generate a lower-energy solution. Thus, generating the appropriate proposals is necessary for the success of the move-making approach. However, not much research efforts has been done on the generation of "good" proposals, especially for non-metric energy functions. In this paper, we propose an application-independent and energy-based approach to generate "good" proposals. With these proposals, we present a graph cuts-based move-making algorithm called GA-fusion (fusion with graph approximation-based proposals). Extensive experiments support that our proposal generation is effective across different classes of energy functions. The proposed algorithm outperforms others both on real and synthetic problems., CVPR2015

Background In South Korea, a cluster of acute lung disease patients included lung injury disease suspected of being caused by the use of humidifier disinfectants. We examined the relationship between humidifier disinfectant exposure and clinically diagnosed humidifier disinfectant-associated lung injury (HDLI) in a family-based study. Methods This case-control study included 169 clinically confirmed HDLI cases and 303 family controls who lived with the HDLI patients. A range of information on exposure to humidifier disinfectants was obtained using a structured questionnaire and field investigations. Odds ratios (ORs) and confidence intervals (CIs) were estimated using unconditional logistic regression models that were adjusted for age, sex, presence of a factory within 1 km of residence, and the number of household chemical products used. Results HDLI risk increased approximately two-fold or more among the highest quartile compared with the lowest quartile in terms of the hours sleeping in a room with an operating humidifier treated with disinfectant (adjusted OR = 2.0, 95 % CI = 1.1-3.7), average hours of disinfectant-treated humidifier use per day (adjusted OR = 2.1, 95 % CI = 1.0-4.5), airborne disinfectant intensity (adjusted OR = 2.6, 95% CI = 1.2-5.3), and cumulative disinfectant inhalation level (adjusted OR = 2.0, 95% CI = 1.0-4.1). HDLI risk increased as the distance of the bed from humidifier gets shorter; compared with longer distance (> 1 m), the odds ratio was 2.7 for 0.5 to 1 m (95 % CI = 1.5-5.1) and 13.2 for

A novel tracking algorithm that can track a highly non-rigid target robustly is proposed using a new bounding box representation called the Double Bounding Box (DBB). In the DBB, a target is described by the combination of the Inner Bounding Box (IBB) and the Outer Bounding Box (OBB). Then our objective of visual tracking is changed to find the IBB and OBB instead of a single bounding box, where the IBB and OBB can be easily obtained by the Dempster-Shafer (DS) theory. If the target is highly non-rigid, any single bounding box cannot include all foreground regions while excluding all background regions. Using the DBB, our method does not directly handle the ambiguous regions, which include both the foreground and background regions. Hence, it can solve the inherent ambiguity of the single bounding box representation and thus can track highly non-rigid targets robustly. Our method finally finds the best state of the target using a new Constrained Markov Chain Monte Carlo (CMCMC)-based sampling method with the constraint that the OBB should include the IBB. Experimental results show that our method tracks non-rigid targets accurately and robustly, and outperforms state-of-the-art methods.

Some immunological biomarkers have been reported to be associated with job-related stress. This study was conducted to explore the intercorrelation between the psychosocial components of job stress and various immunological biomarkers among female nurses. To assess monthly and weekly job stress, 41 nurses have repeatedly completed questionnaires such as the National Institute for Occupational Safety and Health General Job Stress Questionnaire, the profile of mood states short version and the Center for Epidemiologic Studies-Depression scale. Using flow cytometry and radioimmunoassay, the number of white blood cells, lymphocytic proliferation to mitogens, and toxoid were measured. Moreover, levels of hydrocortisol, interleukin-β, interferon-γ, and tumor necrosis factor-α and salivary immunoglobulin A were evaluated by enzyme-linked immunosorbent assay. When the Pearson correlation coefficients between job stress and immunological biomarkers were estimated after adjusting for age and smoking status, “Clashes: conflict at work” was significantly related to the number of CD4 cells (r = 0.36, p-value

Genome-wide association studies (GWAS) have mapped risk alleles for at least 10 distinct cancers to a small region of 63 000 bp on chromosome 5p15.33. This region harbors the TERT and CLPTM1L genes; the former encodes the catalytic subunit of telomerase reverse transcriptase and the latter may play a role in apoptosis. To investigate further the genetic architecture of common susceptibility alleles in this region, we conducted an agnostic subset-based meta-analysis (association analysis based on subsets) across six distinct cancers in 34 248 cases and 45 036 controls. Based on sequential conditional analysis, we identified as many as six independent risk loci marked by common single-nucleotide polymorphisms: five in the TERT gene (Region 1: rs7726159, P = 2.10 x 10(-39); Region 3: rs2853677, P = 3.30 x 10(-36) and P-Conditional = 2.36 x 10(-8); Region 4: rs2736098, P = 3.87 x 10(-12) and P-Conditional = 5.19 x 10(-6), Region 5: rs13172201, P = 0.041 and P-Conditional = 2.04 x 10(-6); and Region 6: rs10069690, P = 7.49 x 10 215 and P-Conditional = 5.35 x 10(-7)) and one in the neighboring CLPTM1L gene(Region 2: rs451360; P = 1.90 x 10(-18) and P-Conditional = 7.06 x 10(-16)). Between three and five cancers mapped to each independent locus with both risk-enhancing and protective effects. Allele-specific effects on DNA methylation were seen for a subset of risk loci, indicating that methylation and subsequent effects on gene expression may contribute to the biology of risk variants on 5p15.33. Our results provide strong support for extensive pleiotropy across this region of 5p15.33, to an extent not previously observed in other cancer susceptibility loci.

