Search

The tropical cyclone genesis number (TCGN) in July-October (JASO) over the western North Pacific (WNP) exhibits a robust interannual variation. It shows a longitudinally tri-pole pattern with a high in the eastern WNP and South China Sea (SCS) and a low in the western WNP, which explain 42.2 and 23.4 % of total TCGN variance in the eastern WNP and SCS, respectively. The high-low-high pattern is similar to that derived from a TC genesis potential index (GPI). To understand the cause of the longitudinal distribution of the dominant interannual mode, we examine the contributions of environmental parameters associated with GPI. It is found that relative humidity and relative vorticity are important factors responsible for TC variability in the SCS, while vertical shear and relative vorticity are crucial in determining TC activity in eastern WNP. A simultaneous correlation analysis shows that the WNP TCGN in JASO is significantly negatively correlated (with a correlation coefficient of −0.5) with sea surface temperature anomalies (SSTA) in the tropical North Atlantic (TNA). The longitudinal distribution of TC genesis frequency regressed onto TNA SSTA resembles that regressed upon the WNP TCGN series. The spatial patterns of regressed environmental variables onto the SSTA over the TNA also resemble those onto TCGN in the WNP, that is, an increase of relative humidity in the SCS and a weakening of vertical shear in the eastern WNP are all associated with cold SSTA in the TNA. Further analyses show that the cold SSTA in the TNA induce a negative heating in situ. In response to this negative heating, a low (upper)-level anomalous aniti-cyclonic (cyclonic) flows appear over the subtropical North Atlantic and eastern North Pacific, and to east of the cold SSTA, anomalous low-level westerlies appear in the tropical Indian Ocean. Given pronounced mean westerlies in northern Indian Ocean in boreal summer, the anomalous westerly flows increase local surface wind speed and surface evaporation and cool the SST in situ. Cold SSTA in northern Indian Ocean further suppress local convection, inducing anomalous westerlies to its east, leading to enhanced cyclonic vorticity and low surface pressure over the WNP monsoon trough region. Idealized numerical experiments further confirm this Indian Ocean relaying effect, through which cold SSTA in the tropical Atlantic exert a remote impact to circulation in the WNP. [ABSTRACT FROM AUTHOR]

As a fundamental task for geographical information updating, 3D city modeling, and other critical applications, the automatic extraction of building footprints from high-resolution remote sensing images has been substantially explored and received increasing attention over recent years. Among different types of building extraction methods, the polygonal segmentation methods produce vector building polygons that are in a more realistic format compared with those obtained from pixel-wise semantic labeling and contour-based methods. However, existing polygonal building segmentation methods usually require a perfect segmentation map and a complex post-processing procedure to guarantee the polygonization quality, or produce inaccurate vertex prediction results that suffer from wrong vertex sequence, self-intersections, fixed vertex quantity, etc. In our previous work, we have proposed a method for polygonal building segmentation from remote sensing images that addresses the above limitations of existing methods. In this paper, we propose PolyCity, which further extends and improves our previous work in terms of the application scenario, methodology design, and experimental results. Our proposed PolyCity contains the following three components: (1) a pixel-wise multi-task network for learning the semantic and geometric information via three tasks, i.e., building segmentation, boundary prediction, and edge orientation prediction; (2) a simple but effective vertex selection module (VSM), which effectively bridges the gap between pixel-wise and graph-based models via transforming the segmentation map into valid polygon vertices; (3) a graph-based vertex refinement network (VRN) for automatically adjusting the coordinates of VSM-generated valid polygon vertices, producing the final building polygons with more precise vertices. Results on three large-scale building extraction datasets demonstrate that our proposed PolyCity generates vector building footprints with more accurate vertices, edges, shapes, etc., achieving significant vertex score improvements while maintaining high segmentation and boundary scores compared with the current state-of-the-art. The code of PolyCity will be released at https://github.com/liweijia/polycity. [ABSTRACT FROM AUTHOR]

Although many criteria were proposed to predict the glass forming ability (GFA) of alloys, the precise correlation between the minor addition of elements and the glass formation remains a challenge. To solve this problem, we propose a simple parameter α' derived from thermal expansion coefficient based on the details of the intrinsic interatomic potential. It has been found that α' has the capability of precisely predict the changes of GFA, indicating the important role of anharmonicity of atomic vibration in glass formation. This anharmonicity-related parameter α' relates to atomic mobility behavior, and could be well embodied by the competition between characteristic types of clusters in terms of the fragility concept upon the liquid-solid evolution. This work not only helps to uncover atomic and cluster origins of glass formation from interatomic potential, but also provides a novel route to precisely evaluate the effect of minor addition of elements on GFA. [Display omitted] [ABSTRACT FROM AUTHOR]

SUMMARY In the context of simulations of wave propagations in unbounded domain, absorbing boundary conditions are often used to truncate the simulation domain to a finite space. Perfectly matched layer (PML) has proven to be an excellent absorbing boundary conditions. However, as this technique was primarily designed for the first-order equation system, it cannot be applied to the second-order equation system directly. In this paper, based on a complex-coordinate stretching technique, we developed a novel, efficient auxiliary-differential equation form of the complex-frequency shifted-PML for the second-order equation system. This facilitates the use of complex-frequency shifted-PML in acoustic simulations based upon wave equations of second-order form. Compared with previous state-of-the-art methods, the proposed one has the advantage of simpler implementation. It is an unsplit-field scheme that can be extended to higher-order discretization schemes conveniently. Numerical results from both homogeneous and heterogeneous computational domains are provided to illustrate the validity of the method. Copyright © 2013 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

