3D MERGE与3D SPACE STIR序列在腰椎间盘突出症检查中的应用比较

目的:对比三维多回波恢复梯度回波(3D MERGE)、三维可变反转角快速自旋回波(3D SPACE STIR)序列在腰椎间盘突出症(LDH)检查中的应用效果。方法:选择2020年1月~2022年11月收治的135例LDH患者,回顾性分析患者临床和磁共振成像(MRI)资料,所有患者均接受常规MRI扫描及3D MERGE、3D SPACE STIR序列扫描,对比3D MERGE、3D SPACE STIR序列测量神经根直径的一致性,评价两种序列的图像质量参数[信噪比(SNR)、对比噪声比(CNR)]、图像清晰度评分。结果:3D MERGE和3D SPACE STIR序列测量的L3~S1神经根直径比较差异无统计学意义(P>0.05),且两组序列测量的L3、L4、L5和S1直径均显示出较高相关性(r=0.957,0.986,0.975,0.972,P<0.05);3D MERGE序列的SNR及CNR均高于3D SPACE STIR序列,神经根显示分级、图像清晰度评分优于3D SPACE STIR序列,差异有统计学意义(P<0.05)。结论:3D MERGE、3D SPACE STIR序列在LDH神经根直径测量中具有极高一致性,3D MERGE序列较3D SPACE STIR序列能够更清晰显示神经跟的解剖形态,图像质量更好。

中医药多元证据整合方法学MERGE框架

在循证医学时代下,依托规范的技术方法和标准化的操作规程发掘中医药独特优势,是实现中医药现代化、国际化发展并惠泽人类的必由之路。中医理论、人用经验和研究证据三结合证据体系的提出标志着中医药特色评价体系思维方法取得了重要进步,经过恰当方法整合后的多元证据体是中医药临床指南推荐意见和循证卫生决策的有力支撑。本文基于当前国际证据合成与分级方法学前沿进展,初步提出中医药多元证据整合的方法学框架——MERGE(Merge Evidence-based Research and artificial intelliGence to support smart dEcision)框架,以期为中医药循证医学方法学体系的完善和发展提供借鉴和参考。

基于VVC的多方向细化Merge模式优化算法

针对多功能视频编码标准VVC帧间预测中,Merge模式未充分考虑运动存在的多种方向性从而降低了预测精度的问题,提出了一种基于VVC的多方向细化Merge模式优化算法。本算法在分析带运动矢量差的Merge技术的基础上,首先调整了步长选择范围并增加多种搜索方向,然后根据步长自适应选择色度块的运动补偿方式,并最终根据率失真代价准则选择最优的运动矢量信息。实验结果表明,本文算法与VTM-12.0参考模型相比,在低延迟P帧配置下Y、U、V三分量的BD-rate分别平均下降了0.57%、0.62%、0.25%,在随机接入配置下的BD-rate中Y分量下降了0.27%、U分量仅提高了0.11%、V分量降低了0.04%,有效地提升了Merge模式的编码性能。

Validation of the multi-satellite merged sea surface salinity in the South China Sea

Sea surface salinity(SSS)is an essential variable of ocean dynamics and climate research.The Soil Moisture and Ocean Salinity(SMOS),Aquarius,and Soil Moisture Active Passive(SMAP)satellite missions all provide SSS measurements.The European Space Agency(ESA)Climate Change Initiative Sea Surface Salinity(CCI-SSS)project merged these three satellite SSS data to produce CCI L4SSS products.We validated the accuracy of the four satellite products(CCI,SMOS,Aquarius,and SMAP)using in-situ gridded data and Argo floats in the South China Sea(SCS).Compared with in-situ gridded data,it shows that the CCI achieved the best performance(RMSD:0.365)on monthly time scales.The RMSD of SMOS,Aquarius,and SMAP(SMOS:0.389;Aquarius:0.409;SMAP:0.391)are close,and the SMOS takes a slight advantage in contrast with Aquarius and SMAP.Large discrepancies can be found near the coastline and in the shelf seas.Meanwhile,CCI with lower RMSD(0.295)perform better than single satellite data(SMOS:0.517;SMAP:0.297)on weekly time scales compared with Argo floats.Overall,the merged CCI have the smallest RMSD among the four satellite products in the SCS on both weekly time scales and monthly time scales,which illustrates the improved accuracy of merged CCI compared with the individual satellite data.