BACKGROUND: In Xuanwei County, Yunnan Province, China, lung cancer mortality rates in both males and females are among the highest in China. METHODS: We evaluated differential effects of smoking on lung cancer mortality before and after household stove improvement with chimney to reduce exposure to smoky coal emissions in the unique cohort in Xuanwei, China. Effects of independent variables on lung cancer mortality were measured as hazard ratios and 95% confidence intervals using a multivariable Cox regression model that included separate time-dependent variables for smoking duration (years) before and after stove improvement. RESULTS AND CONCLUSION: We found that the effect of smoking on lung cancer risk becomes considerably stronger after chimney installation and consequent reduction of indoor coal smoke exposure.

International audience; Graph matching is a powerful tool for computer vision and machine learning. In this paper, a novel approach to graph matching is developed based on the sequential Monte Carlo framework. By constructing a sequence of intermediate target distributions, the proposed algorithm sequentially performs a sampling and importance resampling to maximize the graph matching objective. Through the sequential sampling procedure, the algorithm effectively collects potential matches under one-to-one matching constraints to avoid the adverse effect of outliers and deformation. Experimental evaluations on synthetic graphs and real images demonstrate its higher robustness to deformation and outliers.

Benzene, a recognized hematotoxicant and carcinogen, can damage the human immune system. We studied the association between single nucleotide polymorphisms (SNPs) in genes involved in innate immunity and benzene hematotoxicity in a cross-sectional study of workers exposed to benzene (250 workers and 140 controls). A total of 1,236 tag SNPs in 149 gene regions of six pathways were included in the analysis. Six gene regions were significant for their association with white blood cell (WBC) counts (MBP, VCAM1, ALOX5, MPO, RAC2, and CRP) based on gene-region (P < 0.05) and SNP analyses (FDR < 0.05). VCAM1 rs3176867, ALOX5 rs7099684, and MPO rs2071409 were the three most significant SNPs. They showed similar effects on WBC subtypes, especially granulocytes, lymphocytes, and monocytes. A 3-SNP block in ALOXE3 (rs7215658, rs9892383, and rs3027208) showed a global association (omnibus P = 0.0008) with WBCs even though the three SNPs were not significant individually. Our study suggests that polymorphisms in innate immunity genes may play a role in benzene-induced hematotoxicity; however, independent replication is necessary.

Background Although the role of microRNA’s (miRNA’s) biogenesis pathway genes in cancer development and progression has been well established, the association between genetic variants of this pathway genes and breast cancer survival is still unknown. Methods We used genotype data available from a previously conducted case–control study to investigate association between common genetic variations in miRNA biogenesis pathway genes and breast cancer survival. We investigated the possible associations between 41 germ-line single-nucleotide polymorphisms (SNPs) and both disease free survival (DFS) and overall survival (OS) among 488 breast cancer patients. During the median follow-up of 6.24 years, 90 cases developed disease progression and 48 cases died. Results Seven SNPs were significantly associated with breast cancer survival. Two SNPs in AGO2 (rs11786030 and rs2292779) and DICER1 rs1057035 were associated with both DFS and OS. Two SNPs in HIWI (rs4759659 and rs11060845) and DGCR8 rs9606250 were associated with DFS, while DROSHA rs874332 and GEMIN4 rs4968104 were associated with only OS. The most significant association was observed in variant allele of AGO2 rs11786030 with 2.62-fold increased risk of disease progression (95% confidence interval (CI), 1.41-4.88) and in minor allele homozygote of AGO2 rs2292779 with 2.94-fold increased risk of death (95% CI, 1.52-5.69). We also found cumulative effects of SNPs on DFS and OS. Compared to the subjects carrying 0 to 2 high-risk genotypes, those carrying 3 or 4–6 high-risk genotypes had an increased risk of disease progression with a hazard ratio of 2.16 (95% CI, 1.18- 3.93) and 4.47 (95% CI, 2.45- 8.14), respectively (P for trend, 6.11E-07). Conclusions Our results suggest that genetic variants in miRNA biogenesis pathway genes may be associated with breast cancer survival. Further studies in larger sample size and functional characterizations are warranted to validate these results.

Introduction Although approximately 25 common genetic susceptibility loci have been identified to be independently associated with breast cancer risk through genome-wide association studies (GWAS), the genetic risk variants reported to date only explain a small fraction of the heritability of breast cancer. Furthermore, GWAS-identified loci were primarily identified in women of European descent. Methods To evaluate previously identified loci in Korean women and to identify additional novel breast cancer susceptibility variants, we conducted a three-stage GWAS that included 6,322 cases and 5,897 controls. Results In the validation study using Stage I of the 2,273 cases and 2,052 controls, seven GWAS-identified loci [5q11.2/MAP3K1 (rs889312 and rs16886165), 5p15.2/ROPN1L (rs1092913), 5q12/MRPS30 (rs7716600), 6q25.1/ESR1 (rs2046210 and rs3734802), 8q24.21 (rs1562430), 10q26.13/FGFR2 (rs10736303), and 16q12.1/TOX3 (rs4784227 and rs3803662)] were significantly associated with breast cancer risk in Korean women (Ptrend < 0.05). To identify additional genetic risk variants, we selected the most promising 17 SNPs in Stage I and replicated these SNPs in 2,052 cases and 2,169 controls (Stage II). Four SNPs were further evaluated in 1,997 cases and 1,676 controls (Stage III). SNP rs13393577 at chromosome 2q34, located in the Epidermal Growth Factor Receptor 4 (ERBB4) gene, showed a consistent association with breast cancer risk with combined odds ratios (95% CI) of 1.53 (1.37-1.70) (combined P for trend = 8.8 × 10-14). Conclusions This study shows that seven breast cancer susceptibility loci, which were previously identified in European and/or Chinese populations, could be directly replicated in Korean women. Furthermore, this study provides strong evidence implicating rs13393577 at 2q34 as a new risk variant for breast cancer.