Search

Weak texture and motion blur are always challenging problems for visual-inertial odometry (VIO) systems. To improve accuracy of VIO systems in the challenging scenes, we propose a point-line-based VIO system with novel feature hybrids and with novel predicting-matching for long line track. Point-line features with shorter tracks are categorized into “MSCKF” features and with longer tracks into “SLAM” features. Especially, “SLAM” lines are added into the state vector to improve accuracy of the proposed system. Besides, to ensure the reliability and stability of detection and tracking of line features, we also propose a new “Predicting-Matching” line segment tracking method to increase the track lengths of line segments. Experimental results show that the proposed method outperforms the state-of-the-art methods of VINS-Mono [1], PL-VINS [2] and OpenVINS [3]) on both a public dataset and a collected dataset in terms of accuracy. The collected dataset is full of extremely weak textures and motion blurs. On this dataset, the proposed method also obtains better accuracy than ORB-SLAM3 [4].

Visual Augmented Reality (AR) typically overlays virtual computer graphics or other virtual contents on the real world videos, attracting much interest from both academic and industrial communities. Although AR techniques on planes are well studied, cylindrical objects are seldom used for augmented reality. In this paper, we propose a new method for 3D reconstruction and 6D pose computation for augmented reality on a cylindrical object. The 6D pose is the relative pose between the camera and the cylindrical object, which is very convenient to make augmented reality. First, we capture some images with a cylindrical object and then reconstruct its 3D model with textures offline by using projective invariance and image contours. Second, according to the 3D model, we track the 6D relative pose between the camera and the cylindrical object online, where we propose a linear P3P RANSAC to remove outliers. Finally, the virtual images are exactly aligned with the cylindrical object in the real world. Experimental results show that the proposed method outperforms the state of the arts in terms of 3D mapping and 6D pose computation on cylindrical objects.

Omnidirectional cameras are becoming popular in various applications owing to their ability to capture the full surrounding scene in real-time. However, depth estimation for an omnidirectional scene is more difficult than normal perspective images due to its different system properties and distortions. It is hard to use normal depth estimation methods such as stereo matching or RGB-D sensing. A deep-learning-based single-shot depth estimation approach can be a good solution, but it requires a large labelled dataset for training. The 3D60 dataset, the largest omnidirectional dataset with depth labels, is not applicable for general scene depth estimation because it covers very limited scenes. In order to overcome this limitation, we propose a depth estimation architecture for a single omnidirectional image using domain adaptation. The proposed architecture gets labelled source domain and unlabelled target domain data together as its input and estimated depth information of the target domain using the Generative Adversarial Networks (GAN) based method. The proposed architecture shows >10% higher accuracy in depth estimation than traditional encoder-decoder models with a limited labelled dataset.

Background: Lower limb sensory disturbance presentation can be a false localizing cervical cord compressive myelopathy (CSM). It may lead to delayed or missed diagnosis, resulting in the wrong management plan, especially in the presence of concurrent lumbar lesions.Case presentation:Three Asian patients with lower limb sensory disturbances presentation were treated ineffectively in the lumbar. Magnetic resonance imaging (MRI) showed cervical disc herniation and cervical level spinal cord compression. Anterior cervical discectomy surgery and zero-p interbody fusion were performed. After operations, imagings showed that the spinal cord compression were relieved, and the lower limbs sensory disturbances were also relieved. Three-months follow-up after operation showed good recovery.Conclusions:These three cervical cord compression cases of lower limb sensory disturbance presentation were easily misdiagnosed with lumbar spondylosis. Anterior cervical discectomy and fusion operation had a good therapeutic effect. Therefore, cases that present with lower limb sensory disturbance, but in a non-radicular classical pattern, should always alert a suspicion of a possible cord compression cause at a higher level.

Vascular calcification (VC) involves the deposition of calcium apatite in vascular intima or media. Individuals of advanced age, having diabetes mellitus or chronic kidney disease (CKD) are particularly at risk. The pathogenesis of CKD-associated VC evolves considerably. The core driver is the phenotypic change involving vascular wall constituent cells toward manifestations similar to that undergone by osteoblasts. Gender-related differences are observed regarding the expressions of osteogenesis-regulating effectors, and presumably the prevalence/risk of CKD-associated VC exhibits gender-related differences as well. Despite the wealth of data focusing on gender-related differences in the risk of atherosclerosis, few report whether gender modifies the risk of VC, especially CKD-associated cases. We systematically identified studies of CKD-associated VC or its regulators/modifiers reporting data about gender distributions, and extracted results from 167 articles. A significantly higher risk of CKD-associated VC was observed in males among the majority of original investigations. However, substantial heterogeneity exists, since multiple large-scale studies yielded neutral findings. Differences in gender-related VC risk may result from variations in VC assessment methods, the anatomical segments of interest, study sample size, and even the ethnic origins of participants. From a biological perspective, plausible mediators of gender-related VC differences include body composition discrepancies, alterations involving lipid profiles, inflammatory severity, diversities in matrix Gla protein (MGP), soluble Klotho, vitamin D, sclerostin, parathyroid hormone (PTH), fibroblast growth factor-23 (FGF-23), and osteoprotegerin levels. Based on our findings, it may be inappropriate to monotonously assume that male patients with CKD are at risk of VC compared to females, and we should consider more background in context before result interpretation.

Spin-orbit torque offers an efficient route to manipulate the magnetic state of magnetic materials, which is of great importance for energy-efficient applications of various spintronic devices like memory, logic, oscillator, and neuromorphic computing. Here, we propose a strategy for the realization of a spin torque gate magnetic field sensor with an extremely simple structure by utilizing the longitudinal field dependence of the spin-orbit torque driven magnetization switching. This sensor does not require any magnetic bias to achieve a linear response to the external field, which is the main cause of high cost of all types of magnetoresistance sensors. In addition, zero offset can be achieved in the spin torque gate sensor without complicated offset compensation circuit. By employing the WTe2/Ti/CoFeB structure with both large spin-orbit torque and well-defined PMA, we demonstrate that the sensor can work linearly in the range of ±3-10 Oe with nearly zero dc offset.

