Search

Blind image deconvolution is a fundamental task in image processing, computational imaging, and computer vision. It has earned intensive attention in the past decade since the seminal work of Fergus et al. [1] for camera shake removal. In spite of the recent great progress in this field, this paper aims to formulate the blind problem with a simpler modeling perspective. What is more important, the newly proposed approach is expected to achieve comparable or even better performance towards the real blurred images. Specifically, the core critical idea is the proposal of a pure gradient-based discriminative prior for accurate and robust blur kernel estimation. Numerous experimental results on both the benchmark datasets and real-world blurred images in various imaging scenarios, e.g., natural, manmade, low-illumination, text, or people, demonstrate well the effectiveness and robustness of the proposed approach.

Single-image nonparametric blind super-resolution is a fundamental image restoration problem yet largely ignored in the past decades among the computational photography and computer vision communities. An interesting phenomenon is observed that learning-based single-image super-resolution (SR) has been experiencing a rapid development since the boom of the sparse representation in 2005s and especially the representation learning in 2010s, wherein the high-res image is generally blurred by a supposed bicubic or Gaussian blur kernel. However, the parametric assumption on the form of blur kernels does not hold in most practical applications because in real low-res imaging a high-res image can undergo complex blur processes, e.g., Gaussian-shaped kernels of varying sizes, ellipse-shaped kernels of varying orientations, curvilinear kernels of varying trajectories. The paper is mainly motivated by one of our previous works: Shao and Elad (in: Zhang (ed) ICIG 2015, Part III, Lecture notes in computer science, Springer, Cham, 2015). Specifically, we take one step further in this paper and present a type of adaptive heavy-tailed image priors, which result in a new regularized formulation for nonparametric blind super-resolution. The new image priors can be expressed and understood as a generalized integration of the normalized sparsity measure and relative total variation. Although it seems that the proposed priors are simple, the core merit of the priors is their practical capability for the challenging task of nonparametric blur kernel estimation for both super-resolution and deblurring. Harnessing the priors, a higher-quality intermediate high-res image becomes possible and therefore more accurate blur kernel estimation can be accomplished. A great many experiments are performed on both synthetic and real-world blurred low-res images, demonstrating the comparative or even superior performance of the proposed algorithm convincingly. Meanwhile, the proposed priors are demonstrated quite applicable to blind image deblurring which is a degenerated problem of nonparametric blind SR.

We consider the task of grasping a target object based on a natural language command query. Previous work primarily focused on localizing the object given the query, which requires a separate grasp detection module to grasp it. The cascaded application of two pipelines incurs errors in overlapping multi-object cases due to ambiguity in the individual outputs. This work proposes a model named Command Grasping Network(CGNet) to directly output command satisficing grasps from RGB image and textual command inputs. A dataset with ground truth (image, command, grasps) tuple is generated based on the VMRD dataset to train the proposed network. Experimental results on the generated test set show that CGNet outperforms a cascaded object-retrieval and grasp detection baseline by a large margin. Three physical experiments demonstrate the functionality and performance of CGNet., 7 pages, 2 figures, Accepted to the ICRA2021

In the past decades, the world has experienced several major virus outbreaks, e.g. West African Ebola outbreak, Zika virus in South America and most recently global coronavirus (COVID-19) pandemic. Many vaccines have been developed to prevent a variety of infectious diseases successfully. However, several infections have not been preventable so far, like COVID-19, which induces an immediate urgent need for effective vaccines. These emerging infectious diseases often pose unprecedent challenges for the global heath community as well as the conventional vaccine development paradigm. With a long and costly traditional vaccine development process, there are extensive needs in innovative vaccine trial designs and analyses, which aim to design more efficient vaccines trials. Featured with reduced development timeline, less resource consuming or improved estimate for the endpoints of interests, these more efficient trials bring effective medicine to target population in a faster and less costly way. In this paper, we will review a few vaccine trials equipped with adaptive design features, Bayesian designs that accommodate historical data borrowing, the master protocol strategy emerging during COVID-19 vaccine development, Real-World-Data (RWD) embedded trials and the correlate of protection framework and relevant research works. We will also discuss some statistical methodologies that improve the vaccine efficacy, safety and immunogenicity analyses. Innovative clinical trial designs and analyses, together with advanced research technologies and deeper understanding of the human immune system, are paving the way for the efficient development of new vaccines in the future.