Novel layout design of 4H-SiC merged PiN Schottky diodes leading to improved surge robustness

A method to improve the surge current capability of silicon carbide(SiC)merged PiN Schottky(MPS)diodes is presented and investigated via three-dimensional electro-thermal simulations.When compared with a conventional MPS diode,the proposed structure has a more uniform current distribution during bipolar conduction due to the help of the continuous P+surface,which can avoid the formation of local hotspots during the surge process.The Silvaco simulation results show that the proposed structure has a 20.29%higher surge capability and a 15.06%higher surge energy compared with a conventional MPS diode.The bipolar on-state voltage of the proposed structure is 4.69 V,which is 56.29%lower than that of a conventional MPS diode,enabling the device to enter the bipolar mode earlier during the surge process.Furthermore,the proposed structure can suppress the occurrence of‘snapback'phenomena when switching from the unipolar to the bipolar operation mode.In addition,an analysis of the surge process of MPS diodes is carried out in detail.

一种改进的Split-Merge图像分割算法

图像分割的经典算法Split-Merge算法思路简洁、运算效率高,但存在边界定位不准确和过分割等问题。针对过分割问题提出了一种改进的Split-Merge算法,在合并过程中引入全局目标函数,并在目标函数中定义了长度项,使用贪心法完成合并过程。长度项以及贪心法的使用促使相邻区域充分合并,有效地克服了过分割等问题,取得了较好的分割效果。

高效视频编码的Merge模式候选决策研究与改进

高效视频编码的Merge模式基于四叉树的图像编码方式,可以将多个相邻且具有相似运动情况的预测单元(PU)块合并成同一个区域并使用相同的运动参数。研究候选PU的合并决策过程,针对Merge模式所带来的高计算复杂度且合并候选过程中常进行多次不必要的判断问题,提出一种低复杂度的候选决策算法。该算法基于时空域候选在不同情形下的分布特征,从改变候选列表选择过程的角度减少候选判断次数从而降低编码复杂度。实验结果表明,与原始算法相比,该算法的编码时间缩短、码率减少,能有效降低编码复杂度。

在Split-Merge框架下进行基于图的图像分割

Split-Merge算法是图像分割领域的经典算法,思路简洁、运算效率高,但因分裂位置和形状固定,导致出现方块效应和边界定位不准确等问题。现有的各种改进算法不能从根本上解决这些问题,为此本文将基于图的分割算法纳入到Split-Merge框架中来,提出了在Split-Merge框架下进行基于图的分割。新算法把传统的基于图割值的算法作为分裂阶段的处理手段,在合并阶段采用本文提出的一种与访问次序无关的合并算法。新算法既解决了Split-Merge算法中因分裂位置和形状固定而带来的各种问题,也有效地消除了在基于图的分割中常见的"过分割"现象,取得了满意的分割效果。

基于Excel数据源的Mail Merge在教学管理中的应用

教育教学管理过程中常常涉及到追考证、成绩单、录取通知书等教育对象数据的通知工作,利用常规办法将成绩统计表中数据逐一复制的方式不仅繁琐且容易出错。基于EXCEL数据源的Mail Merge(Word邮件合并)使用可以迅速、准确地完成数据处理、个案分析及个人成绩通知。

基于域的Mail Merge在文秘工作中的应用探究

域是文本处理软件内置的一个文档部件,应用广泛,功能强大。在文档合并、数据更新等方面有独到的用途。文秘工作经常涉及大量不同类型文档信息的合并处理,应用域及Mail Merge功能,可以快速高效地将电子表格中相应的记录单数据合并进文本文档,一键生成多个个性化文档。

Regional Merge K-means图像分割算法及其质量评价

K值是基于K-means聚类算法进行图像分割效果的关键因素,为了解决K值过小,图像分割不明显,K值过大,图像分割信息过于碎片化的问题,提出一种Regional Merge K-means(RMK)算法.首先对图像运用K-means算法进行聚类,然后运用区域合并(Regional Merge)优化图像分割效果,最后利用不同图像进行图像分割试验,并选用不同的质量评价指标对试验结果进行评价.结果表明,该算法能显著优化图像分割效果.

