一种基于多门限二进制编码的多光谱图像分类新方法

不同类型的地物具有不同的反射光谱 ,在多维光谱空间中构成不同的特征向量 ,这便是遥感多光谱图象分析与识别的物理依据。传统方法中有基于单个象元波段亮度的、空间纹理的、变换空间的多光谱图象特征提取与分析方法 ,但这些方法并没有直接描述地物的最本质特征———反射光谱曲线 ,从上个世纪 80年代开始 ,当二进制编码的方法提出并在多光谱匹配识别中获得成功应用后 ,多光谱图象分析处理便可以在这种特征提取的基础上 ,研究新的方法。本文提出一种基于光谱形状描述的多门限的二进制编码分类方法 ,并给出了图象实验的结果。     

高血压病动态血压与中医辨证分型的关系

对高血压病的肝阳上亢和阴虚阳亢型与 2 4h动态血压的关系进行观察 ,结果 :肝阳上亢型 2 4h收缩压的最大、平均及最小值均略低于阴虚阳亢型 ,但差异无显著性 (P >0 .0 5 ) ,2 4h平均压的最大、平均及最小值显著低于阴虚阳亢型 (P <0 .0 5 ) ,2 4h舒张压的最大、平均及最小值均显著高于阴虚阳亢型 (P <0 .0 5 )。两种证型的血压昼夜变化规律亦有所不同。白昼收缩压肝阳上亢型略低于阴虚阳亢型 ,但差异无显著性 (P >0 .0 5 ) ;白昼的舒张压肝阳上亢型略高于阴虚阳亢型 (P >0 .0 5 )。夜间收缩压肝阳上亢型明显低于阴虚阳亢型 (P <0 .0 5 ) ,而夜间舒张压肝阳上亢型仍高于阴虚阳亢型 (P <0 .0 5 )。虽然两种证型皆呈现昼高夜低的特点 ,但肝阳上亢型昼夜比显著高于阴虚阳亢型 ,可见肝阳上亢型血压昼夜变化大 ,而阴虚阳亢型变化较小 ;肝阳上亢型的收缩压累计值和收缩压负荷低于阴虚阳亢型 ,而舒张压累计值和舒张压负荷则高于后者。阴虚阳亢型的脉压差亦显著高于肝阳上亢型。可见 ,肝阳上亢型以舒张压负荷升高为主 ,而阴虚阳亢型以收缩压负荷升高为主     

相似系数法在川贝母群落分类中的应用研究

目的　探讨数值分类方法应用于川贝母分布群落类型的划分。方法　以群落相似系数百分率划分川贝母分布的群落类型。结论　应用相似系数法可将川贝母野生抚育的适宜群落划定为窄叶鲜卑花灌丛、理塘杜鹃灌丛、硬叶柳灌丛、金露梅 +绣线菊灌丛、香柏灌丛、珠芽蓼 +圆穗蓼草甸等 6种类型     

高效液相色谱系统聚类分析方法在风轮菜属药材分类中的应用

目的 :探讨高效液相色谱技术应用于风轮菜属中药的鉴定研究。方法 :利用高效液相色谱提供药材的内在成分信息 ,结合模式识别技术 ,对风轮菜属植物进行了分类及鉴定。结果 :色谱数据的系统聚类结果得到了预期效果 ,这对于研究风轮菜属药材的类别及近缘性提供了依据。结论 :本方法具有简便、快捷、重现性好等优点。     

Diagnostic Rules for Children with PDD-NOS and Multiple Complex Developmental Disorder

This study was designed to examine the classification performance of diagnostic rules for pervasive developmental disorder not otherwise specified (PDD-NOS) and multiple complex developmental disorder (McDD), with clinical diagnosis as the gold standard. McDD is an heuristic concept of a developmental disorder characterised by social impairments, affective dysregulation, and thought disturbance. Detailed information on the symptoms, reliably extracted from the charts of 103 children with PDD-NOS and McDD, 32 with autistic disorder, and 96 with non-PDD disorders, was used to determine the presence of the DSMIV criteria of autistic disorder and the criteria of McDD. A scoring rule for PDD-NOS based on a short set of seven DSM-IV criteria with a cut-off point of three items and one social interaction item set as mandatory had the best balance between high sensitivity and high specificity. The most effective and simple rule based on McDD criteria had a cut-off of three items, out of six items of anxieties and thought disturbance.     