Energy Router(ER) is the core of metro energy Internet construction. The paper proposes an ER based on dual DC bus. First, with the high voltage 1500V DC bus and the low voltage 600V DC bus as the core, 35kV AC grid, catenary, station load, Photovoltaic (PV) and energy storage are interconnected through multiple ports. Second, to analyze the working mode of the ER based on the traction characteristics of the metro. Then combined with the control method of each port, a coordinated control strategy is proposed to realize the flexible and autonomous distribution of electric energy. Finally, the metro power supply network based on the ER is built in Simulink. The results show that the combination of ER and metro power supply network solves the problems of serious regenerative braking energy loss, uncontrollability of electric energy, and low new energy consumption capacity.

This paper proposes a system configuration of the half-bridge submodule based modular multilevel converter (HBSM-MMC) combined with SCOTT transformer, which can be applied in the 50Hz three-phase to 50Hz single-phase conversion for railway supply. Three AC terminals of MMC are respectively connected to the public- and railway- grids. The railway and SCOTT transformer are grounded together to provide circuit path for the catenary current. A control strategy based on fundamental-frequency zero-sequence voltage injection is proposed, to implement separate control of public-and railway- grids. Simulation results verify the validation of the proposed system and control strategy. The proposed system can eliminate the imbalance of public grid, and MMC arm output voltages are fully and evenly used. It is economical advantageous due to the reduction in converter capacity, capacitor requirement, and converter loss. In addition, the bulky transformers and filters can be avoided.

Continuous co-phase traction power supply system is an effective method to eliminate neutral sections and provide high power quality for both the public grid and the catenary. Modular multilevel converter (MMC) can be applied in co-phase power supply system to perform three-phase/single-phase conversion. However, the existing schemes of MMC have some problems including high cost, bulky size and coupling between public grid and catenary. This paper proposed a new main circuit scheme based on four-port MMC (4P-MMC) and three-phase Y/D transformer, which significantly reduces the cost and realizes the decoupling of three-phase and single-phase sides. Simulation results are presented to verify the validation of the proposed scheme.

With the advance of urban railway transportation and wireless communication technology, an increasing concern is placed on the reliability and low latency of urban mass transit communication system. Nowadays, both of WLAN and LTEM systems, known as two mainstream subway communication systems, are deployed redundantly with single communication network in case of unexpected failure. However, using a single communication system has irresistibility to frequency interference and traditional redundant cold backup manual has an intolerable delay in switching. In this paper, we adopt a multipath transmission network architecture with heterogeneous communication systems to solve the problems mentioned above. To realize the hot backup networks, the Redundant Forward Gateway (RFG) device is designed to connect terminals to the two networks. Information forwarded by source RFGs can be transmitted redundantly and simultaneously via both WLAN and LTE-M communication systems. At the other side, the redundant data packets will be removed before being forwarded to the destination terminals. Besides, tests were performed in various communication scenarios to evaluate the availability and effectiveness of the RFG and the network architecture. The results indicate that the RFG has a negligible processing delay and improves the robustness and performance compared to the traditional LTE-M network according to the transmission delay and packet loss tests. In conclusion, it is considered that the RFG and the heterogeneous LTE-M and WLAN networks are of vital importance in enhancing reliability in the metro communication system.

