Jitter considerations in high bit rate digital video signals

While digital signals tend to be more immune to noise as measured in analog systems, other noise related phenomena such as bit errors can cause problems which are equally as undesirable. This paper addresses one such noise artifact which is jitter in a digital bit stream. While the typical method of measuring jitter is with a sampling oscilloscope, a new technique using modulation domain analyzers was used. This method allows direct measurement of the spectral content of jitter. This technique has been used in the telecommunications industry and examples of their specifications are provided     

Enterococcal bacteremia in a medical center

BACKGROUND: The enterococci have become important nosocomial pathogens. They can cause multiple site infections and enterococcal bacteremia becomes more frequently associated with a high mortality rate. Previous studies of enterococcal bacteremia showed a variety of results. To establish the significance and importance of enterococci as nosocomial pathogens in this hospital, to characterize their clinical pictures and to search for the risk factors for mortality, this retrospective study was performed. METHODS: There were 208 cases of enterococcal bacteremia which occurred from 1988 to 1992. Twenty-seven cases had no medical charts, dismissing possibility of evaluation. Finally, 181 cases of enterococcal bacteremia were analysed. RESULTS: One hundred and eighteen episodes were nosocomial infections. Polymicrobial bacteremia occurred in 68.5% of the patients and the most common co-isolate was Pseudomonas aeruginosa. Those patients (78.5%) with underlying diseases and malignancies were the most common underlying problems. The portal of entry could be found in 69.6 percent of patients, with the gastrointestinal tract the most common sources. Antimicrobial susceptibility testing showed high gentamicin resistance rate (89.5%), and ampicillin still had about 80 percent sensitivity rate. The group who received specific antibiotic therapy for enterococcus showed lower mortality (36.4% versus 47.6%). Only one case had infective endocarditis. Forty-nine patients suffered from septic shock, the cause of 30 deaths. Totally 75 patients died during hospitalization. Besides sepsis, another major cause of death was their underlying diseases itself. CONCLUSIONS: Enterococci have no doubt become important nosocomial pathogens and enterococcal bacteremia were associated with high mortality, especially in elderly patients with underlying diseases such as malignancy or diabetes. When clinically dealing with sepsis from the gastrointestinal or biliary tract, especially when previous cephalosporins therapy showed no response, the possibility of enterococcal bacteremia should always be considered.     

Formation and growth mechanism of TiC crystal in TiC_p/Ti composites

1　ＩＮＴＲＯＤＵＣＴＩＯＮＰａｒｔｉｃｌｅｓｒｅｉｎｆｏｒｃｅｄｔｉｔａｎｉｕｍａｌｌｏｙｍａｔｒｉｘｃｏｍｐｏｓｉｔｅｓ(ＴＭＣｐ)ｈａｖｅｗｉｄｅａｐｐｌｉｃａｔｉｏｎｆｉｅｌｄｓｆｏｒｔｈｅｉｒａｔｔｒａｃｔｉｖｅｈｉｇｈｓｐｅｃｉｆｉｃｓｔｒｅｎｇ     

Dynamic endocrine testing. The Mayo Clinic model

An endocrine testing center (ETC) is a universal requirement for the practice of endocrinology. Modifications of the Mayo Clinic model for an ETC should be applicable to most endocrine practices. Key components of an ETC include a centralized testing area, registered nurse-physician team, detailed testing protocols, and patient education programs.     

Activity-dependent changes in voltage-dependent calcium currents and transmitter release

The use of capillary electrophoresis (CE) for clinically relevant assays is attractive since it often presents many advantages over contemporary methods. The small-diameter tubing that holds the separation medium has led to the development of multicapillary instruments, and simultaneous sample analysis. Furthermore, CE is compatible with a wide range of detectors, including UV-Vis, fluorescence, laser-induced fluorescence, electrochemistry, mass spectrometry, radiometric, and more recently nuclear magnetic resonance, and laser-induced circular dichroism systems. Selection of an appropriate detector can yield highly specific analyte detection with good mass sensitivity. Another attractive feature of CE is the low consumption of sample and reagents. However, it is paradoxical that this advantage also leads to severe limitation, namely poor concentration sensitivity. Often high analyte concentrations are required in order to have injection of sufficient material for detection. In this regard, a series of devices that are broadly termed 'analyte concentrators' have been developed for analyte preconcentration on-line with the CE capillary. These devices have been used primarily for non-specific analyte preconcentration using packing material of the C18 type. Alternatively, the use of very specific antibody-containing cartridges and enzyme-immobilized microreactors have been demonstrated. In the current report, we review the likely impact of the technology of capillary electrophoresis and the role of the CE analyte concentrator-microreactor on the analysis of biomolecules, present on complex matrices, in a clinical laboratory. Specific examples of the direct analysis of physiologically-derived fluids and microdialysates are presented, and a personal view of the future of CE in the clinical environment is given.     

一类非线性MIMO不确定系统的动态输出反馈镇定

考察其标称系统的相对阶大于{1,1,…,1}同时含匹配和非匹配不确定性的MIMO 非线性系统的动态输出反馈镇定问题.文中直接用Lyaunov方法构造一类输出反馈动态补 偿器,该补偿器可以实现对所论非线性不确定系统的动态输出反馈渐近镇定.     

一种多晶硅高温压力传感器

制作压力传感器时，在二氧化硅层上淀积多晶硅膜，既可利用优良的机械特性，又可保证压敏电阻与衬底间具有良好的绝缘性，由此可大大提高器件的温度特性。介绍了一种多晶硅压力传感器的原理和设计。实验结果表明，这类传感器具有灵敏度好，精度高等特点，电路工作范围为０－２５０℃，且具有良好的温度稳定性。     

冶金过程中复杂夹杂物和钢液化学平衡计算

本文提出一个同温冶炼过程中预测非金属夹杂物沉淀的计算机程序。本研究中，多元硅酸盐溶体的Ｇｉｂｂｓ自由能由基于统计热力学的晶胞模型来计算，该晶胞模型按照氧与周围的阳离子构成的对称和非对称晶胞来描述硅酸盐熔体结构，假定高温冶炼时认与非金属夹杂物平衡，钢液中氧化物元素的活度由Ｗａｇｎｅｒ交互作用系数来计算，利用本程序可以估计炼钢时沉淀的非金属氧化物的成分，为调整冶炼工艺和提高产品质量提供指导，进一步可计     

Learning concepts in parallel based upon the strategy of versionspace

Applies the technique of parallel processing to concept learning. A parallel version-space learning algorithm based upon the principle of divide-and-conquer is proposed. Its time complexity is analyzed to be O(k log₂n) with n processors, where n is the number of given training instances and k is a coefficient depending on the application domains. For a bounded number of processors in real situations, a modified parallel learning algorithm is then proposed. Experimental results are then performed on a real learning problem, showing that our parallel learning algorithm works, and being quite consistent with the results of theoretical analysis. We conclude that when the number of training instances is large, it is worth learning in parallel because of its faster execution