超声左心耳图像的轮廓自动提取

针对经食道超声左心耳图像的分辨率低、对比度低、含有斑点噪声等问题,提出了一种结合左心耳解剖位置和超声图像灰度及相位信息的方法,全自动定位常规切片中的左心耳。首先,根据医生采集习惯,以左心耳在标准切面中的解剖位置为先验知识,结合其灰度特性,自动确定分割模型中的初始轮廓;然后,通过线型加权相位和梯度信息构造新的外力项,改进向量场卷积模型,完成左心耳轮廓的自动提取。300张左心耳超声图片测试结果表明,以医生手动勾勒的轮廓作为“金标准”,该方法自动提取左心耳的准确性为0.8969 0.0494、敏感性为0.9058 0.0762、特异性为0.9645 0.1687。分割效果优于传统的向量场卷积模型,能够解决自动定位超声图像中左心耳的初始轮廓和弱边界分割的问题。     

基于ＡＤＳＬ接入的核心网体系结构

ＡＤＳＬ技术是基于铜双绞线的一种重要的宽带接入方式,在讨论了一组能够支持基于ＰＰＰ／ＡＴＭ／ＡＤＳＬ的接入网体系结的核心网络 体系结构后,分析了几种不同的组网方式的特征与协议栈模型。     

基于深度学习的超声图像左心耳自动分割方法

从超声图像中分割出左心耳（LAA）是得出临床诊断指标的重要步骤,而准确自动分割的首要步骤和难点就是实现目标的自动定位。针对这一问题,提出了一种结合基于深度学习框架的自动定位和基于模型的分割算法的方法来实现超声图像中LAA的自动分割。首先,训练YOLO模型作为LAA自动定位的网络架构;其次,通过验证集确定最优的权重文件,并预测出LAA的最小包围盒;最后,在正确定位的基础上,将YOLO预测的最小包围盒放大1.5倍作为初始轮廓,利用C-V模型完成LAA的自动分割。分割结果用5项指标加以评价：正确性、敏感性、特异性、阴性、阳性。实验结果表明,所提方法能够实现不同分辨率条件和不同显示模式下LAA的自动定位,小样本数据在1000次迭代时已经达到最优的定位效果,正确定位率达到72.25%,并且在正确定位的基础上,C-V模型的分割准确率能够达到98.09%。因此,深度学习技术在实现LAA超声图像的自动分割上具备较大的潜力,能够为基于轮廓的分割算法提供良好的初始轮廓。     

Impact of limited drying on Momordica cochinchinensis Spreng. aril carotenoids content and antioxidant activity

Food product based on gac fruit (Momordica cochinchinensis Spreng.) arils have a high potential due to the high carotenoids content of this fruit. Drying is a key preparation step for carotenoids extraction from gac fruit in a economically viable process. The impact of different drying technics, temperature, final product moisture content on the carotenoid content, hydrophilic and lipophilic antioxidant activity (evaluated with three methods) and color of the gac arils is discussed based on laboratory scale experimental tests. The results highlight an optimal temperature between 50 °C and 60 °C to conserve the color, the carotenoid content and the antioxidant activity. Also, these properties are better preserved by limiting the drying to dry based moisture content between 15% and 18% while the advantages of drying for further processing and for refrigerated conservation for a few months are achieved.     

LTE-U关键技术与网络规划策略

LTE-U是一种在非授权频谱上部署LTE技术以提供电信级移动通信服务的无线接入技术.本文首先研究了LTE-U关键技术,然后从频率使用、载波聚合方案、网络部署策略、干扰规避等方面分析了LTE-U引入对网络规划的影响,针对性的提出了以提升中国移动LTE网络质量与建设效率为目标的LTE-U网络规划策略,为中国移动未来LTE-U部署提供参考.     