Cellular vehicle-to-everything (C-V2X) communication, as a part of 5G wireless communication, has been considered one of the most significant techniques for Smart City. Vehicles platooning is an application of Smart City that improves traffic capacity and safety by C-V2X. However, different from vehicles platooning travelling on highways, C-V2X could be more easily eavesdropped and the spectrum resource could be limited when they converge at an intersection. Satisfying the secrecy rate of C-V2X, how to increase the spectrum efficiency (SE) and energy efficiency (EE) in the platooning network is a big challenge. In this paper, to solve this problem, we propose a Security-Aware Approach to Enhancing SE and EE Based on Deep Reinforcement Learning, named SEED. The SEED formulates an objective optimization function considering both SE and EE, and the secrecy rate of C-V2X is treated as a critical constraint of this function. The optimization problem is transformed into the spectrum and transmission power selections of V2V and V2I links using deep Q network (DQN). The heuristic result of SE and EE is obtained by the DQN policy based on rewards. Finally, we simulate the traffic and communication environments using Python. The evaluation results demonstrate that the SEED outperforms the DQN-wopa algorithm and the baseline algorithm by 31.83 % and 68.40 % in efficiency. Source code for the SEED is available at https://github.com/BandaidZ/OptimizationofSEandEEBasedonDRL.

Background: Functional Magnetic Resonance Imaging (fMRI) is a presurgical planning technique used to localize functional cortex so as to maximize resection of diseased tissue and avoid viable tissue. In this retrospective study, we examined differences in morbidity and mortality of brain tumor patients who received preoperative fMRI in comparison to those who did not. Methods: Brain tumor patients (n = 206) were selected from a retrospective review of neurosurgical case logs from 2001 to 2009 at the University of Wisconsin-Madison. Results: Univariate analysis showed improved mortality in the fMRI group and the fMRI + Electrical Cortical Stimulation Mapping (ECM) group compared to the No-fMRI group. Multivariate analyses showed improved mortality of the fMRI group and the fMRI + ECM group compared to the No-fMRI group, with age and tumor grade being the most significant influencers. Overall, the fMRI group showed survival benefits at 3 years; twice that of the No-fMRI group. Furthermore, patients with high-grade tumors showed significant survival benefits in the fMRI group, while patients with low-grade tumors did not (controlling for age and ECM). There was also a significant difference in the two groups with respect to morbidity, with patients receiving fMRI showing improved outcomes in the motor and language domains. Conclusions: This study analyzing a large retrospective series of brain tumor patients with and without the use of fMRI in the preoperative planning has resulted in improved mortality and morbidity outcomes with the use of fMRI. These results point to the importance of incorporating fMRI in presurgical planning in the clinical management of patients with brain tumors. Keywords: Brain mapping, Brain tumors, Functional neuroimaging

A segmentation-based architecture is proposed to decompose objects into multiple primitive shapes from monocular depth input for robotic manipulation. The backbone deep network is trained on synthetic data with 6 classes of primitive shapes generated by a simulation engine. Each primitive shape is designed with parametrized grasp families, permitting the pipeline to identify multiple grasp candidates per shape primitive region. The grasps are priority ordered via proposed ranking algorithm, with the first feasible one chosen for execution. On task-free grasping of individual objects, the method achieves a 94% success rate. On task-oriented grasping, it achieves a 76% success rate. Overall, the method supports the hypothesis that shape primitives can support task-free and task-relevant grasp prediction., 8 pages

Responder analysis is in common use in clinical trials, and has been described and endorsed in regulatory guidance documents, especially in trials where “soft” clinical endpoints such as rating scales are used. The procedure is useful, because responder rates can be understood more intuitively than a difference in means of rating scales. However, two major issues arise: 1) such dichotomized outcomes are inefficient in terms of using the information available and can seriously reduce the power of the study; and 2) the results of clinical trials depend considerably on the response cutoff chosen, yet in many disease areas there is no consensus as to what is the most appropriate cutoff. This article addresses these two issues, offering a novel approach for responder analysis that could both improve the power of responder analysis and explore different responder cutoffs if an agreed-upon common cutoff is not present. Specifically, we propose a statistically rigorous clinical trial design that pre-specifies multiple tests of responder rates between treatment groups based on a range of pre-specified responder cutoffs, and uses the minimum of the p-values for formal inference. The critical value for hypothesis testing comes from permutation distributions. Simulation studies are carried out to examine the finite sample performance of the proposed method. We demonstrate that the new method substantially improves the power of responder analysis, and in certain cases, yields power that is approaching the analysis using the original continuous (or ordinal) measure.