Evolutionary Classification of Homeodomains

Purpose and Methods: We performed multiple comparisons between available amino acid (aa) sequences of homeodomain(HOM)-containing proteins from a wide spectrum of animals to create an evolutionary classification of the proteins. Results: Based on results of statistical and special computational analyses of over 500 homeodomain aa sequences (HOMs) a novel system of concepts describing complex structural correlations between homologous proteins is proposed. This system includes such notions as differentiated isofunctionality of aa, chemotype, stereotype, local functional motifs, gradual conservativeness of aa positions, and group-specific domain patterns, as well as major categories of the evolutionary classification of HOMs (Division, Type, Branch, Class, Family, Series, Variety, Sort). Using this approach, a complete structural systematics of HOMs belonging to proteomes of eukaryotic animals is proposed. Conclusions: The proposed structural classification of HOMs is in full agreement with the bulk of experimental data revealing complex functional similarities and differences among HOMs in terms of their expression patterns in developing embryos. It turn, this classification can provide answers regarding homology among homeodomains when experimental data are conflicting.     

Impact of Varying Stages of Endometriosis on the Outcome of In Vitro Fertilization–Embryo Transfer

Purpose: The impact of severity of endometriosis on the outcome of in vitro fertilization (IVF) was analyzed in an uncontrolled, retrospective study in an academic IVF program. Methods: Sixty-one patients with a primary diagnosis of endometriosis undergoing 85 cycles of IVF were included in the study. Patients were divided according to the severity of disease based on the revised American Fertility Society (AFS) classification into groups A (stages I/II, or minimal/mild) and B (stages III/IV, or moderate/severe). Group A included 32 patients undergoing 45 IVF-embryo transfer (ET) cycles; group B included 29 patients undergoing 40 IVF cycles. Exclusion criteria were age older than 40 years, basal day 3 follicle stimulating hormone (FSH) greater than 20 IU/L, male-factor infertility, assisted hatching, and gamete intrafallopian transfer cases. Stimulation for IVF cycles was standard using pituitary down-regulation with gonadotropin-releasing hormone agonist in a midluteal protocol. Controlled ovarian hyperstimulation (COH) was achieved using a combination of FSH and human menopausal gonadotropin. Outcomes assessed included response to COH and number, maturity, and quality of oocytes retrieved. Fertilization, implantation, and pregnancy rates after IVF-ET were also analyzed. Results: The response to COH and the number, maturity, and quality of the oocytes was comparable between patients with varying severity of endometriosis. Fertilization rates for oocytes of patients in group B (stages III/IV) were significantly impaired compared to those in group A (stages I/II) (P = 0,004). The rates for implantation, clinical pregnancy, and miscarriage were comparable between the two groups. Conclusions: The reduced fertilization potential of the oocytes obtained from patients with severe endometriosis in the absence of male-factor infertility suggests an adverse biological impact of the advanced disease on the oocytes. The outcome of IVF-ET, however, is unaffected by increasing severity of endometriosis. This suggests that IVF may compensate for or overcome this reduction in the biological potential of the oocytes associated with severe disease, thus accounting for a comparable outcome irrespective of the severity of endometriosis.     

住院病例医疗费用分布及影响因素Logistic回归分析

目的探讨住院病例医疗费用分布及其影响因素.方法用SAS6.12软件对我国南方某市5年来近21万条住院病例进行统计分析,以各分型病例医疗费的上、下四分位数为界,将医疗费分为高、中、低三个等级及高、适度两个等级,用logistic回归分析找出影响费用的因素,并用建模数据和另外6万条数据验证模型预测效果.结果医疗费用的影响因素有:住院日、病人身份、危重病例、手术病例、年龄、医院等级、输血、转归情况(好转)、诊断个数、一级护理等18项.费用分为三个等级的logistic回归模型,对建模病例和验证病例预测符合率分别为63.81%、54.82%;费用分为两个等级的logistic回归模型,预测符合率分别为82.59%、77.32%.结果增加医疗费用划分等级,会增加模型拟合的难度,降低预测结果;利用logistic回归模型判断和预测住院病例医疗费的高低,可以及时地了解费用控制情况,为加强费用管理提供依据.     

一个快速的多维数据包分类算法

通过引入位图映射以及元组空间的概念,提出了一个数据包分类算法,该算法可以满足高性能网络对骨干路由器的要求,能够以线速完成对输入数据包的分类处理。算法在时间及空间复杂度上性能较好,并具有很好的扩展性。     

农村公共卫生投融资机制研究样本抽取的区域分类方法

目的：构建农村公共卫生投融资机制抽样研究的区域分类指标体系,并计算分类标志值.方法：本文在广泛收集我省农村公共卫生投融资方面的资料数据的基础上,运用专家咨询法确立符合本研究领域特征的指标体系和指标权重,并运用极差法对实际数据进行无量纲化处理,采用加权平均法计算出我省14个地州市的区域分类标志值.结果：我省14个地（州、市）的分类标志值虽然极差较大（反映发展不平衡）,但主要分布于（40,70）这一区间内,其中,大于60的有5个地（州、市）,小于50的有5个,在50-60之间的有4个,我们依此将其分为三个类别的区域.结论：经过专家咨询所筛选的9个指标所计算的标志值能够较好地反映一个地区的农村公共卫生投融资的现状,可以作为本研究抽样的依据.