Feasibility study of using montmorillonite for stability enhancement of l-ascorbic acid

This study tended to construct new l-ascorbic acid (LAA) composites in low toxicity and high stability for feasible application. LAA is chemically very unstable, since it is easily oxidized into biologically inactive compounds naturally. Our finding showed that introduction of montmorillonite (MMT) could significantly attenuate its toxicity and to sustain the stability of LAA with economic feasibility for practical uses. In addition, as phosphoric acid was biologically compatible, it was used for the pretreatment of MMT to obtain a promising stabilization of LAA. Toxicity assessment also showed that MMT treated with low-concentration acids should be considered as biologically safe according to our assessment. Thus, using acid treated MMT to stabilize LAA in a long-term might be technically feasible for further uses.     

非授权频段长期演进系统中的混合动态分簇算法

针对非授权频段长期演进（LTE）系统中动态子帧配置引起的交叉子帧干扰问题,提出了一种综合考虑大尺度损耗及小区业务量情况的混合动态分簇算法。首先,通过基站端对大尺度损耗及小区业务量情况的周期性测量,计算出对应的相关度值;然后,根据相关度值对小区进行轮询式分簇,实现小区分簇结果的周期性更新;最后,根据更新后的小区分簇结果执行动态子帧配置。仿真实验中,相比传统的静态分簇算法,中业务到达率条件下混合动态分簇算法的用户上下行平均吞吐量分别提升了约16.92%和34.33%;用户上下行平均时延分别降低了约14.18%和36.32%。仿真结果表明,混合动态分簇算法可以有效减小交叉子帧干扰的影响,提升系统吞吐量,性能优于传统的静态分簇算法。     

Mutual coupling compensation of linear antenna array using evolutionary optimization technique

This paper presents an optimization-based approach to compensate for the mutual coupling effect and to reduce the sidelobe level (SLL) of the normalized radiation pattern by optimizing the current excitation amplitude of the antenna array elements. Due to the mutual coupling effect, the SLL of the radiation pattern is increased, and thereby, the field pattern of the antenna array is severely degraded. This causes interference with other communication systems working at the same frequency. Toward the compensation of the mutual coupling effect and reduction of SLL, the Cat swarm optimization (CSO) algorithm is employed, and the excitation amplitudes of the antenna elements are optimized. In this regard, optimizing the cost function is defined by introducing the impedance factor (IF), calculated by using the self-impedance and mutual impedance of the antenna elements. The proposed method for the synthesis of a mutually coupled linear antenna array is less expensive, simpler to use, and more effective. Array sets of 4, 6, 8, and 10 elements are considered for optimization.     

测井属性在砂砾岩地层评价中应用初探

砂砾岩地层评价中岩性识别、层序划分以及地层对比是亟待解决的难题。在测井曲线中提取有效的地层信息是解决此问题的有效途径。在总结前人研究工作基础上,提出了测井属性的概念,并初步进行了属性分类,目前主要分为曲线多元参数统计属性、小波分析属性、频谱分析属性和积分属性4类。辅以典型实例分析,取得了明显的地质效果,提升了测井地质应用能力:多元参数统计属性能够有效地识别砂砾岩地层的复杂岩性;频谱分析属性对信号突变点具有敏感性,可以反映地层界面;根据小波的多尺度属性实现了由粗到细逐级地划分层序界面;依据测井曲线的积分属性在砂砾岩地区地层对比工作中取得了良好的效果。     

非授权频段中LTE和Wi-Fi共存系统性能研究

为了应对移动数据日益增长的需求,3GPP(3rd Generation Partnership Project)对LTE(Long Term Evolution)扩展到非授权频段展开了相关讨论。LAA(Licensed-Assisted Access),亦称为LTE-U(Unlicensed LTE),与目前已经运行在5GHz非授权频段上的其他无线接入技术(如Wi-Fi)的共存问题,成为了研究的焦点。本文主要研究了在非授权频段上运行LTE和Wi-Fi系统共存的机制,设计实现了基于LAA和Wi-Fi共存场景的系统级仿真平台。仿真结果表明,在非授权频段上LAA能与Wi-Fi和谐共存。针对LAA系统设计一种合理的退避机制,能够同时提升LAA和Wi-Fi系统的吞吐量,提高资源分配的公平性和使用效率。