The rapidly changing drug development landscapes have brought unique challenges to sponsors in designing clinical trials in a faster and more efficient way. With the ability to accelerate development timeline, reduce redundant sample size, and select the right dose and patient population during the clinical trial, adaptive designs help to increase the probability of success of clinical trials and eventually contribute to bringing the promising drugs to patients earlier and fulfilling their unmet medical needs. Although extensive adaptive design methods have been proposed in recent years, a comprehensive review of how to implement adaptive design in the practical confirmatory trials is still lacking. In this paper, we will review the evolving history of adaptive designs, updates of newly released regulatory guidance and emerging practical adaptive designs, including but not limited to sample size re-estimation, seamless design and surrogate endpoint used in the interim analysis. Furthermore, we will discuss the current practice of adaptive design implementation by demonstrating a complex oncology seamless phase 2/3 adaptive design case study. Through this example, we will introduce the critical roles of each cross disciplinary function, communication process and important documents when adaptive designs are implemented in real-world setting.

Blind image deblurring or deconvolution is a very hot topic towards which numerous methods have been put forward in the past decade, demonstrating successful deblurring results on one or another benchmark natural image dataset. However, most of existing algorithms are found not robust enough as dealing with images in specific scenarios, such as images with text, saturated area or face. In this paper, a robust blind deblurring approach is presented using relative total variation as a regularization penalty. To the best of our knowledge, none of previous studies have followed this modeling principle. The underlying idea is to pursue more accurate intermediate image for more reliable kernel estimation by harnessing relative total variation which could extract salient structures from textures and suppress ringing artifacts and noise as well. An iterative estimating algorithm is finally deduced for alternatively update of sharp image and blur kernel by operator splitting and augmented Lagrangian. Extensive experiments on a challenging synthetic dataset and real-world images as well are conducted to validate the new method with comparison against the state-of-the-art approaches from 2006 to 2014. Based on both PSNR and SSIM and a non-reference evaluation metric, the results demonstrate that our approach is comparatively more effective to process blurred images in more challenging scenarios.

This paper aims to exploit the full potential of gradient-based methods, attempting to explore a simple, robust yet discriminative image prior for blind deblurring. The specific contributions are three-fold: Above all, a pure gradient-based heavy-tailed model is proposed as a generalized integration of the normalized sparsity and the relative total variation. On the second, a plug-and-play algorithm is deduced to alternatively estimate the intermediate sharp image and the nonparametric blur kernel. With the numerical scheme, image estimation is simplified to an image smoothing problem. Lastly, a great many experiments are performed accompanied with comparisons with state-of-the-art approaches on synthetic benchmark datasets and real blurry images in various scenarios. The experimental results show well the effectiveness and robustness of the proposed method.

Randomization is fundamental to the design and conduct of clinical trials. Simple randomization ensures independence among subject treatment assignments and prevents potential selection biases, yet it does not guarantee balance in covariate distributions across treatment groups. Ensuring balance in important prognostic covariates across treatment groups is desirable for many reasons. A broad class of randomization methods for achieving balance are reviewed in this paper; these include block randomization, stratified randomization, minimization, and dynamic hierarchical randomization. Practical considerations arising from experience with using the techniques are described. A review of randomization methods used in practice in recent randomized clinical trials is also provided.