Background: Generally, it is difficult to obtain accurate pose and depth for a non-rigid moving object from a single RGB camera to create augmented reality (AR). In this study, we build an augmented reality system from a single RGB camera for a non-rigid moving human by accurately computing pose and depth, for which two key tasks are segmentation and monocular Simultaneous Localization and Mapping (SLAM). Most existing monocular SLAM systems are designed for static scenes, while in this AR system, the human body is always moving and non-rigid. Methods: In order to make the SLAM system suitable for a moving human, we first segment the rigid part of the human in each frame. A segmented moving body part can be regarded as a static object, and the relative motions between each moving body part and the camera can be considered the motion of the camera. Typical SLAM systems designed for static scenes can then be applied. In the segmentation step of this AR system, we first employ the proposed BowtieNet, which adds the atrous spatial pyramid pooling (ASPP) of DeepLab between the encoder and decoder of SegNet to segment the human in the original frame, and then we use color information to extract the face from the segmented human area. Results: Based on the human segmentation results and a monocular SLAM, this system can change the video background and add a virtual object to humans. Conclusions: The experiments on the human image segmentation datasets show that BowtieNet obtains state-of-the-art human image segmentation performance and enough speed for real-time segmentation. The experiments on videos show that the proposed AR system can robustly add a virtual object to humans and can accurately change the video background. Keywords: Augmented reality, Moving object, Reconstruction and tracking, Camera pose, Human segmentation

The 3D reconstruction is an important topic in computer vision with many applications, such as robotics and augmented reality. Since the raw depth images captured by consumer RGB-D cameras are often low resolution (LR), noisy, and incomplete. How to obtain high-quality 3D models with a consumer RGB-D camera is still a challenge for the existing systems. In this paper, we propose a new depth super-resolution and completion method implemented in a deep learning framework and build a high-quality 3D reconstruction system. We first improve the resolution of LR depth image with a depth super-resolution network and remove the outliers in high-resolution (HR) depth image based on gradient saliency. To further enhance the quality of HR depth image with the guide of HR color image, we learn surface normal and occlusion boundary images from the corresponding HR color image through two deep fully convolutional networks. In particular, the blurriness of HR color image is also detected and pixel-wise quantized. Finally, we obtain a completed HR depth image by optimizing the HR depth image with the surface normal, occlusion boundary, and color image blurriness. We have carried out qualitative and quantitative evaluations with baseline methods on public datasets. The experimental results demonstrate that our method has better performance both on single depth image enhancement and 3D reconstruction.

In recent years, simultaneous localization and mapping in dynamic environments (dynamic SLAM) has attracted significant attention from both academia and industry. Some pioneering work on this technique has expanded the potential of robotic applications. Compared to standard SLAM under the static world assumption, dynamic SLAM divides features into static and dynamic categories and leverages each type of feature properly. Therefore, dynamic SLAM can provide more robust localization for intelligent robots that operate in complex dynamic environments. Additionally, to meet the demands of some high-level tasks, dynamic SLAM can be integrated with multiple object tracking. This article presents a survey on dynamic SLAM from the perspective of feature choices. A discussion of the advantages and disadvantages of different visual features is provided in this article.

Eutrophication refers to the nutrient enrichment, leading to blooms of algae. Such blooms in lakes can happen throughout the year because of the changes in nutrient and hydrological cycles. Nutrient export to lakes from rivers is the main cause of eutrophication problems. Seasonal trends in nitrogen (N) export by rivers to lakes are still not well understood. The objective of this study is, therefore, to better understand the seasonal trends in river export of dissolved inorganic N (DIN) to lakes of the Hai He Basin. To this end, we selected Guanting and Baiyangdian as representative lakes, whose drainage areas include various cropping systems. We developed a seasonal version of the Model to Assess River Inputs of Nutrients to lAkes (MARINA-Lakes) model for Guanting and Baiyangdian while assessing N flows from the land to the lakes. The model accounts for the seasonality in human activities (e.g., cropping systems and fertilizer practices), climate, and hydrology. The effective seasons are winter (December–February), spring (March–May), summer (June–August), and fall (September–November). The model results for the year 2012 indicate that river export of DIN was highest in winter and lowest in summer. Point sources accounted for over 50% of DIN exports to Guanting and Baiyangdian across seasons. Avoiding direct discharges of animal manure (point source) in winter is needed to reduce future lake pollution. We argue that effective lake pollution control requires accounting for seasonal N cycles. Our study can support effective nutrient management and environmental policies.

The Field Programmable Gate Array (FPGA) accelerator for CNN-based object detection has been attracting widespread attention in computer vision. For most existing FPGA accelerators, the inference accuracy and speed are affected negatively by the low power-efficient and performance-density. To address this problem, we propose a software and hardware co-designed FPGA accelerator for accurate and fast object detection with high power-efficient and performance-density. To develop the FPGA accelerator on CPU+FPGA heterogeneous platforms, a resource sensitive and energy aware FPGA accelerator framework is designed. In hardware, a hardware sensitive neural network quantization called Dynamic Fixed-point Data Quantization (DFDQ) is proposed to improve the power-efficient. In software, an algorithm-level convolution (CONV) optimization scheme is further proposed to improve the performance-density by paralleling block execution of CONV cores. To validate the proposed FPGA accelerator, a Zynq FPGA is used to build the acceleration platform of You Only Look Once (YOLO) network. Results demonstrate that the proposed FPGA accelerator outperforms the state-of-the-art methods in power-efficient and performance-density. Besides, the speed of object detection is increased by at most 16.5 times along with less than 1.5% accuracy degradation.

Visual localization to compute 6DoF camera pose from a given image has wide applications. Both local and global descriptors are crucial for visual localization. Most of the existing visual localization methods adopt a two-stage strategy: image retrieval first is performed by global descriptors, and then 2D-3D correspondences are made by local descriptors from 2D query image points and its nearest neighbor candidates which are the 3D points visible by these retrieved images. The above two stages are serially performed in these methods. However, due to the fact that 3D points obtained from the retrieval feedback are only rely on global descriptors, these methods cannot fully take the advantages of both local and global descriptors. In this paper, we propose a novel parallel search framework, which fully leverages advantages of both local and global descriptors to get nearest neighbor candidates of a 2D query image point. Specifically, besides using deep learning based global descriptors, we also utilize local descriptors to construct random tree structures for obtaining nearest neighbor candidates of the 2D query image point. We propose a new probability model and a new deep learning based local descriptor when constructing the random trees. In addition, a weighted Hamming regularization term to keep discriminativeness after binarization is given in loss function for the proposed local descriptor. The loss function co-trains both real and binary local descriptors of which the results are integrated into the random trees. Experiments on challenging benchmarks show that the proposed localization method can significantly improve the robustness and accuracy compared with the ones which get nearest neighbor candidates of a query local feature just based on either local or global descriptors.

