Current top-performing object detectors depend on deep CNN backbones, such as ResNet-101 and Inception, benefiting from their powerful feature representations but suffering from high computational costs. Conversely, some lightweight model based detectors fulfil real time processing, while their accuracies are often criticized. In this paper, we explore an alternative to build a fast and accurate detector by strengthening lightweight features using a hand-crafted mechanism. Inspired by the structure of Receptive Fields (RFs) in human visual systems, we propose a novel RF Block (RFB) module, which takes the relationship between the size and eccentricity of RFs into account, to enhance the feature discriminability and robustness. We further assemble RFB to the top of SSD, constructing the RFB Net detector. To evaluate its effectiveness, experiments are conducted on two major benchmarks and the results show that RFB Net is able to reach the performance of advanced very deep detectors while keeping the real-time speed. Code is available at https://github.com/ruinmessi/RFBNet.

Objectives This study sought to determine the relationship of cardiovascular magnetic resonance (CMR) measures of tissue composition to age in the Multi-Ethnic Study of Atherosclerosis (MESA). Background Animal and human studies have demonstrated increased collagen deposition in senescent hearts. New CMR indices of tissue composition by using T1 mapping are sensitive to the presence of myocardial fibrosis. Methods A total of 1,231 study participants (51% women; age range 54 to 93 years) of the MESA cohort were evaluated with T1 mapping by using 1.5-T CMR scanners. None of the participants had focal scar on delayed enhancement CMR. Single-slice T1 mapping was performed at the midventricular level before and at 12- and 25-min delay after administration of gadolinium contrast by using a modified Look-Locker inversion recovery sequence. The partition coefficient was determined by the slope of the linear relationship of (1/T1myo vs. 1/T1blood). The extracellular volume fraction (ECV) was derived accounting for the hematocrit level. Multivariable regression analyses were performed, adjusting for traditional risk factors and left ventricular structure. Results Women had significantly greater partition coefficient, ECV, and precontrast T1 than men, as well as lower post-contrast T1 values (all p < 0.05). In general, linear regression analyses demonstrated that greater partition coefficient, pre-contrast T1 values, and ECV were associated with older age in men (multivariate regression coefficients = 0.01; 5.9 ms; and 1.04% per 10 years’ change; all p < 0.05). ECV was also significantly associated with age in women after multivariable adjustments. Conclusions CMR parameters that have been associated with myocardial fibrosis were related to older age in the MESA study. Women had higher ECV than men but less ECV change over time.

No abstract

Purpose:To determine the utility of cardiac magnetic resonance (MR) T1 mapping for quantification of diffuse myocardial fibrosis compared with the standard of endomyocardial biopsy. Materials and Methods:This HIPAA-compliant study was approved by the institutional review board. Cardiomyopathy patients were retrospectively identified who had undergone endomyocardial biopsy and cardiac MR at one institution during a 5-year period. Forty-seven patients (53% male; mean age, 46.8 years) had undergone diagnostic cardiac MR and endomyocardial biopsy. Thirteen healthy volunteers (54% male; mean age, 38.1 years) underwent cardiac MR as a reference. Myocardial T1 mapping was performed 10.7 minutes 6 2.7 (standard deviation) after bolus injection of 0.2 mmol/kg gadolinium chelate by using an inversion-recovery Look-Locker sequence on a 1.5-T MR imager. Late gadolinium enhancement was assessed by using gradientecho inversion-recovery sequences. Cardiac MR results were the consensus of two radiologists who were blinded to histopathologic findings. Endomyocardial biopsy fibrosis was quantitatively measured by using automated image analysis software with digital images of specimens stained with Masson trichrome. Histopathologic findings were reported by two pathologists blinded to cardiac MR findings. Statistical analyses included Mann-Whitney U test, analysis of variance, and linear regression. Results:Median myocardial fibrosis was 8.5% (interquartile range, 5.7-14.4). T1 times were greater in control subjects than in patients without and in patients with evident late gadolinium enhancement (466 msec 6 14, 406 msec 6 59, and 303 msec 6 53, respectively; P , .001). T1 time and histologic fibrosis were inversely correlated (r = 20.57; 95% confidence interval: 20.74, 20.34; P , .0001). The area under the curve for myocardial T1 time to detect fibrosis of greater than 5% was 0.84 at a cutoff of 383 msec. Conclusion:Cardiac MR with T1 mapping can provide noninvasive evidence of diffuse myocardial fibrosis in patients referred for evaluation of cardiomyopathy.q RSNA, 2012

Pedestrian detection in a crowd is a very challenging issue. This paper addresses this problem by a novel Non-Maximum Suppression (NMS) algorithm to better refine the bounding boxes given by detectors. The contributions are threefold: (1) we propose adaptive-NMS, which applies a dynamic suppression threshold to an instance, according to the target density; (2) we design an efficient subnetwork to learn density scores, which can be conveniently embedded into both the single-stage and two-stage detectors; and (3) we achieve state of the art results on the CityPersons and CrowdHuman benchmarks.