Colorectal cancer is the second leading cause of death from cancer in the United States. To facilitate the efficiency of colorectal cancer screening, there is a need to stratify risk for colorectal cancer among the 90% of US residents who are considered “average risk.” In this article, we investigate such risk stratification rules for advanced colorectal neoplasia (colorectal cancer and advanced, precancerous polyps). We use a recently completed large cohort study of subjects who underwent a first screening colonoscopy. Logistic regression models have been used in the literature to estimate the risk of advanced colorectal neoplasia based on quantifiable risk factors. However, logistic regression may be prone to overfitting and instability in variable selection. Since most of the risk factors in our study have several categories, it was tempting to collapse these categories into fewer risk groups. We propose a penalized logistic regression method that automatically and simultaneously selects variables, groups categories, and estimates their coefficients by penalizing the [Formula: see text]-norm of both the coefficients and their differences. Hence, it encourages sparsity in the categories, i.e. grouping of the categories, and sparsity in the variables, i.e. variable selection. We apply the penalized logistic regression method to our data. The important variables are selected, with close categories simultaneously grouped, by penalized regression models with and without the interactions terms. The models are validated with 10-fold cross-validation. The receiver operating characteristic curves of the penalized regression models dominate the receiver operating characteristic curve of naive logistic regressions, indicating a superior discriminative performance.

Although, the relationship of spontaneous subarachnoid hemorrhage (SAH) to climatic or circadian factors has been widely studied, epidemiologic, circardian and climatic factors in non-aneurysmal SAH (naSAH), particularly perimesencephalic SAH (PMH), has not been reported before.For the first time, demographic, climatic, and circadian variables are examined together as possible contributing factors comparing aSAH and naSAH.We reviewed records for 384 patients admitted to University of Wisconsin Neurosurgery Service from January 2005 to December 2010 with spontaneous non-traumatic SAH. Patients were grouped as aSAH (n=338) or naSAH (n=46) on clinical and radiological criteria. PMH (n=32) was identified as a subgroup of naSAH based on radiological criteria. We logged demographic data, time of SAH, temperature at onset and atmospheric pressure at onset. The three subgroups were compared.Aneurysmal SAH occurred most often from 6am to 12pm (p0.001); this correlation was not found in naSAH or PMH subgroups. Demographic analysis demonstrated predominance of female gender (p=0.008) and smoking (p=0.002) in aSAH, with predominance of hypercholesterolemia in naSAH (p=0.033). Atmospheric pressure, correlated with aSAH in the main county referral area, where we had detailed weather data (p0.05); however, there was no weather correlation in the entire referral region taken together. Multivariate analysis supported a statistical difference only in smoking status between aSAH and naSAH groups (p=0.0159).Statistical differences in gender, smoking status, and history of hypercholesterolemia support a clinical distinction between aSAH and naSAH. Furthermore, circadian patterning of aSAH is not reproduced in naSAH, supporting pathophysiologic differences. Only smoking status provides a robust difference in aSAH and naSAH groups. Our data prompt further investigation into the relationship between aSAH and atmospheric pressure.

BACKGROUND ABT-126 is a potent, selective α7 nicotinic acetylcholine receptor agonist with putative procognitive effects as a monotherapy in treating Alzheimer's disease (AD). OBJECTIVE This randomized, double-blind, placebo-controlled multicenter study (NCT01549834) investigated the efficacy and safety of ABT-126 in subjects with mild-to-moderate AD who were taking stable doses of acetylcholinesterase inhibitors (AChEIs). METHODS Subjects received 25 mg ABT-126 (n = 143), 75 mg ABT-126 (n = 145), or placebo (n = 146) once daily for 24 weeks. Subjects who completed the 24-week double-blind study were eligible to enroll in a 28-week open-label extension study (NCT01690195) and received 75 mg ABT-126 daily. The primary efficacy endpoint was the change from baseline to week 24 in the 11-item total score of the Alzheimer's Disease Assessment Scale- Cognitive Subscale (ADAS-Cog). RESULTS Neither dose of ABT-126 demonstrated significant improvement compared with placebo in the primary efficacy endpoint. However, 25 mg ABT-126 demonstrated significant improvement compared with placebo in ADAS-Cog scores at week 4 (least squares mean difference, -1.21; standard error, 0.51; p < 0.010, one-sided); 75 mg ABT-126 did not demonstrate significant improvements in ADAS-Cog scores compared with placebo at any time point. A treatment effect was not observed for any secondary efficacy measures of cognition, function, or global improvement. ABT-126 was generally well tolerated; the most common adverse events were agitation, constipation, diarrhea, fall, and headache. CONCLUSIONS Overall, the efficacy profile of ABT-126 did not warrant further development as add-on therapy to AChEIs to treat mild-to-moderate AD.