In this paper, we present a deep learning-based approach to monocular visual odometry. We propose a LCGR(Local Convolution and Global RNN) module which utilizes several independent 3D convolution layers to filter noise from features extracted by FlowNet, as well as to model local information, and a Bi-ConvLSTM layer to model time series and capture global information. In addition, our network jointly predicts optical flow as an auxiliary task by measuring photometric consistency in a self-supervised way to help the encoder for better motion feature extraction. In order to alleviate the effects of non-Lambertian surfaces and dynamical objects in the scene, a confidence mask layer is estimated and epipolar constraint is added to the training process. Experiment results indicate competitive performance of the proposed framework to the state-of-art methods.

Visual localization is a key technology in the field of 3D robot vision. One of its major difficulties lies in how to deal with the challenges brought by the appearance changes of query images and database images caused by large time spans. Many methods focus on extracting more robust features from images to deal with the impact of complex scenes. In this paper, we explore the impact of image translation on visual localization tasks in complex scenes. We propose UniGAN - a modified image translation model, fusing semantic label constraints and finer reconstruction losses, to unify images captured under different environmental conditions to a standard scene more suitable for localization tasks. To estimate the 6-DOF camera pose, a two-stage localization framework composed of image retrieval and local matching is utilized. Experiments show that our method outperforms the state-of-the-art in terms of both accuracy and robustness to environmentally sensitive scenes.

Visual localization is a fundamental problem in computer vision and robotics. Recently, deep learning has shown to be effective for robust monocular localization. Most deep learning-based methods utilize convolution neural network (CNN) to regress global 6 degree-of-freedom (Dof) pose. However, these methods suffer from pose sparsity, leading to over-fitting during training and poor localization performance on unseen data. In this paper, we try to alleviate this issue by implementing randomly geometric augmentation (RGA) during training. Specifically, we firstly estimate the depth map using a depth estimation network for the initial training image. Combing the estimated depth, RGB image and its corresponding pose, we can randomly synthesize new images of different views. The synthesized and initial images are used to train the pose regression network. Experiment results show our geometric augmentation strategy can significantly improve the localization accuracy.

For the additive manufacturing (AM) of metal objects, the powder-based fusion (PBF) method is routinely utilized to fabricate macroscale parts. On the other hand, electrochemical additive manufacturing (ECAM), in which metallic structures are deposited through the electrochemical reduction of metal ions, is a promising technique for producing micro- and nanoscale objects. However, a gap exists in terms of fabricating mesoscale objects within the current AM techniques. The PBF method is limited by fabrication precision due to pronounced residual stresses, and most current ECAM systems are difficult to scale up to print mesoscale objects. In the present paper, the novel design of a low-cost ECAM 3D printer based on a microfluidic system is proposed for fabricating mesoscale metal parts. The meniscus-guided electrodeposition approach is utilized, in which a meniscus is formed between the print head and substrate, and electrodeposition is confined within the meniscus. A 3D object is fabricated by the meniscus moving with the print head according to the programmed pattern and the material subsequently being deposited at the designated locations. The key to the proposed design is to maintain a mesoscale meniscus, which normally cannot be sustained by the electrolyte surface tension with a print nozzle having a mesoscale diameter. Therefore, a microfluidic system, called the fountain pen feed system, constituting a semi-open main channel and comb structure, was designed to maintain a mesoscale meniscus throughout the printing process. Two materials, copper and nickel, with various geometric shapes were attempted to print by the proposed ECAM system, and, during the printing process, both fluid leaking and meniscus breaking were completely prevented. Free standing tilted copper pillars with controlled angles were printed to show the ability of the proposed design in fabricating 3D structures. A copper circuit was also printed on a non-conductive substrate to demonstrate a possible application of the proposed ECAM system in the fabrication of functional electronics.

Resistant starch (RS) is a variant of starch that is indigestible by human enzymes and has been acknowledged for multiple physiological benefits including attenuation of postprandial glycemia when incorporated into foods. Distarch phosphate is a RS type 4 (RS4) containing phosphodiester cross-links within and between starch molecules. Considering the importance of RS to human health, the present study aimed to investigate the dose response effect of a novel RS4 (potato-derived distarch phosphate - (VERSAFIBE 1490™) on acute postprandial glycemic responses compared with energy/available carbohydrate and sugar-matched control. The study was designed as a controlled, single-center randomized, single-blinded, cross-over trial, in which 31 healthy adults consumed a baked breakfast cereal bar containing 0, 10 or 20 g RS4 followed by serial blood samples over two hours to determine glucose and insulin concentrations and calculate the incremental area under the curve (iAUC). Results suggest that the addition of RS4 did not reduce iAUC glucose or insulin responses significantly (P > 0.05) at the doses provided. No significant changes in the glucose or insulin maximum concentration (Cmax) and time to reach maximal glucose and insulin concentrations (Tmax) were observed (P > 0.05). Overall, this particular RS4 did not affect measures of glycemia in healthy individuals at doses provided in ready-to-eat baked-good.

This paper proposes a novel deep learning based approach for monocular visual odometry (VO) called FlowVO-Net. Our approach utilizes CNN to extract motion information between two consecutive frames and employs Bi-directional convolution LSTM (Bi-ConvLSTM) for temporal modelling. ConvLSTM can encode not only temporal information but also spatial correlation, and the bidirectional architecture enables it to learn the geometric relationship from image sequences pre and post. Besides, our approach jointly predicts optical flow as an auxiliary task in a self-supervised way by measuring photometric consistency. Experiment results indicate competitive performance of the proposed FlowVO-Net to the state-of-art methods.

Densely connected convolutional networks (DenseNet) have reached unprecedented parameter-performance efficiencies and alleviated problems of vanishing gradients by concatenating each layer to every other layer. However, with the increase in network depth, the cross-channel interaction of dense blocks has become increasingly complex. Hence, it is now more difficult to optimize networks. Moreover, the way combining features in DenseNet restricts its flexibility and scalability in learning more expressive combination strategies. In this study, we aim to answer the question of how to simultaneously ensure the benefits of feature reuse, reduce the complexity of cross-channel interactions, and increase the flexibility of the network. Hence, the components of DenseNet are refined and then used as a basis to develop a universal densely connected convolutional network (UDenseNet). Based on the proposed architecture, the impact of different component configurations on the network performance is empirically analyzed to determine the optimal architectural configuration. Extensive experiments are conducted to validate the proposed UDenseNet on benchmark datasets (CIFAR, SVHN and ImageNet). Results show that, compared to most other methods, the proposed UDenseNet can significantly improve performance in image recognition tasks.

