摘要：GeO_(2) is commonly used as dopant to adjust the refractive index profile(RIP)and the acoustic velocity profile(AVP)in the fiber,thereby forming different Brillouin gain spectrum(BGS)characteristics such as Brillouin gain,acoustic mode number and peak intensity *** an optical fiber is used in optical fiber sensing or communication system,its BGS characteristics may play an important role in determining the performance of the *** this paper,finite element analysis(FEA)method is used to study the influence of refractive index distribution and its corresponding AVP on the BGS in step-index,graded-index,and complex-index optical fibers.A new method has also been proposed to efficiently discriminate acoustic mode solution and obtain the new and full images of total Brillouin gain and acoustic modes number of the fiber as a function of the refractive index distribution,considering the influence of changing the refractive index difference and the geometric size *** each type of optical fiber,the recommended parameter range is provided for optical fiber sensing and optical fiber ***,the suitable optical fiber with close peak intensity in its multi-peak BGS is explored and achieved,which can be used in Brillouin beat spectrum detection systems to improve sensing accuracy.

摘要：目的基于深静脉血栓形成(deep vein thrombosis,DVT)发生风险的评估数据,对泌尿外科肿瘤患者DVT发生风险的相关性因素及预后模型进行探究,以此辅助临床更好地进行风险评估,做出准确的预后判断并采取相应预防措施。方法抽取选用复旦大学附属肿瘤医院2019年12月—2021年12月收治的泌尿外科肿瘤患者住院期间建立的3814条DVT发生风险评估表单记录的数据。首先,对数据样本进行相关性因素提取,并行数据清洗、脱敏及结构化处理;然后,使用Mann-Whitney U检验对特征数据进行单因素分析,使用Logistic回归模型进行回归性分析,得到患者DVT发生风险的显著性相关因素;最后,基于机器学习支持向量机(support vector machine,SVM)算法和决策树算法,采用交叉验证方法训练分类器并检验相关性因素对患者DVT发生风险的预测能力。结果Mann-Whitney U检验分析结果显示,身体质量指数(body mass index,BMI)、活动、特殊风险、外科手术与患者DVT发生风险相关(P<0.05)。Logistic回归分析显示,BMI、活动、特殊风险、外科手术与患者DVT发生风险显著相关(P<0.05),SVM分类器分类结果显示最高分类准确率为0.876,ROC曲线下的面积(area under the curve,AUC)为0.904,即这4种特征可以对患者的DVT发生风险做出较为准确的预测。结论BMI、活动、特殊风险以及外科手术4种因素是泌尿外科肿瘤患者DVT发生风险的显著性相关因素。