用灰色预测模型对病毒性肝炎发病率作预测

本文应用的灰色预测模型是在原GM(1,1)残差的基础上建立的新模型,在此残差基础上建立更新的系统模型,如此循环直到r阶残差,达到精度更高的预测。现介绍如下。计算方法(1)对原始数据X_(t)作累加生成数据列:Y_(t)=sum from i=1 to t X。(2)由累加数据y_(t)作均值数据列:Z_(t)=1/2[Y_(t) Y_(t-1)]。(3)建立GM(1,1)模型:Y_(t) =(X_(1)-μ/a)     

37年麻风病流行趋势灰色预测研究

利用垫江县1952-1989年麻风各年发病数计算累计5年年均发病率(11/10万)X(t)。用GM(1,1)时间序列灰色系统建立预测模型。经计算D=10.6179 a=0.1280u=0.4594 (?)=-3.4287e~(-0.1280(t-1))+3.5891。依据(?)(t)=(?)(t)-(?)(t-1)计算(?)(t)并与X(t)对比,经拟合决定后解数     

生物医学信号处理(二)

第二讲相关技术一、概述由随机信号的观察值估计它的统计特征是一项基本处理技术。其中,数字特征(均值、均方和方差)的估计是大家熟知的(按照定义式用有限样本来估计。例如,x_1,x_2…,x_NN 个样本的均值估计(?)_x=1/N sum from n=1 to N X_n     

1990～1999年我国居民糖尿病死亡率分析及预测

目的分析 1990～ 1999年我国城市和农村居民糖尿病死亡率的变化并进行趋势预测。方法采用 1990～ 1999年《全国卫生统计年报资料》,对我国 1990～ 1999年糖尿病死亡率进行流行病学分析 ,并利用灰色动态模型 GM(1,1) ,预测我国 2 0 0 0～ 2 0 0 5年城市和农村糖尿病死亡率趋势。结果 1999年 ,糖尿病死亡率城市为 15 .37/10万 ,农村为 5 .13/10万 ;分别较 1990年上升 1.89倍和 1.71倍 ;女性高于男性 ,城市和农村男女之比分别为 0 .6∶ 1和 0 .7～ 0 .8∶ 1;所建立的灰色模型预测方程城市为 Xt=- 10 6 .4 817(1- e0 .0 74 58t) y0 ( t- 1 ) (t=1,2 ...,N) ;农村为 :Xt=- 33.6 0 5 3(1- e0 .0 7537t) y0 ( t- 1 )(t=1,2 ,...,N) ;预测到 2 0 0 5年我国城市和农村糖尿病死亡率将分别达到 2 5 .2 4 /10万和 8.15 /10万 ,精度检验证明 ,城市和农村预测模型拟合良好 ,外推测值是可信的。结论 1我国人群糖尿病死亡率上升趋势明显 ,提示有关部门应加强防治力度。 2灰色模型 GM(1,1) ,可作为糖尿病死亡率近期预测使用。     

秩和比法在疾病流行预报中的应用

1 秩和比法简介1988年作者提出了秩和比(Rank sumratio,简称RSR)法,这里的RSR是一个内涵极为丰富的统计量,具有0～1连续变量的特征。积和比的计算需按行(R)或按列(C)分别进行: RSR_R=sum from 1 to m R/m·n或RSR_c=sum from 1 n R/m·n     

应用灰色数列模型预测成都铁路局传染病发病率

目的：了解成都铁路局传染病发病变化趋势。方法：应用灰色系统基本模型建立各类传染病发病率预测模型,用相关指数R^2检验模型预测效果,并用所建模型预测成铁局传染病发病率。结果：所建灰色模型预测方程分别为总发病率-/y(t)=1280.49-11506.28e^-0.0446（t-1),肝炎-/y(t)=2659.60-2456.70e^-0.0837(t-1)。痢疾-/y(t)=4351.17-4002.46e^-0.0649(t-1),淋病-/y(t)=196.48-156.96e^-0.179(t-1),各模型预测效果较好。预测出2001－2005年发病率呈下降趋势。结论：应用灰色数列模型预测是可靠和可行的。     

