Many existing approaches for mammogram analysis are based on single view. Some recent DNN-based multi-view can perform either bilateral or ipsilateral analysis, while in practice, radiologists use both to achieve the best clinical outcome. MommiNet is first tri-view mass identification approach, which simultaneously and of mammographic images, turn, fully emulate radiologists' reading practice. In this paper, we present MommiNet-v2, with improved network architecture performance. Novel...

Current state-of-the-art document retrieval solutions mainly follow an index-retrieve paradigm, where the index is hard to be directly optimized for final target. In this paper, we aim show that end-to-end deep neural network unifying training and indexing stages can significantly improve recall performance of traditional methods. To end, propose Neural Corpus Indexer (NCI), a sequence-to-sequence generates relevant identifiers designated query. optimize NCI, invent prefix-aware...

At present, ultrasound is one of the essential tools for noninvasive medical diagnosis. However, speckle noise inherent in images and it cause decreased resolution contrast-to-noise ratio. Low image quality an obstacle effective feature extraction, recognition, analysis, edge detection; also affects interpretation by doctor accuracy computer-assisted diagnostic techniques. Thus, reduction significant critical step pre-processing images. Many techniques have been studied researchers, but to...

Ring artifacts in cone beam computed tomography (CBCT) images are caused by pixel gain variations using flat-panel detectors, and may lead to structured non-uniformities deterioration of image quality. The purpose this study is propose a method general ring artifact removal CBCT images. This based on the polar coordinate system, where manifest as stripe artifacts. Using relative total variation, first smoothed generate template with fewer details By subtracting from images, residual...

Blind image quality assessment can be modeled as feature extraction followed by score prediction. It necessitates considerable expertise and efforts to handcraft features for optimal representation of perceptual quality. This paper addresses blind sharpness using a shallow convolutional neural network (CNN). The takes single layer unearth intrinsic utilizes multilayer perceptron (MLP) rate Different from traditional methods, CNN integrates prediction into an optimization procedure retrieves...

Tumor recurrence and bone defects are two key challenges in the surgical treatment of osteosarcoma (OS). Therefore, it is highly necessary to develop a multifunctional scaffold that can simultaneously eradicate tumor cells promote regeneration. Herein, hierarchically porous shape memory consisting hydroxyapatite, silica, Poly(D,L-lactide-co-trimethylene carbonate) Fe3O4 (HSP-Fe3O4) constructed by Pickering emulsion 4D printing technique. The HSP-Fe3O4 demonstrates advantages multimodal...

A new algorithm for feature and contrast enhancement of mammographic images is proposed in this paper. The approach bases on multiscale transform mathematical morphology. First all, the Laplacian Gaussian pyramid operator applied to mammography into different scale subband images. In addition, detail or high frequency subimages are equalized by limited adaptive histogram equalization (CLAHE) low-pass processed Finally, enhanced image reconstructed from coefficients modified at one more...

Abstract Background Fetal spinal magnetic resonance imaging (MRI) is a prenatal routine for proper assessment of fetus development, especially when suspected malformations occur while ultrasound fails to provide details. Limited by hardware, fetal spine MR images suffer from its low resolution. High-resolution can directly enhance readability and improve diagnosis accuracy. Image interpolation higher resolution required in clinical situations, many methods fail preserve edge structures. Edge...

As an emerging modality for whole breast imaging, ultrasound tomography (UST), has been adopted diagnostic purposes. Efficient segmentation of entire in UST images plays important role quantitative tissue analysis and cancer diagnosis, while major existing methods suffer from considerable time consumption intensive user interaction. This paper explores three-dimensional GrabCut (GC3D) isolation thirty reflection (B-mode) volumetric images. The algorithm can be conveniently initialized by...

We propose a more accurate and efficient reconstruction method used in testing large aspheric surfaces with annular subaperture interferometry. By the introduction of Zernike polynomials that are orthogonal over region, proposed here eliminates coupling problem earlier algorithm based on circle polynomials. Because complexity recurrence definition polynomials, general symbol representation computing program is established. The implementation for provided detail. performance evaluated some...

A new algorithm for feature and contrast enhancement of mammographic images is proposed in this paper. The approach based on multiscale transform mathematical morphology. First all, the Laplacian Gaussian pyramid applied to decompose mammography into different subband sub-images. In addition, detail or high frequency sub-images are equalized by limited adaptive histogram equalization (CLAHE) low processed Finally, enhanced image reconstructed from coefficients modified at one more levels...