基于学习模型的3D-HEVC提前Merge模式终止算法

作为高效视频编码(HEVC)的扩展,3D-HEVC 标准有效地提高了 3D 视频的压缩效率,但是也带来了很高的编码计算复杂度。为了显著地降低 3D-HEVC 编码复杂度,提出了一种提前 Merge 模式终止算法。首先,提取 Merge 模式编码后的残差信号作为特征信息;然后,根据当前编码帧内已经编码的编码单元(CU)的最优Merge 模式残差信号建立学习模型;最后,提取当前 CU 的 Merge 模式的残差信号,并且利用学习模型预测 Merge模式是否为最优模式。实验结果表明,提出的提前 Merge 模式终止算法分别将 3D-HEVC 纹理视点和深度图编码的时间降低了 41.9%和 24.3%,且编码性能的降低几乎可忽略不计。相较于现有的提前 Merge 模式算法,提出的提前 Merge 模式终止算法能进一步降低 3D-HEVC 的编码时间,并且设计简单,易于集成到 3D-HEVC 测试模型。

标准化3D MERGE序列在腰骶丛神经成像中的应用研究

目的:研究和分析在腰骶丛神经成像中标准化3D MERGE序列的应用价值。方法:随机选取33例腰骶椎病变患者,为患者实施常规MRI腰骶椎扫描,加扫STIR序列及3D MERGE序列,对比两种序列腰骶丛神经显示情况及图像信噪比、对比噪声比。结果:3D MERGE序列Ⅰ级腰骶丛神经显示率高于STIR序列,Ⅲ级显示率低于STIR序列,差异有统计学意义(P<0.05),两种序列Ⅱ级腰骶丛神经显示率差异不存在统计学意义(P>0.05)。结论:标准化3D MERGE序列能够使患者腰骶丛神经解剖结构等获得清晰显示,值得临床推荐。

A wave energy resource assessment in the China's seas based on multi-satellite merged radar altimeter data

Wave energy resources are abundant in both offshore and nearshore areas of the China＇s seas. A reliable assessment of the wave energy resources must be performed before they can be exploited. First, for a water depth in offshore waters of China, a parameterized wave power density model that considers the effects of the water depth is introduced to improve the calculating accuracy of the wave power density. Second, wave heights and wind speeds on the surface of the China＇s seas are retrieved from an AVISO multi-satellite altim-eter data set for the period from 2009 to 2013. Three mean wave period inversion models are developed and used to calculate the wave energy period. Third, a practical application value for developing the wave energy is analyzed based on buoy data. Finally, the wave power density is then calculated using the wave field data. Using the distribution of wave power density, the energy level frequency, the time variability indexes, the to-tal wave energy and the distribution of total wave energy density according to a wave state, the offshore wave energy in the China＇s seas is assessed. The results show that the areas of abundant and stable wave energy are primarily located in the north-central part of the South China Sea, the Luzon Strait, southeast of Taiwan in the China＇s seas; the wave power density values in these areas are approximately 14.0–18.5 kW/m. The wave energy in the China’s seas presents obvious seasonal variations and optimal seasons for a wave energy utilization are in winter and autumn. Except for very coastal waters, in other sea areas in the China＇s seas, the energy is primarily from the wave state with 0.5 m≤Hs≤4 m, 4 s≤Te≤10 s whereHs is a significant wave height andTe is an energy period; within this wave state, the wave energy accounts for 80% above of the total wave energy. This characteristic is advantageous to designing wave energy convertors （WECs）. The practical application value of the wave energy is higher which can be as an effective supplement for an energy con-sumption in some areas. The above results are consistent with the wave model which indicates fully that this new microwave remote sensing method altimeter is effective and feasible for the wave energy assessment.

数据集成增量检测中Sort Merge算法的改进和实现

在数据集成应用中,增量检测一个最主要的职责就是检测计算出数据源的增量数据,有多种方式可以检测分析出数据源的变化数据:触发器、日志分析、快照差分等等。而快照差分又可以采用不同的算法对两个快照进行计算,针对差分算法中SortMerge算法的一个改进和实现,主要利用有序文件来提高增量计算的效率。由于快照差分在字符串比对、文件差异计算、大数据量的差异对比、自然连接等实际问题有应用,对此类问题都有一定借鉴意义。

A new merged dataset of global ocean chlorophyll a concentration with higher spatial and temporal coverage