Automatic and reliable segmentation of the hippocampus from magnetic resonance (MR) brain images is extremely important in a variety of neuroimage studies. To improve the hippocampus segmentation performance, a local binary pattern based feature extraction method is developed for machine learning based multi-atlas hippocampus segmentation. Under the framework of multi-atlas image segmentation (MAIS), a set of selected atlases are registered to images to be segmented using a non-linear image registration algorithm. The registered atlases are then used as training data to build linear regression models for segmenting the images based on the image features, referred to as random local binary pattern (RLBP), extracted using a novel image feature extraction method. The RLBP based MAIS algorithm has been validated for segmenting hippocampus based on a data set of 135 T1 MR images which are from the Alzheimer’s Disease Neuroimaging Initiative database (adni.loni.usc.edu). By using manual segmentation labels produced by experienced tracers as the standard of truth, six segmentation evaluation metrics were used to evaluate the image segmentation results by comparing automatic segmentation results with the manual segmentation labels. We further computed Cohen’s d effect size to investigate the sensitivity of each segmenting method in detecting volumetric differences of the hippocampus between different groups of subjects. The evaluation results showed that our method was competitive to state-of-the-art label fusion methods in terms of accuracy. Hippocampal volumetric analysis showed that the proposed RLBP method performed well in detecting the volumetric differences of the hippocampus between groups of Alzheimer’s disease patients, mild cognitive impairment subjects, and normal controls. These results have demonstrated that the RLBP based multi-atlas image segmentation method could facilitate efficient and accurate extraction of the hippocampus and may help predict Alzheimer’s disease. The codes of the proposed method is available (https://www.nitrc.org/frs/?group_id=1242).

Place recognition as one of the most significant requirements for long-term simultaneous localization and mapping (SLAM) has been developed rapidly in recent years. Also, deep learning is proved to be more capable than traditional methods to extract features under some complex environments. However, in real-world environments, there are many challenging problems such as viewpoint changes and illumination changes. The existing deep learning-based place recognition in extracting feature phases and matching process is both time-consuming. Moreover, features extracted from convolution neural network (CNN) are floating-point type with high dimension. In this paper, we propose deep supervised hashing for place recognition, where we design a similar hierarchy loss function to learn a model. The model can distinguish the similar images more accurately which is well suitable to place recognition. Besides the model can learn high quality hash codes by maximizing the likelihood of triplet labels. Experiments on several benchmark datasets for place recognition show that our approach is robust to viewpoints, illuminations and season changes with high accuracy. Furthermore, the trained model can extract features and match in real time on CPU with less memory consumption.

Generally, due to the partial occlusion, it is very difficult to correctly estimate the central catadioptric line image from its visible part. Except for the necessary and sufficient conditions that must be satisfied by a set of paracatadioptric line images, we find that if the antipodal points of image points on the visible arc are known, the fitting accuracy can be improved greatly. In this paper, we propose a new method for fitting line images, which can be applied to all central catadioptric projection. Firstly, a new relationship between a pair of antipodal image points and the camera principal point is derived. Next, using this relationship, a new method is proposed to estimate the line images. These estimated line images are used to calibrate camera intrinsic parameters to evaluate the performance of our fitting method. Experimental results on both simulated and real data have demonstrated the effectiveness of our method.

Semantic segmentation of medical images with deep learning models is rapidly being developed. In this study, we benchmarked state-of-the-art deep learning segmentation algorithms on our clinical stereotactic radiosurgery dataset. The dataset consists of 1688 patients with various brain lesions (pituitary tumors, meningioma, schwannoma, brain metastases, arteriovenous malformation, and trigeminal neuralgia), and we divided the dataset into a training set (1557 patients) and test set (131 patients). This study demonstrates the strengths and weaknesses of deep-learning algorithms in a fairly practical scenario. We compared the model performances concerning their sampling method, model architecture, and the choice of loss functions, identifying suitable settings for their applications and shedding light on the possible improvements. Evidence from this study led us to conclude that deep learning could be promising in assisting the segmentation of brain lesions even if the training dataset was of high heterogeneity in lesion types and sizes.

Rationale Diffuse idiopathic skeletal hyperostosis (DISH) is a skeletal disease characterized by calcification of anterolateral ligaments of the spine and the rest of the body. DISH combined with disc herniation induces complex symptoms and is more difficult to treat. Here, we describe a complicated case of a patient diagnosed with DISH as well as cervical disc herniation that was successfully treated using anterior cervical surgery. Patients concern A 69-year-old Asian male experienced dysphagia and weakness in his left limbs. He also experienced a stiff neck and right slant over a 6-month period. Diagnosis An X-ray revealed calcification of the C4-7 vertebral front edge and a narrowed intervertebral space between C5/6. Coronal and sagittal computerized tomography (CT) and magnetic resonance imaging (MRI) both showed compression of the spinal cord at the cervical 5/6. Esophagography revealed that osteophytes in the anterior margin of vertebral body (C4-C7) oppressed the esophagus. Interventions An operation focused on anterior cervical C5/6 disc fusion and anterior vertebral C4-7 osteophyte removal was performed. Outcomes After the operation, X-ray and CT showed that most osteophytes were removed and spinal cord compression was relieved. One day following the operation, both dysphagia and numbness in limbs were improved. Strong recovery was observed at the three-month follow-up exam. Lessons This complex DISH combined with disc herniation case is rarely observed in patients. Anterior cervical disc fusion and anterior vertebral osteophyte removal were effective treatment measures. This case study provides insight into treating cases presented with cervical spine complications associated with DISH combined with other ailments.

The present study was aimed to evaluate the isoxanthanol against Staphylococcus aureus chronic obstructive pulmonary disease (COPD) in rat model. The isoxanthanol decreased the parasitic load by almost 99% in the Staphylococcus aureus infected rats. It significantly (P 0.05) decreased mortality rate of the rats, prevented pulmonary tissue damage and aggregation of inflammatory cytokines. In Staphylococcus aureus infected rats, isoxanthanol treatment inhibited production of interleukin-18, interleukin-1β and TNF-α significantly (P 0.05) in the BALF and pulmonary tissues. Treatment of the Staphylococcus aureus-infected rats with isoxanthanol inhibited up-regulation of NLRP3, ASC and caspase-1 expression. In Staphylococcus aureus-infected rats the expression of miR-145-5p was remarkably increased on treatment with isoxanthanol. In summary, isoxanthanol prevents Staphylococcus aureus-induced COPD in rats through up-regulation of miR-145-5p and suppression of inflammatory cytokines. Therefore, isoxanthanol can be of therapeutic importance for the treatment of Staphylococcus aureus induced COPD.