The eigenspace-based minimum variance (ESBMV) beamformer can provide good imaging resolution and contrast; however, the performance is achieved at the cost of high computational complexity. In adaptive array processing, the beamspace method is an efficient way to lower the computational complexity. In this paper, we combine the beamspace method with the ESBMV beamformer and propose a beamdomain ESBMV beamformer. We demonstrate the feasibility of introducing the beamspace into the ESBMV beamformer and propose an effective method of forming the transform matrix based on the spatial spectrum of the array signals. We also illustrate the performance of the proposed beamformer when resolving point scatterers and a cyst phantom with both simulated and experimental data. The results show that the proposed method can achieve performance comparable to the ESBMV beamformer within much shorter time. [ABSTRACT FROM AUTHOR]

• A 3D UNet-based model was proposed to conduct cross-modality synthesis of MR images. • Multi-scale and residual learning framework were developed to improve the effect. • Experimental results on three datasets demonstrate the promising performance. Multi-modality MRI has been increasingly used for disease diagnosis and prognosis evaluation. However, the acquisitions of multi-modal images always face some practical difficulties such as the increased cost and patient discomfort. In this paper, we proposed an improved 3D UNet model for MRI modality synthesis and use the synthetic modalities to improve the performance of the following image analysis tasks such as tumor segmentation and tumor grading. We developed our model with three improvements on the basic 3D UNet model. First, we constructed a coarse-to-fine resolution fusion network to improve the ability of structural texture preservation. Second, we designed a convolutional fusion-based residual learning framework to learn mapping function, which leads to more detail preservation as well as less artifacts in the background. Third, for many-to-one synthesis, to combine the input multi-modalities more effectively, we introduced a new network module to fuse the UNet output feature maps of each modality. One private and two public datasets including a total of 493 cases were used for synthesis validation. Our proposed method surpasses the best existing method about 2.0 dB and 0.8 dB in PSNRs on the two datasets, respectively. Furthermore, two separate radiomics applications were used to demonstrate the value of our modality synthesis method. In the image segmentation application, integrating the modalities synthesized by our method has improved the segmentation Dice by 1.3%. In gliomas grading diagnosis application, integrating the synthesized modalities improved the diagnostic accuracy by 8.3%. [ABSTRACT FROM AUTHOR]

Tracking pylorus in ultrasonic image sequences is an important step in the analysis of duodenogastric reflux (DGR). We propose a joint prediction and segmentation method (JPS) which combines optical flow with active contour to track pylorus. The goal of the proposed method is to improve the pyloric tracking accuracy by taking account of not only the connection information among edge points but also the spatio-temporal information among consecutive frames. The proposed method is compared with other four tracking methods by using both synthetic and real ultrasonic image sequences. Several numerical indexes: Hausdorff distance (HD), average distance (AD), mean edge distance (MED), and edge curvature (EC) have been calculated to evaluate the performance of each method. JPS achieves the minimum distance metrics (HD, AD, and MED) and a smaller EC. The experimental results indicate that JPS gives a better tracking performance than others by the best agreement with the gold curves while keeping the smoothness of the result. [ABSTRACT FROM AUTHOR]

Objective: Compared with normal birth weight fetuses (2,500-4,000 g), accurate fetal weight estimation for fetuses with low or excessive weight is considered more important for fetal compromise prediction and labor management. New formulas were developed to estimate weight for low birth weight (LBW) fetuses and macrosomic fetuses, respectively. Methods: A total of 523 fetuses with birth weight less than 2,500 g, 652 fetuses with normal birth weight, and 239 fetuses with birth weight more than 4,000 g were included in the study. As much as 25 existing formulas which incorporate regularly defined fetal measurements were evaluated and compared. Performance evaluation of existing formulas showed that no formulas can provide consistently accurate weight estimation both for LBW fetuses and macrosomic fetuses. A total of 1,034 cases were utilized to generate an overall regression formula. If the pre-estimated weight fell into the suspicious LBW and macrosomia range, the value was then updated by using the new regression formulas for LBW fetuses and macrosomic fetuses. As the training group, 262 LBW fetuses and 120 macrosomic fetuses were employed in a stepwise linear regression to obtain two update regression formulas for suspicious LBW fetuses and macrosomic fetuses. As the validation group, another 261 LBW fetuses and 119 macrosomic fetuses were assessed. Results: The new overall formula has the form of Log(10)BW = 0.180(HC) + 0.00628(AC) - 0.00318(HC)(2) + 0.00173(AC)(FL) + 0.0000430(BPD)(HC)(2). The update formula for suspicious LBW fetuses is LnBW = 1.470(BPD) + 0.0169(HC) - 0.0873(BPD)(2) + 0.00518(AC)(FL) and for macrosomic fetuses is Log(10)BW = 0.730(BPD) - 0.0375(BPD)(2) + 0.000264(AC)(FL). For LBW fetuses, the new method gave 7.6 ± 209.0 (g) of estimation error and 8.3 ± 7.8 (%) of absolute percentage error, while the best existing formula provided -0.7 ± 226.0 (g) and 9.1 ± 8.3 (%). With the new method, 71.3% of estimates fell within ±10% of the actual birth weight, while the best existing formula gave 65.5%. For macrosomic fetuses, the new method gave -87.9 ± 231.0 (g) of estimation error and 4.4 ± 3.9 (%) of absolute percentage error, while the best existing formula provided 115.6 ± 345.1 (g) and 6.8 ± 5.4 (%). With the new method, 89.1% of estimates fell within ±10% of the actual birth weight, while the best existing formula gave 75.6%. Conclusions: To improve the weight estimation accuracy for low or excessive weight fetuses, separate formulas are necessary. The new method provides significant improvement on fetal weight estimation for LBW fetuses and macrosomic fetuses. [ABSTRACT FROM AUTHOR]