图像引导调强放射治疗胸腹部肿瘤的摆位误差分析

目的 研究胸腹部肿瘤中应用图像引导调强放射治疗(Image Guided Radiation Therapy,IGRT)。方法 观察对象选择2020年2～4月广东省农垦中心医院收治的20例胸腹部肿瘤患者,全部患者接受瓦里安VitalBEAM IGRT加速器治疗,并经由锥形束CT(Cone beam CT,CBCT)技术获得线性摆位误差[左右(x)、头脚(y)、前后(z)]以及旋转摆位误差[u(对应x)、v(对应y)、w(对应z)],展开摆位误差分析。结果 ①20例患者采集CBCT共441次,各方向的平均摆位误差以“系统误差±随机误差”代表,分别为：x轴是(0.16±0.08)cm,y轴是(0.09±0.14)cm、z轴是(-0.03±0.13)cm、u轴是0.77°±0.26°、v轴是-0.21°±0.28°、w轴是0.21°±0.18°;②x方向的误差最小,而y方向的误差最大;x方向、y方向、z方向的外扩值为0.61cm、1.12cm、0.69cm。结论 在胸腹部肿瘤治疗中应用IGRT,可在线纠正摆位误差,做到低损伤、高精度治疗;且为避免摆位误差影响,当计算临床耙体积(Cli...     

灰色GM(1,1)预测模型在疾病预测中的应用

目的 预测糖尿病死亡率的变化趋势。方法 利用灰色系统GM(1，1)预测模型Y(t)=[x(1)-u／α]e^a(t-1) u/α分别预测1999—2004年糖尿病的死亡率趋势。结果 依据某市1991—1998年糖尿病死亡率资料。所建立的灰色预测模型为Y(t)=492．36e^0.0328(t-1)-477．03，拟合结果显示。模型的平均误差率为1．8％，精度为优(C=0．29，P=1)。结论 预测结果表明：近10年来糖尿病的死亡率呈持续上升趋势，建议有关部门加强糖尿病的防治工作。     

杭州市病毒性肝炎时间系列分析

本文系将杭州市病毒性肝炎的逐年发病指数进行初步分析,以较清楚地显示本病长期变化趋势。1.长期趋势:设趋势直线为 (?)=a+bt,其中 t=1、2、3……n,n 为时间系列的期数,t 为时间。(?)为时间 t 时发病指数 y 的估计值,a 为 t=0时 y 的估计值。b 为y 的平均增长率。用最小二乘法得:     

对氧磷酶和谷胱甘肽S转移酶基因多态性与农药暴露花卉种植者的健康关系研究

Objective In this study we conducted a cross-sectional study and reported on the distribution of two common genetic polymorphisms of the PON1 gene and two common genetic polymorphisms of the GST gene as well as the association between those polymorphisms and other predictors in a population of floriculture workers from Kunming city. Methods 136 pesticide-exposed farmers were recruited. PON1 and GST T1, M1 genotype were determined by means of polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Results (1)Compared with exposure group, the GLB in control group was higher(β=-0.16,t=-2.30, P=0.02), but the A/G in control group was lower (β=0.14,t=2.09,P=0.04); (2) Compared with the persons who carried QQ genotype, the accumulative symptom scores were lower in the persons who carried RR genotype(t=-2.78, P=0.006); (3) Compared with GST T1 carriers, non-carriers' GLB (β=-0.20, t=-3.01, P=0.00)and IgG(β=-0.20,t=-3.01, P=0.00)were higher, but the A/G was lower; (4) The abnormities of cardiograms among people who had been exposed to pesticides were higher compared to people who did not expose to any pesticides (β=1.147, P=0.042). Conclusion PON1-192 and GST T1 gene were associated with the farmers health condition after pesticides exposure.     

对氧磷酶和谷胱甘肽S转移酶基因多态性与农药暴露花卉种植者的健康关系研究

Objective In this study we conducted a cross-sectional study and reported on the distribution of two common genetic polymorphisms of the PON1 gene and two common genetic polymorphisms of the GST gene as well as the association between those polymorphisms and other predictors in a population of floriculture workers from Kunming city. Methods 136 pesticide-exposed farmers were recruited. PON1 and GST T1, M1 genotype were determined by means of polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Results (1)Compared with exposure group, the GLB in control group was higher(β=-0.16,t=-2.30, P=0.02), but the A/G in control group was lower (β=0.14,t=2.09,P=0.04); (2) Compared with the persons who carried QQ genotype, the accumulative symptom scores were lower in the persons who carried RR genotype(t=-2.78, P=0.006); (3) Compared with GST T1 carriers, non-carriers' GLB (β=-0.20, t=-3.01, P=0.00)and IgG(β=-0.20,t=-3.01, P=0.00)were higher, but the A/G was lower; (4) The abnormities of cardiograms among people who had been exposed to pesticides were higher compared to people who did not expose to any pesticides (β=1.147, P=0.042). Conclusion PON1-192 and GST T1 gene were associated with the farmers health condition after pesticides exposure.     