Multiplicity control is an important statistical issue in clinical trials where strong control of the type I error rate is required. Many multiple testing methods have been proposed and applied to address multiplicity issues in clinical trials. This paper provides an application oriented and comprehensive overview of commonly used multiple testing procedures and recent developments in statistical methodology in multiple testing in clinical trials. Commonly used multiple testing procedures are applied to test non-hierarchical hypotheses and gatekeeping procedures can be used to test hierarchically ordered hypotheses while controlling the overall type I error rate. The recently developed graphical approach has the flexibility to integrate hierarchical and non-hierarchical procedures into one framework. A graphical multiple testing procedure with "no-dead-end" provides an opportunity to fully recycle α across hypothesis families. Two hypothetical clinical trial examples are used to illustrate applications of these procedures. The advantages and disadvantages of the different procedures are briefly discussed.

Background: The role of postoperative radiation (RT) in atypical meningioma remains controversial. Materials and Methods: We report a retrospective review of outcomes and prognostic factor analysis in 158 patients treated between 2000 and 2010, and extensively review the literature. Results: Following resection, 23 patients received immediate RT, whereas 135 did not. Median progression-free survival (PFS) with and without RT was 59 (range 43-86) and 88 (range 64-123) months. For Simpson grade (G) 1-3 resection, with and without RT, median PFS was 48 (2-80) versus 96 (88-123) months and for Simpson G4, it was 59 (6-86) versus 47 (15-104) months (P = 0.4). The rate of 5-year overall survival (OS) with and without RT was 89% and 83%, respectively. On univariate analysis, Simpson G4 (HR 3.2, P = 0.0006) and brain invasion (HR 2.2, P = 0.03) were significantly associated with progression, whereas age >60 years (HR 9.7, P = 0.002), mitoses >5 per 10 high-power field (0.2, P = 0.0056), and Simpson G4 (HR 2.4, P = 0.07) were associated with higher risk of death. We summarized 22 additional reports, which provide very divergent results regarding the benefit of RT. Conclusions: In our series, adjuvant RT is surprisingly associated with worse PFS and OS, and this is more likely to be due to selection bias of referring tumors with more aggressive characteristics such as elevated Ki-67 and brain invasion for adjuvant RT, rather than a direct causal effect of adjuvant RT. Although there is a trend toward improved PFS with adjuvant RT after subtotal resection, no improvement was noted in OS. Multivariate analysis did not yield statistical significance for any of the factors including Simpson grades of resection, adjuvant RT, or six pathological defining features. The relatively divergent results in the literature are most likely explained by patient selection variability; therefore, randomized trials to adequately address this question are clearly necessary.

The tumor–node–metastasis (TNM) staging system has been the anchor of cancer diagnosis, treatment, and prognosis for many years. For meaningful clinical use, an orderly, progressive condensation of the T and N categories into an overall staging system needs to be defined, usually with respect to a time-to-event outcome. This can be considered as a cutpoint selection problem for a censored response partitioned with respect to two ordered categorical covariates and their interaction. The aim is to select the best grouping of the TN categories. A novel bootstrap cutpoint/model selection method is proposed for this task by maximizing bootstrap estimates of the chosen statistical criteria. The criteria are based on prognostic ability including a landmark measure of the explained variation, the area under the receiver operating characteristic (ROC) curve, and a concordance probability generalized from Harrell’s c-index. We illustrate the utility of our method by applying it to the staging of colorectal cancer.