Understanding the ocean＇s role in the global carbon cycle and its response to environmental change requires a high spatio-temporal resolution of observation.Merging ocean color data from multiple sources is an effective way to alleviate the limitation of individual ocean color sensors（e.g.,swath width and gaps,cloudy or rainy weather,and sun glint） and to improve the temporal and spatial coverage.Since the missions of Sea-Viewing Wide Field-of-View Sensor（Sea Wi FS） and Medium-spectral Resolution Imaging Spectrometer（MERIS） ended on December 11,2010 and May 9,2012,respectively,the number of available ocean color sensors has declined,reducing the benefits of the merged ocean color data with respect to the spatial and temporal coverage.In present work,Medium Resolution Spectral Imager（MERSI）/FY-3 of China is added in merged processing and a new dataset of global ocean chlorophyll a（Chl a） concentration（2000–2015） is generated from the remote sensing reflectance（Rrs（λ）） observations of MERIS,Moderate-resolution imaging spectra-radiometer（MODIS）-AQUA,Visible infrared Imaging Radiometer（VIIRS） and MERSI.These data resources are first merged into unified remote sensing reflectance data,and then Chl a concentration data are inversed using the combined Chl a algorithm of color index-based algorithm（CIA） and OC3.The merged data products show major improvements in spatial and temporal coverage from the addition of MERSI.The average daily coverage of merged products is approximately 24% of the global ocean and increases by approximately 9% when MERSI data are added in the merging process.Sampling frequency（temporal coverage） is greatly improved by combining MERSI data,with the median sampling frequency increasing from 15.6%（57 d/a） to 29.9%（109 d/a）.The merged Chl a products herein were validated by in situ measurements and comparing them with the merged products using the same approach except for omitting MERSI and Glob Colour and MEa SUREs merged data.Correlation and relative error between the new merged Chl a products and in situ observation are stable relative to the results of the merged products without the addition of MERSI.Time series of the Chl a concentration anomalies are similar to the merged products without adding MERSI and single sensors.The new merged products agree within approximately 10% of the merged Chl a product from Glob Colour and MEa SUREs.

Simulation study of a mixed terminal structure for 4H-SiC merged PiN/Schottky diode

In this paper, a mixed terminal structure for the 4H-SiC merged PiN/Schottky diode （MPS） is investigated, which is a combination of a field plate, a junction termination extension and floating limiting rings. Optimization is performed on the terminal structure by using the ISE-TCAD. Further analysis shows that this structure can greatly reduce the sensitivity of the breakdown voltage to the doping concentration and can effectively suppress the effect of the interface charge compared with the structure of the junction termination extension. At the same time, the 4H-SiC MPS with this termination structure can reach a high and stable breakdown voltage.

A 4H-SiC merged P–I–N Schottky with floating back-to-back diode

A novel 4 H-Si C merged P–I–N Schottky(MPS)with floating back-to-back diode(FBD),named FBD-MPS,is proposed and investigated by the Sentaurus technology computer-aided design(TCAD)and analytical model.The FBD features a trench oxide and floating P-shield,which is inserted between the P+/N-(PN)junction and Schottky junction to eliminate the shorted anode effect.The FBD is formed by the N-drift/P-shield/N-drift and it separates the PN and Schottky active region independently.The FBD reduces not only the Vturn to suppress the snapback effect but also the Von at bipolar operation.The results show that the snapback can be completely eliminated,and the maximum electric field(Emax)is shifted from the Schottky junction to the FBD in the breakdown state.

面向硬件实现的HEVC中Merge的高效编码方案

HEVC帧间预测中,Merge模式是对相邻已编码的PU进行合并处理,其需要重复访问数据并进行大量的复杂计算,是影响硬件实现效率的模块之一。针对该问题,本文通过对帧间预测技术的研究,提出了一种能实现高清视频实时编码的多级流水并行硬件架构设计,其使用多级流水线计算Merge模块,达到了较高的数据吞吐率;合理规划同个CTU不同深度的CU扫描顺序,实现了CTU内流水线不断流的高效编码设计;整个帧间的插值模块和SATD运算模块合理复用,提高了硬件资源利用率。

HEVC中Merge模式的算法优化和硬件实现

高效视频编码HEVC是最新的视频编码标准,比之前的如H.264/AVC有更高的压缩比,Merge模式是HEVC为了提高帧间编码效率新改进的技术之一。Merge过程是一个选择候选运动矢量的过程,需要大量复杂的计算和内存访问,导致Merge硬件实现效率低下。本文首先从算法上用2N×2N预测单元PU的Merge模式实现其他划分类型的PU Merge模式,在性能损失不多的情况下,便于硬件实现。其次,设计硬件时,多层次采用流水线结构设计,提高了Merge硬件模块的吞吐量,在插值模块和SATD运算模块都重复高效地利用了16×8的最小缓存空间,提高了硬件资源利用率,降低了硬件开支。RTL代码实现并DC综合结果表明,与现有文献提出的架构相比,本文架构在资源利用上有较大的优势,逻辑门数为279k,主频可达到980MHz,每秒可以处理319帧分辨率为1920×1080的视频。