Memristive devices are able to store and process information, which offers several key advantages over the transistor-based architectures. However, most of the two-terminal memristive devices have fixed functions once made and cannot be reconfigured for other situations. Here, we propose and demonstrate a memristive device "memlogic" (memory logic) as a nonvolatile switch of logic operations integrated with memory function in a single light-gated memristor. Based on nonvolatile light-modulated memristive switching behavior, a single memlogic cell is able to achieve optical and electrical mixed basic Boolean logic of reconfigurable "AND", "OR", and "NOT" operations. Furthermore, the single memlogic cell is also capable of functioning as an optical adder and digital-to-analog converter. All the memlogic outputs are memristive for in situ data storage due to the nonvolatile resistive switching and persistent photoconductivity effects. Thus, as a memdevice, the memlogic has potential for not only simplifying the programmable logic circuits but also building memristive multifunctional optoelectronics.

A visual simultaneous localization and mapping (SLAM) system usually contains a relocalization module to recover the camera pose after tracking failure. The core of this module is to establish correspondences between map points and key points in the image, which is typically achieved by local image feature matching. Since recently emerged binary features have orders of magnitudes higher extraction speed than traditional features such as scale invariant feature transform, they can be applied to develop a real-time relocalization module once an efficient method of binary feature matching is provided. In this paper, we propose such a method by indexing binary features with hashing. Being different from the popular locality sensitive hashing, the proposed method constructs the hash keys by an online learning process instead of pure randomness. Specifically, the hash keys are trained with the aim of attaining uniform hash buckets and high collision rates of matched feature pairs, which makes the method more efficient on approximate nearest neighbor search. By distributing the online learning into the simultaneous localization and mapping process, we successfully apply the method to SLAM relocalization. Experiments show that camera poses can be recovered in real time even when there are tens of thousands of landmarks in the map.

Paralleled high-voltage SF6 circuit breakers with a highly coupled split reactor (HCSR) could realize the function of automatic current-sharing and current-limiting. Aiming at investigating the arc characteristics of paralleled SF6 circuit breakers with an HCSR, the single-phase tests for the paralleled high-voltage SF6 circuit breakers with an HCSR were carried out, including the opening and closing–opening operations. The influence of the capacitor paralleled to one of the breaks was studied. A mathematical model was developed for the paralleled high-voltage SF6 circuit breakers with an HCSR by coupling the equivalent circuit equations into the magnetohydrodynamic (MHD) arc model through the embedded codes, laying the foundation for the further research. The variation of the short-circuit currents in the branches during the breaking process was analyzed in detail. The current-sharing and current-limiting effects are quite satisfying. The temperature distribution of both arc chambers was presented. The calculated results were in agreement with the experimental results to some extent.

With constructions of demonstrative microgrids, the realistic optimal economic dispatch and energy management system are required eagerly. However, most current works usually give some simplifications on the modeling of microgrids. This paper presents an optimal day-ahead scheduling model for a microgrid system with photovoltaic cells, wind turbine units, diesel generators and battery storage systems. The power flow constraints are introduced into the scheduling model in order to show some necessary properties in the low voltage distribution network of microgrids. Besides a hybrid harmony search algorithm with differential evolution (HSDE) approach to the optimization problem is proposed. Some improvements such as the dynamic F and CR , the improved mutation, the additional competition and the discrete difference operation have been integrated into the proposed algorithm in order to obtain the competitive results efficiently. The numerical results for several test microgrids adopting the IEEE 9-bus, IEEE 39-bus and IEEE 57-bus systems to represent their transmission networks are employed to show the effectiveness and validity of the proposed model and algorithm. Not only the normal operation mode but also some typical fault modes are used to verify the proposed approach and the simulations show the competitiveness of the HSDE algorithm.

Frailty exhibits diverse influences on health-related outcomes and represents a surrogate of increased susceptibility to harmful injuries. Patients with chronic kidney disease (CKD) are at a higher risk of accelerated biologic aging, and, in this population, the concept of frailty emerges as an instrumental measurement of physiologic reserves. However, a comprehensive description of known independent contributors to, and risk associates of, frailty in these patients remain unavailable. In the present review, original studies up to 28 February 2019 that assessed frailty in patients with all stages of CKD were retrieved and reviewed, with results extracted and summarized. By pooling 62 original investigations, 58.1% and 49.1% used cohort and cross-sectional designs, respectively. Dialysis-dependent end-stage renal disease patients ( n = 39; 62.9%) were the most commonly examined population, followed by those with nondialysis CKD ( n = 12; 19.4%) and those receiving renal transplantation ( n = 11; 17.7%). Contributors to frailty in CKD patients included sociodemographic factors, smoking, CKD severity, organ-specific comorbidities, depression, hypoalbuminemia, and low testosterone levels. Conversely, the development of frailty was potentially associated with the emergence of cardiometabolic, musculoskeletal, and cerebral complications; mental distress; and a higher risk of subsequent functional and quality-of-life impairment. Moreover, frailty in CKD patients increased healthcare utilization and consistently elevated mortality among affected ones. Based on the multitude of contributors to frailty and its diverse health influences, a multifaceted approach to manage CKD patients with frailty is needed, and its potential influences on outcomes besides mortality need to be considered.

Accurate hand segmentation is important for human segmentation. However, in videos, hand regions usually have serious motion blur, which reduces segmentation performance obviously. To solve this problem, we propose an automatic matting network to deal with motion-blurred hands. Then we combine the hand alpha mattes provided by matting network and the human segmentation results provided by segmentation network to generate our final human soft segmentation results. In addition, to train the matting network, we need a huge amount of motion-blurred hand images and their groundtruth alpha mattes. However these images are very difficult to obtain. To solve this problem, we propose an efficient semi-automatic synthetic data generation method and generate 36186 synthetic motion-blurred hand images and their alpha mattes. Experiments on synthetic images and real videos show that our method achieves state-of-art matting performance and successfully solve the problem of bad hand segmentation caused by serious motion blur.