Abstract Several tough and flexible fluorinated polyamide-imide films were prepared from trimellitic anhydride chloride 2,2′-bis(trifluoromethyl)benzidine through a facile one-pot in-situ silylation method. By incorporating side groups, the solubility of was greatly enhanced. Meanwhile, due to their linear chain configuration existence hydrogen bonding, revealed high tensile strength, modulus, glass transition temperature (Tg) most interestingly, very low coefficient thermal expansion (CTE)...

Image sharpness is one of the most determining factors for image readability and scene understanding. How to accurately quantify it a hot topic. This paper systematically validates previously proposed index full-reference assessment (edge preservation ratio, EPR). Based on Gaussian blurring images in LIVE, CSIQ, TID2008 TID2013 databases, we firstly evaluated EPR accuracy five edge detectors LIVE selected an optimal further analysis. Then nine state-of-the-art quality metrics are compared,...

Mammography is a major technique for early detection of breast cancer, typically through masses or calcifications. However, how to help radiologists efficiently recognize these lesions remains challenging problem. In this paper, we propose comprehensive deep learning based solutions respectively detect and segment calcifications in mammograms. To achieve the optimal mass performance, our method combines Faster R-CNN with Feature Pyramid Networks, Focal Loss, Non-Local Neural Networks. We...

Many interpolation methods have been developed for high visual quality, but fail preserving image structures. Edges carry heavy structural messages tasks. Importance of edge preservation imposes edge-directed (EDI) a center focus. How to measure edge-preserving ability has not mentioned. In this paper, two metrics are proposed the by ratio from accuracy and robustness. Performance four with traditional evaluated on groups standard images other six commonly-used metrics. Experimental results...

Chinese narcissus (Narcissus tazetta var. chinensis) was introduced to China 1300–1400 years ago, and has grown naturally in southeastern China. It is a popular traditional potted flower well-known flowering bulb cultivated worldwide with only two white-tepal triploid cultivars, ‘Jinzhan Yintai’ ‘Yulinglong’. Recently, mutant yellow tepals observed promptly became popular. To clarify the key pigment for color difference its molecular mechanism of accumulation, we conducted metabolite...

Diffractive telescope technology is an innovation solution in construction of large light-weight space telescope. However, the nondesign orders diffractive optical elements (DOEs) may affect imaging performance as stray light. To study light characteristics a telescope, prototype was developed and its analysis model established. The including ghost, point source transmittance, veiling glare index (VGI) were analyzed. During star test prototype, ghost images appeared around image exposure...

Edge preservation ratio (EPR) is a full-reference metric for objective image quality assessment (IQA). It under the assumption that key messages to human visual systems are mainly from structures, and these structures can be extracted by edge detection. EPR measure twofold: accuracy robustness, color map synthesized reveal structure changes before after processing. The feasibility superiority of have been validated via magnification noise reduction. Experimental results suggest that: (1) it...

Ultrasound tomography (UST) image segmentation is fundamental in breast density estimation, medicine response analysis, and anatomical change quantification. Existing methods are time consuming require massive manual interaction. To address these issues, an automatic algorithm based on GrabCut (AUGC) proposed this paper. The presented method designs automated initialization for incomplete labeling sped up with multicore parallel programming. verify performance, AUGC applied to segment...

We present a novel method to accurately measure the flatness of large flat. The can be seen as task-specific error correction laser tracker. Laser trackers are positioned at two specific measuring stations and coordinates same sample points. angular errors primary tracker compensated with constraint information provided by an additional Using this method, we 4.5 m measurement uncertainty is 1.4 µm in rms (root mean square).

Background Computed tomography (CT) and magnetic resonance imaging (MRI) are indicated for use in preoperative planning may complicate diagnosis place a burden on patients with lumbar disc herniation. Purpose To investigate the diagnostic potential of MRI-based synthetic CT conventional Material Methods After obtaining prior institutional review board approval, 19 who underwent were enrolled this prospective study. Synthetic images generated from MRI data using U-net. The two sets compared...

In response to the pressing need for robust disease diagnosis from gastrointestinal tract (GIT) endoscopic images, we proposed FLATer, a fast, lightweight, and highly accurate transformer-based model. FLATer consists of residual block, vision transformer module, spatial attention which concurrently focuses on local features global attention. It can leverage capabilities both convolutional neural networks (CNNs) transformers (ViT). We decomposed classification images into two subtasks: binary...

We present the measurement methods for off-axis parabolic mirrors using Coordinates Measurement Machine and Swing Arm Profilometer during grinding process. An mirror with 1200mm in parent focal length, 360mm displacement 410mm diameter is measured introduced this paper. The convergence curves of surface error are reported, which reduced from 204.7μm PV, 55.3μm rms to 1.1μm 0.3μm after 39 iterations grinding. Besides, able control length less than 0.1mm tolerance. interferometric result...