简单催化模型拟合方法的探讨

用微分拟合法并借鉴灰色均值生成理论来拟合简单催化模型 ,取得了较好的效果。1　原理和公式简单催化模型微分方程为 :dydt=r(k- y) (1)y代表在给定年龄组中有阳性病史的比率 ,r为在单位时间内每个个体的有效接触数 ,t为时间 ,k代表所有经过有效接触且试验结果为阳性的比率。满足 t=0 ,y=0的解为 :y=k(1- e- rt) (2 )根据微分拟合法 ,取 dy=Δ y,即 y的累减值 ,dt=Δt,即年龄组中值的累减值。令 I=Δ yΔ t代入 (1)式则有 x=rk- ky,可用最小二乘法求解参数。2　举例及效果评价某地百日咳曾患率与年龄关系资料 ,试作催化模型拟合。表 1　某…     

中毒患者血液中褪黑素浓度的气相色谱-串联质谱联用检测

目的 建立血液中褪黑素的气相色谱-串联质谱联用(GC-MS-MS)检测方法,为临床中毒患者的救治提供依据。方法 乙酸乙酯液-液提取血液中的褪黑素,进样1μl,色谱柱:BR-5MS,载气为氦气,流速为1. 0 ml/min;柱温起始温度60℃,保持2 min,以速率20℃/min升至280℃,保持10 min,EI离子源。通过Full Scan确定保留时间(RT)、多反应监测模式(MRM)采集数据,进行定性定量检测。结果 褪黑素色谱峰保留时间为13. 998 min,以基峰m/z 173为母离子进行MS-MS分析,最终确定子离子m/z 129. 7的最佳CID为18 e V,子离子m/z 158. 5的最佳CID为15 e V。配制系列褪黑素浓度(x)和峰面积(y)进行线形回归,得标准曲线y=230. 44x-9 535. 2,r=0. 998 4。选择信噪比(S/N) 3∶1定性、S/N 10∶1定量,得到褪黑素定性检出限为1. 0 ng/ml,定量检出限为3. 5 ng/ml;日内精密度RSD为2. 9%,日间精密度RSD为5. 3%。结论 与GC-MS相比,GC-MS-MS方法选择性好、灵敏度高、排除干扰强。     

原吸法测水中铜的不确定度评定

水样中铜离子被原子化后 ,吸收来自铜元素空心阴极灯发出的共振线 ,吸收共振线的量与该元素的含量成正比。根据这原理 ,我们采用铜标准溶液不同浓度时仪器的信号响应值作为工作曲线 ,被测水样产生的信号响应值经工作曲线查得其浓度进行测定。1　数学模型y =bx a表 1　工作曲线数据铜浓度 (mg/L)第 1次吸光值第 2次吸光值0 .1 0 .0 1 4 90 .0 1 4 50 .2 0 .0 2 71 0 .0 2 6 40 .3 0 .0 4 0 0 0 .0 4 0 00 .5 0 .0 65 1 0 .0 64 20 .7 0 .0 897 0 .0 887　　注 :铜浓度为x,吸光值为y。得a =0 0 0 2 2 2 ,b =0 12 4,r=0 99994吸光值分别为 …     

南通市通州区梅毒、淋病发病趋势预测

目的探究南通市通州区梅毒、淋病的发病趋势，为防治工作提供依据。方法 根据通州区2004-2010年梅毒、淋病的发病情况，利用灰色系统GM（1，1）建立预测模型进行预测。结果通州区2004-2010年梅毒年发病率呈上升趋势，淋病年发病率有下降趋势，趋势性x^2检验有统计学意义。梅毒、淋病趋势预测方程分别为：Y（t）=0．903957e^0.194487（t-1）-0．468266，y（t）=2．311341e^0.094 813（t-1）-2．038763。结论若无特殊情况，通州区今后2年梅毒发病率会将持续上升，淋病发病率可能会有小幅波动上升。     