In human parsing task, it is important to fully exploit global and local structure information and get accurate and coherent results. In this paper, we propose a classification assisted segmentation network, in which a multi-label classification task can obtain the probability of each class in an image that used to learn better weights for parsing task. Our method takes advantages of both the global information from classification and the detail information from segmentation. Experiments demonstrate that our method could efficiently avoid the confusion between similar categories and get more reasonable results. Particularly, it significantly boosts performances of rare categories such as scarf, belt and sunglasses with mean IoU increased by 6.29%.

Abnormal cervical curvature and cervical disc herniation are closely related to neck pain and should be taken into account before any treatment. However, studies have rarely reported on the correlation between cervical lordosis and cervical disc herniation in patients with neck pain. Therefore, in this study, we collect young neck pain patients with abnormal cervical curvature to evaluate the relationship between cervical lordosis and cervical disc herniation. Three hundred patients below 40 years old with neck pain were enrolled. Patient sex, age, apical vertebra, segment of intervertebral disc protrusionl, sagittal diameter of spinal duramater, saggital diameter of spinal canal, height of disc space were recorded, and the cervical curvature, and degree of cervical spinal cord compression (G/F ratio) were calculated. The change of degree of disc herniation and degree of cervical spinal cord compression were analyzed in different cervical curvature groups. Further more, collected these patients who had improved cervical curvature over a period of time, to compare the changes of degree of disc herniation, G/F ratio, and height of disc space. The median age of patients with kyphosis was lower than those with lordosis and straight cervical spine. The degree of disc herniation was higher in the straight and kyphosis groups compared to the lordosis group. Cervical lordosis was inversely correlated with the degree of disc herniation and positively with G/F ratio. Cervical curvature was significantly affected by sex, age, and the degree of disc herniation. With the improvement of cervical lordotic curvature, the degree of disc herniation decreased and height of disc space increased. The degree of disc herniation and cervical spinal cord compression are inversely correlated to cervical lordosis in young neck pain patients, and the degree of disc herniation and height of disc space can recover with the recovery of cervical lordotic curvature. These findings may indicating a link between cervical curvature and degenerative changes which have important clinical implications.

International audience; We develop and exploit a new catalog of coronal pressure waves modeled in 3D to study the potential role of these waves in accelerating solar energetic particles (SEPs) measured in situ. Our sample comprises modeled shocks and SEP events detected during solar cycle 24 observed over a broad range of longitudes. From the 3D reconstruction of shock waves using coronagraphic observations we derived the 3D velocity along the entire front as a function of time. Combining new reconstruction techniques with global models of the solar corona, we derive the 3D distribution of basic shock parameters such as Mach numbers, compression ratios, and shock geometry. We then model in a time-dependent manner how the shock wave connects magnetically with spacecraft making in situ measurements of SEPs. This allows us to compare modeled shock parameters deduced at the magnetically well-connected regions, with different key parameters of SEPs such as their maximum intensity. This approach accounts for projection effects associated with remote-sensing observations and constitutes the most extensive study to date of shock waves in the corona and their relation to SEPs. We find a high correlation between the maximum flux of SEPs and the strength of coronal shock waves quantified, for instance, by the Mach number. We discuss the implications of that work for understanding particle acceleration in the corona.

We propose a novel stereo visual SLAM framework considering both accuracy and speed at the same time. The framework makes full use of the advantages of key-feature-based multiple view geometry (MVG) and direct-based formulation. At the front-end, our system performs direct formulation and constant motion model to predict a robust initial pose, reprojects local map to find 3D-2D correspondence and finally refines pose by the reprojection error minimization. This frontend process makes our system faster. At the back-end, MVG is used to estimate 3D structure. When a new keyframe is inserted, new mappoints are generated by triangulating. In order to improve the accuracy of the proposed system, bad mappoints are removed and a global map is kept by bundle adjustment. Especially, the stereo constraint is performed to optimize the map. This back-end process makes our system more accurate. Experimental evaluation on EuRoC dataset shows that the proposed algorithm can run at more than 100 frames per second on a consumer computer while achieving highly competitive accuracy.

Hai He basin is located in the densely populated North China Plain that is providing food to more than 100 million people. The fast developing agriculture and urbanization in Hai He basin have resulted in discharging nutrient-rich wastewater into lakes and reservoirs, leading to eutrophication and water scarcity such as in Guanting reservoir and Baiyangdian lake. In this study we analyzed future trends in nutrient inputs into Guanting reservoir and Baiyangdian lake by applying the MARINA (Model to Assess River Inputs of Nutrients to seAs) model. We analyzed trends between 2012 and 2050 for a business-as-usual scenario (SSP3) and a scenario based on Current Environmental Policies (CEP). In addition, we assessed future impacts of two important events on river export of nutrients: the 2022 Olympic Winter Games in the Guanting basin (OLY scenario), and the development of Xiong’an in the Baiyangdian basin (URB scenario). Finally, we assumed implementation of advanced technologies to reduce nutrients in rivers (OLY+ and URB+). Our study has five main findings. First, nutrients in Guanting reservoir and Baiyangdian lake were mainly from agriculture in 2012. Second, nutrient export doubles between 2012 and 2050 in SSP3. Third, effective implementation of current environmental policies could reduce the future pollution to levels below that in 2012. Fourth, improved sewage systems associated with the 2022 Winter Olympic Games could not reduce nutrient pollution effectively in Guanting reservoir, indicating that reducing nutrient losses from agriculture may be more effective to improve water quality than urban waste water treatment. Fifth, urbanization in the Baiyangdian basin may increase river export of nutrients to the lake by 28-43% compared to the CEP scenario (URB scenario). Highly effective waste treatment is needed not only in Xiong’an but also in surrounding areas to ensure the availability of clean water (URB+ scenario). Our results could improve our understanding of nutrient management for specific lakes and reservoirs, and highly relevant for policy making for effective environmental policies.

When we use video chat, video game, or other video applications, motion-blurred hands often appear. Accurately extracting these hands is very useful for video editing and behavior analysis. However, existing motion-blurred object extraction methods either need user interactions, such as user supplied trimaps and scribbles, or need additional information, such as background images. In this letter, a novel method which can automatically extract the semi-transparent motion-blurred hand just according to the original RGB image is proposed. The proposed method separates the extraction task into two subtasks: alpha matte prediction and foreground prediction. These two subtasks are implemented by Xception based encoder-decoder networks. The images of extracted motion-blurred hands are calculated by multiplying the predicted alpha mattes and foreground images. Experiments on synthetic and real datasets show that the proposed method has promising performance.