The tumor-node-metastasis staging system has been the lynchpin of cancer diagnosis, treatment, and prognosis for many years. For meaningful clinical use, an orderly grouping of the T and N categories into a staging system needs to be defined, usually with respect to a time-to-event outcome. This can be reframed as a model selection problem with respect to features arranged on a partially ordered two-way grid, and a penalized regression method is proposed for selecting the optimal grouping. Instead of penalizing the L1-norm of the coefficients like lasso, in order to enforce the stage grouping, we place L1 constraints on the differences between neighboring coefficients. The underlying mechanism is the sparsity-enforcing property of the L1 penalty, which forces some estimated coefficients to be the same and hence leads to stage grouping. Partial ordering constraints is also required as both the T and N categories are ordinal. A series of optimal groupings with different numbers of stages can be obtained by varying the tuning parameter, which gives a tree-like structure offering a visual aid on how the groupings are progressively made. We hence call the proposed method the lasso tree. We illustrate the utility of our method by applying it to the staging of colorectal cancer using survival outcomes. Simulation studies are carried out to examine the finite sample performance of the selection procedure. We demonstrate that the lasso tree is able to give the right grouping with moderate sample size, is stable with regard to changes in the data, and is not affected by random censoring.

To determine whether the mammographic density of noncalcified solid breast masses is associated with malignancy and to measure the agreement between prospective and retrospective assessment.The institutional review board approved this study and waived informed consent. Three hundred forty-eight consecutive breast masses in 328 women who underwent image-guided or surgical biopsy between October 2005 and December 2007 were included. All 348 biopsy-proved masses were randomized and assigned to a radiologist who was blinded to biopsy results for retrospective assessment by using the Breast Imaging Reporting and Data System (retrospectively assessed data set). Clinical radiologists prospectively assessed the density of 180 of these masses (prospectively assessed data set). Pathologic result at biopsy was the reference standard. Benign masses were followed for at least 1 year by linking each patient to a cancer registry. Univariate analyses were performed on the retrospectively assessed data set. The association of mass density and malignancy was examined by creating a logistic model for the prospectively assessed data set. Agreement between prospective and retrospective assessments was calculated by using the κ statistic.In the retrospectively assessed data set, 70.2% of high-density masses were malignant, and 22.3% of the isodense or low-density masses were malignant (P.0001). In the prospective logistic model, high density (odds ratio, 6.6), irregular shape (odds ratio, 9.9), spiculated margin (odds ratio, 20.3), and age (β = 0.09, P.0001) were significantly associated with the probability of malignancy. The κ value for prospective-retrospective agreement of mass density was 0.53.High mass density is significantly associated with malignancy in both retrospectively and prospectively assessed data sets, with moderate prospective-retrospective agreement. Radiologists should consider mass density as a valuable descriptor that can stratify risk.http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.10100328/-/DC1.

Author(s): Burnside, Elizabeth S; Lin, Yunzhi; Munoz del Rio, Alejandro; Pickhardt, Perry J; Wu, Yirong; Strigel, Roberta M; Elezaby, Mai A; Kerr, Eve A; Miglioretti, Diana L | Abstract: BackgroundMotivated by the challenges in assessing physician-level cancer screening performance and the negative impact of misclassification, we propose a method (using mammography as an example) that enables confident assertion of adequate or inadequate performance or alternatively recognizes when more data is required.MethodsUsing established metrics for mammography screening performance-cancer detection rate (CDR) and recall rate (RR)-and observed benchmarks from the Breast Cancer Surveillance Consortium (BCSC), we calculate the minimum volume required to be 95% confident that a physician is performing at or above benchmark thresholds. We graphically display the minimum observed CDR and RR values required to confidently assert adequate performance over a range of interpretive volumes. We use a prospectively collected database of consecutive mammograms from a clinical screening program outside the BCSC to illustrate how this method classifies individual physician performance as volume accrues.ResultsOur analysis reveals that an annual interpretive volume of 2770 screening mammograms, above the United States' (US) mandatory (480) and average (1777) annual volumes but below England's mandatory (5000) annual volume is necessary to confidently assert that a physician performed adequately. In our analyzed US practice, a single year of data uniformly allowed confident assertion of adequate performance in terms of RR but not CDR, which required aggregation of data across more than one year.ConclusionFor individual physician quality assessment in cancer screening programs that target low incidence populations, considering imprecision in observed performance metrics due to small numbers of patients with cancer is important.