首起社区甲型H1N1流行性感冒伴季节性流行性感冒暴发的调查分析

Objective To timely summarize past experience and to provide more pertinent reference for control and prevention in A/H1N1 cases in influenza season. Methods During May 25 to 31,2009,2 secondary community cases caused by a influenza A/H1N1 imported case. In the close contacts of 3 A/ H1N1 cases, 14 had some aspiratory symptoms onset, such as fever(≥ 37.5℃), cough, sore throat and etc. Laboratory tests excluded the infection of A/H1N1 influenza. For throat swab test for the 14 cases,7 were tested for seasonal influenza virus. A face-to-face or telephone interview was conducted by CDC staff to collect information of 62 close contacts. Results Of 14 fever cases,there was no significant by differences by age [15-age group: 19. 2% (5/26), over 25-age group: 25.0% (9/36); χ2=0. 287, P=0. 592]; by sex group [24. 0% (6/25) for male and 21.6% (8/37) for female; χ2=0. 048, P=0. 826], by working units [dressing and design, photograph, saleroom and others, consumer group: 42. 1% (8/19), 27. 3% (3/11), 12. 5% (2/16) and 6. 3% (1/16); χ2=7. 653, P=0. 054], by dormitory style [dormitory style=33.3% (4/12),non-dormitory style=29.4%(10/34); χ2=0. 699,P=0. 403]. All the cases had fever (37.5-37.9℃),no case had diarrhea. One in 3 A/H1N1 cases had diarrhea. All the 14 cases were negative result for A/H1N1 RNA. Six from 7 cases were positive for seasonal influenza test. Conclusion This was a seasonal influenza outbreak happened in the close contacts of first confirmed A/H1N1 cases in community in mainland China. It showed that we should exclude the seasonal influenza in the investigation of A/H1N1 cases in the seasonal influenza period in some time. It is necessary to take effective measure to strengthen the control and prevention of seasonal influenza.     

离子选择电极和酶电极在环境与临床研究中的应用

离子选择电极离子选择电极对活度为A_i、电荷为z的离子i的响应由能斯脱(Nernst)方程给出: E=常数 2.303RT/2Flog[A_i K_(ij)(A_j)~(z/y)] (1)式中E为测量电位;R是气体常数,等     

新沂市肺结核病患病率的灰色预测

本文应用GM（1．1）模型,根据新沂市1985～1994年活动性肺结核病资料对1995年～2000年肺结核病的患病率作一预测。1资料来源：本文资料取自本站历年结核病统计年报,删除其中正组病人及重报病人。原始数据列于表五。2建模方法：①按公式：求几a、。分别为：②将x（1）和a、u值代人公式：马按此模型算得Y（;）的估计值Y（;）数列和按公式：X（;）、Y《t）－Y（t－1）,t＝2、3……N,算得x（;）的估计值X（;）数列见表2。3拟合度检验：依值计算8。X（t;植计算已分别为8。＝0．10762又一0．915511。,。。。I。＿。民。L＿＿＿＿…     

甲型H1N1流感病毒基因组序列分析及其特性研究

Objective To analyse the genome of influenza A (H1N1) vires so as to elucidate its molecular characteristics and evolution status. Methods DNA sequences of the influenza viruses were collected from NCBI, and compared with the genomes of referenced intluenza viruses. The phylogenetic trees were constructed by the neighbor-joining method, and the pathogenicity, drug susceptibility and vaccine protection were analyzed. Results Phyiogenetic analysis showed that the genes encoding HA, PB2, PBI, PA, NP, and NS protein were most closely related to those influenza A viruses circulating in swine populations in North America. NA and M gene belonged to Eurasia lineages swine influenza vires. The amino acid sequence of the cleavage site between HA1 and HA2 was PARSSR ↓ GLFGAI with the typical characteristics of the low pathogenic influenza virus. Influenza A(H1N1) virus can spread from person-to-person. It is sensitive to oseltamivir and zanamivir but resistant to amantadine and remantadine. The current human seasonal influenzavaccines confered little protection against influenza A/H1N1 because of the great diversity on antigenic domains between A/H1N1 virus and vaccine virus. Conclusions Influenza A(H1N1) virus is a reassortant virus of North America and Eurasia hneages swine influenza virus. It is important to develop a vaccine against the currently circulating virus strain to control the disease spread.     

甲型H1N1流感病毒基因组序列分析及其特性研究

Objective To analyse the genome of influenza A (H1N1) vires so as to elucidate its molecular characteristics and evolution status. Methods DNA sequences of the influenza viruses were collected from NCBI, and compared with the genomes of referenced intluenza viruses. The phylogenetic trees were constructed by the neighbor-joining method, and the pathogenicity, drug susceptibility and vaccine protection were analyzed. Results Phyiogenetic analysis showed that the genes encoding HA, PB2, PBI, PA, NP, and NS protein were most closely related to those influenza A viruses circulating in swine populations in North America. NA and M gene belonged to Eurasia lineages swine influenza vires. The amino acid sequence of the cleavage site between HA1 and HA2 was PARSSR ↓ GLFGAI with the typical characteristics of the low pathogenic influenza virus. Influenza A(H1N1) virus can spread from person-to-person. It is sensitive to oseltamivir and zanamivir but resistant to amantadine and remantadine. The current human seasonal influenzavaccines confered little protection against influenza A/H1N1 because of the great diversity on antigenic domains between A/H1N1 virus and vaccine virus. Conclusions Influenza A(H1N1) virus is a reassortant virus of North America and Eurasia hneages swine influenza virus. It is important to develop a vaccine against the currently circulating virus strain to control the disease spread.