End-to-end visual navigation based on deep reinforcement learning (DRL) has recently attracted much attention. For most existing navigation methods, a robot moves along only fixed directions (e.g., up, down, left and right) on a grid. Obviously, they are not flexible and efficient, which worsens the navigation performance (i.e., the distance of movement and times of rotation). To address this problem, we propose a novel pose-guided end-to-end visual navigation framework, which is flexible and efficient. In the pose-guided navigation framework, a robot can move along arbitrary directions, which are determined by poses between adjacent objects. Further, to select a proper motion and finally form an optimal path, we propose a DRL based action-selected strategy, where a dynamic action select space on the basis of deep siamese actor-critic network is developed. Besides, to validate the proposed method, we propose a novel pose-guided dataset. Experimental results demonstrate that the proposed method outperforms the state of the arts in both flexibility and efficiency.

RATIONALE Anatomical variations in aortic arch (AA) branching are not unusual. Generally, these variations are asymptomatic and are diagnosed incidentally. Here, we report a rare case of a middle-aged female patient with an aberrant right subclavian artery (ARSA) associated with anomalous origins of the bilateral vertebral arteries (VAs). PATIENT CONCERNS The patient treated for urolithiasis complained of repeated dizziness for several years. DIAGNOSES Echocardiography and computed tomography angiography (CTA) confirmed arterial variations. Moreover, mild stenosis was found in the left common carotid artery (LCCA), which was considered to be the cause of dizziness. INTERVENTIONS Congenital anomalous arteries were not necessary to intervene urgently, but aspirin and atorvastatin were administered to prevent potential thrombosis attributed to vascular stenosis after completing the operation for urolithiasis. OUTCOMES Whether the symptoms will be alleviated or not should be continuously followed up, and the patient may accept interventional therapy in the future if necessary. LESSONS Here, we report the rare variation of AA branches and highlight the importance of preoperative vascular assessment in surgical or interventional procedures for the affected body regions.

Topological insulators (TIs) show bright prospects in exerting spin–orbit torques (SOTs) and inducing magnetization switching in the adjacent ferromagnetic (FM) layer. However, a variation of the SOT efficiency values may be attributed to the ex situ deposition of the FM layer or the complex capping/decapping processes of the protection layer. We have employed an in situ fabrication of Bi2Se3/Fe heterostructures and investigated the SOT efficiency by spin torque ferromagnetic resonance. An enhanced SOT efficiency and large effective spin mixing conductance have been obtained especially below 100 K as compared with ex situ methods. The enhancement of the SOT efficiency is attributed to a much thinner interfacial layer (0.96 nm) in the in situ case and thus the enhanced interface spin transparency. Our results reveal the crucial role of interface engineering in exploring highly efficient TI-based spintronic devices.

Objective: To explore the effect of transitional care mode on the exclusive breastfeeding rate of patients with gestational diabetes mellitus patients (GDM). Methods: A retrospective survey of patients who had undergone regular obstetrics and delivery in the hospital was carried out from December 2016 to December 2017. A random selection of 110 patients with GDM was included as the control group, and routine measures were taken. A total of 110 patients with GDM were selected as the experiment group. The whole nursing intervention was conducted with nurses in transitional care department on the basis of routine measures. The patients’ knowledge about breastfeeding and the breastfeeding rate of the two groups were compared and analyzed. Results: A total of 220 patients with GDM were recruited. The awareness rate of breastfeeding could help to regulate glucose and lipid metabolism of mothers and newborns were higher than that in the control group (P

To find the optimum threshold of an image is still an important research topic in the recent years. This paper presents a segmentation of liver cyst for ultrasound image through combining Wellner's thresholding algorithm with particle swarm optimization (PSO). The proposed method firstly obtains an optimal parameter, which expressed as a percentage or fixed amount of dark objects against a white background in a gray image, of Wellner's thresholding algorithm by PSO method. And then the gray image is binarized according to the optimized parameter. Finally, a semi-automatic method for locating and identifying multiple liver cysts or single liver cyst of ultrasound images is performed. For a validation, the results of the proposed technique are compared with those of other segmented methods. We also tested 92 ultrasound images of the liver cysts by our software. The corrected identification rate of the single liver cysts is 97.7 %, and that of multiple liver cysts is 87.5 %. Experimental results demonstrate that the proposed technique is reliable on segmenting the contour of liver cyst and identifying single or multiple liver cysts.

A novel label fusion method for multi-atlas based image segmentation method is developed by integrating semi-supervised and supervised machine learning techniques. Particularly, our method is developed in a pattern recognition based multi-atlas label fusion framework. We build random forests classification models for each image voxel to be segmented based on its corresponding image patches of atlas images that have been registered to the image to be segmented. The voxelwise random forests classification models are then applied to the image to be segmented to obtain a probabilistic segmentation map. Finally, a semi-supervised label propagation method is adapted to refine the probabilistic segmentation map by propagating its reliable voxelwise segmentation labels, taking into consideration consistency of local and global image appearance of the image to be segmented. The proposed method has been evaluated for segmenting the hippocampus in MR images and compared with alternative machine learning based multi-atlas based image segmentation methods. The experiment results have demonstrated that our method could obtain competitive segmentation performance (average Dice index > 0.88), compared with alternative multi-atlas based image segmentation methods under comparison. Source codes of the methods under comparison are publicly available at www.nitrc.org/frs/?group_id=1242.

3D scene reconstruction is an important topic in computer vision. A complete scene is reconstructed from views acquired along the camera trajectory, each view containing a small part of the scene. Tracking in textureless scenes is well known to be a Gordian knot of camera tracking, and how to obtain accurate 3D models quickly is a major challenge for existing systems. For the application of robotics, we propose a robust CPU-based approach to reconstruct indoor scenes efficiently with a consumer RGB-D camera. The proposed approach bridges feature-based camera tracking and volumetric-based data integration together and has a good reconstruction performance in terms of both robustness and efficiency. The key points in our approach include: (i) a robust and fast camera tracking method combining points and edges, which improves tracking stability in textureless scenes, (ii) an efficient data fusion strategy to select camera views and integrate RGB-D images on multiple scales, which enhances the efficiency of volumetric integration, (iii) a novel RGB-D scene reconstruction system, which can be quickly implemented on a standard CPU. Experimental results demonstrate that our approach reconstructs scenes with higher robustness and efficiency compared to state-of-the-art reconstruction systems.

Virtual reality, augmented reality, robotics, and autonomous driving, have recently attracted much attention from both academic and industrial communities, in which image-based camera localization is a key task. However, there has not been a complete review on image-based camera localization. It is urgent to map this topic to enable individuals enter the field quickly. In this paper, an overview of image-based camera localization is presented. A new and complete classification of image-based camera localization approaches is provided and the related techniques are introduced. Trends for future development are also discussed. This will be useful not only to researchers, but also to engineers and other individuals interested in this field.

The Magnetic Resonance Images (MRI) which can be used to segment brain tumors are 3D images. To make use of 3D information, a method that integrates the segmentation results of 3 2D Fully Convolutional Neural Networks (FCNNs), each of which is trained to segment brain tumor images from axial, coronal, and sagittal views respectively, is applied in this paper. Integrating multiple FCNN models by fusing their segmentation results rather than by fusing into one deep network makes sure that each FCNN model can still be tested by 2D slices, guaranteeing the testing efficiency. An averaging strategy is applied to do the fusing job. This method can be easily extended to integrate more FCNN models which are trained to segment brain tumor images from more views, without retraining the FCNN models that we already have. In addition, 3D Conditional Random Fields (CRFs) are applied to optimize our fused segmentation results. Experimental results show that, integrating the segmentation results of multiple 2D FCNNs obviously improves the segmentation accuracy, and 3D CRF greatly reduces false positives and improves the accuracy of tumor boundaries.

Fitting conic from images is a preliminary step for its plentiful applications. It's a common sense that geometric distance based fitting methods are better than algebraic distance based ones. However, for a long time, there has not been a geometric distance between a data point and a general conic that allows easy computation and achieves high accuracy simultaneously. In this paper, we derive a new geometric distance between a data point and a conic by revisiting Sampson distance. The new geometric distance is accurate and simultaneously still explicit analytical representation, which is greatly easy to be implemented. Then, based on the distance, a new cost function with combining Sampson distance is constructed. The conic fitting optimization by minimizing this cost function has all the merits of the geometric distance based methods and simultaneously avoids their limitations.

Ferromagnetic nanofilms with high saturation magnetization (Ms), particularly FeCo alloys with the highest Ms value (theoretically 2.45T), are promising candidates for emerging compact high-performance magnetoelectronic, microwave, and spintronic devices. Thin films with diverse and specific dynamic microwave properties are essential components for these different applications. For example, a high-frequency mobile phone requires a ferromagnetic element of high resonant frequency (fr) ; [4] a thin single-band microwave absorber requires ferromagnetic films with large resonance-type permeability (m), whereas a broadband absorber requires films with wide frequency dispersion or large relaxation-type m. The demand for a wide range of functions poses challenges for both scientific understanding and fabrication of these thin films, in particular, for applications in gigahertz (GHz) range. Hence despite being one of the earliest themes in magnetism, the topic of ferromagnetic thin films for high-frequency applications still generates much interest from scientists and engineers. Although the influence of the composition of materials on static or dynamic magnetic properties is well studied and described mathematically, for example, by Snoek’s law and constant (S), 3, 15] most formulae are insensitive to the crystal structure or nanostructure of thin ferromagnetic films. The understanding of the influence of an altered structure on the static and dynamic magnetic properties is still limited. A notable example is the preparation of appropriate nanostructure or crystal texture of soft ferromagnetic thin films with high Ms, such as FeCo, for achieving low coercivity (Hc 50 Oe), as well as ultrahigh fr ( 5 GHz) or wide frequency dispersion ( 3 GHz) by simple diverse atom arrangements without the addition of other materials. 17] Due to the inherent limitations of pure homogeneous ferromagnetic single-layer films (e.g. undesirable magnetic softness, anisotropy, or crystallinity of FeCo films), and a lack of evidence on the relationship between the nanostructure or crystal texture and dynamic microwave properties, doping or insertion of nonmagnetic materials is often used to address the stringent requirements of functional devices. This has be achieved by adding AlOx to a FeCo(HfO) film to attain a high fr ( 3 GHz), utilizing FeCo–MnIr exchange-coupled multilayers for wideband microwave noise filtering, or by doping Zr or B into a ferromagnetic film to adjust the m type. However, there are two critical challenges associated with these methods. The foreign elements decrease the Ms of the ferromagnetic film and subsequently deteriorate the device efficiency or performance. Moreover, most of the modifications are tailored The application of a pure FeCo film directly to devices is limited by the intrinsic properties of a homogeneous monolayer of the alloy, despite it having the highest saturation magnetization. A feasible methodology based on alternating current-density electrodeposition to tune the microwave properties of FeCo films through diverse atom-stacking arrangements is reported. The properties range from a large relative resonance permeability (up to 728 in the real part) to ultrahigh resonant frequency (up to 5.2 GHz), and extremely wide frequency dispersion (1.5–6.0 GHz); these are not observed in other materials or as a result of conventional doping methods. They are attained using different single-layer, semi-multilayer, and multilayer FeCo nanofilms, fabricated by diverse stacking arrangements of Fe and Co atoms during deposition in a single bath. This technique firstly exemplifies the significant effect of atom arrangement on magnetic inhomogeneity, nanoscale morphology, composition, crystal texture, stress, and damping, which in turn largely alters both the static and dynamic magnetic properties of the ferromagnetic film. These thin FeCo films might be used directly in various devices operating in the gigahertz frequency range.

A novel groupwise image registration framework is developed for registering MR brain images with tumors. Our method iteratively estimates a normal-appearance counterpart for each tumor image to be registered and constructs a directed graph (digraph) of normal-appearance images to guide the groupwise image registration. Particularly, our method maps each tumor image to its normal appearance counterpart by identifying and inpainting brain tumor regions with intensity information estimated using a low-rank plus sparse matrix decomposition based image representation technique. The estimated normal-appearance images are groupwisely registered to a group center image guided by a digraph of images so that the total length of 'image registration paths' to be the minimum, and then the original tumor images are warped to the group center image using the resulting deformation fields. We have evaluated our method based on both simulated and real MR brain tumor images. The registration results were evaluated with overlap measures of corresponding brain regions and average entropy of image intensity information, and Wilcoxon signed rank tests were adopted to compare different methods with respect to their regional overlap measures. Compared with a groupwise image registration method that is applied to normal-appearance images estimated using the traditional low-rank plus sparse matrix decomposition based image inpainting, our method achieved higher image registration accuracy with statistical significance (p = 7.02 × 10-9).