基于NIRS技术和PCA-SVM算法6种树脂及其他类中药的快速鉴别

目的:利用近红外漫反射光谱(NIRS)法,结合主成分分析(PCA)和支持向量机(SVM)联用算法,建立6种树脂及其他类中药安息香(Benzoinum),琥珀(Succinum),没药(Myrrha),乳香(Olibanum),松香(Colophonium),天竺黄(Bambusaen Concretio Silicea)的NIR模式识别模型,用于该6味中药的快速鉴别。方法:收集上述6种中药样品,经性状鉴别和理化分析确定正品药材55批,粉碎成均匀粉末,在4 000~12 000 cm~(-1)光谱区,采集各样品粉末的NIR光谱,选取特征谱段9 000~5 400,5 000~4 000 cm~(-1)为建模谱段,分别采用矢量归一化法(vector normalization,VN),一阶导数法(first derivative,FD),二阶导数法(second derivative,SD)3种不同光谱预处理方法进行预处理并分别进行PCA降维。根据主成分空间散点图,优选最佳预处理方法。利用最佳预处理方法处理后光谱的PCA降维数据,建立SVM模式识别模型,SVM模型参数c和g采用网格搜索法结合五折交叉验证进行寻优。对比不同主成分数所建PCA-SVM模型的预测准确率,确定最佳的主成分数,最终建立6种中药NIR快速鉴别模型。结果:在9 000~5 400,5 000~4 000 cm~(-1)建模谱段,确定最佳光谱预处理方法为SD,SD预处理光谱PCA降维后,确定最佳主成分数为3个,累计贡献率达93.57%。经网格搜索法确定最佳SVM建模参数组为c=65 536,g=512。所建PCA-SVM模型对训练集和验证集样品预测正确率均达100%,模型五折交叉验证准确率亦达100%。结论:所建的6种中药NIR光谱PCA-SVM鉴别模型,预测准确率高,模型预测能力强,结合NIRS技术无损、快速的优点,该模型可用于上述6种中药的无损、快速鉴别。     

体表胃电的特征提取方法研究

目的提取表征性强、稳定性高且能够为临床应用提供诊断参考的体表胃电图(EGG)特征参数。方法采用经验模态分解方法对EGG信号进行预处理,提取EGG信号的时域、频域以及非线性多维特征参数,并通过统计学方法筛选出最优特征参数组成特征向量,对功能性消化不良(FD)患者的EGG信号进行识别。结果基于时-频-非多维特征向量的FD患者的EGG信号识别率显著优于基于传统标准所构建的特征向量,其识别率最高可达90%以上。结论所提出的多维特征提取方法能够有效识别FD等胃肠疾病患者的EGG特征,可以为胃电的相关研究提供一种可靠的分析工具。     

Development and validation of a mobile application based on a machine learning model to aid in predicting dosage of vitamin K antagonists among Indian patients post mechanical heart valve replacement

Patients who undergo heart valve replacements with mechanical valves need to take Vitamin K Antagonists (VKA) drugs (Warfarin, Nicoumalone) which has got a very narrow therapeutic range and needs very close monitoring using PT-INR. Accessibility to physicians to titrate drugs doses is a major problem in low-middle income countries (LMIC) like India. Our work was aimed at predicting the maintenance dosage of these drugs, using the de-identified medical data collected from patients attending an INR Clinic in South India. We used artificial intelligence (AI) - machine learning to develop the algorithm. A Support Vector Machine (SVM) regression model was built to predict the maintenance dosage of warfarin, who have stable INR values between 2.0 and 4.0. We developed a simple user friendly android mobile application for patients to use the algorithm to predict the doses. The algorithm generated drug doses in 1100 patients were compared to cardiologist prescribed doses and found to have an excellent correlation.     

乳腺肿瘤超声图像识别模式分类方法的比较研究

目的利用乳腺肿瘤超声图像良恶性的不同特征,借助于模式分类方法对乳腺肿瘤良恶性进行识别,作为医生的计算机辅助诊断。方法本文研究基于乳腺肿瘤超声图像的原始特征参数已提取情况下,采用顺序前进搜索方法获得最优特征矢量,然后利用支撑矢量机、贝叶斯分类器、BP网络和Fisher线性判别器四种模式识别方法分别对乳腺肿瘤良恶性进行识别。结果基于200例病例随机划分为训练集100例和测试集100例进行测试,支撑矢量机、贝叶斯分类器、BP网络和Fisher线性判别器的Accuracy分别为0.960,0.940,0.932±0.013,0.930。结论支撑矢量机的分类性能优于其它分类器,能有效地对超声图像乳腺肿瘤进行良恶性识别。     

组合核支持向量机在放电模式识别中的优化策略

传统的单一核函数SVM无法实现局部放电中的多特征空间向量的映射分类,且目前SVM大都采用不同尺度的径向基函数作为核函数,核的调整空间较为有限,无法针对不同特征空间的放电参数达到普适的效果。针对上述问题,本文提出一种基于多分组特征的组合核多分类SVM的局部放电识别方法。该方法首先完成多类特征空间在不同类型SVM核函数中的映射分类,再采用骨干粒子群(BBPSO)优化方法选取最佳核参数,并求解核函数权值系数,最终形成最优核函数组合分类模型。实验结果表明,该方法对多个特征空间数据具有普适性,且融合效果理想,分类精度高于误差反向传播神经网络(BPNN)和SVM识别方法。     

Prediction of drug’s Anatomical Therapeutic Chemical (ATC) code by integrating drug–domain network

Predicting Anatomical Therapeutic Chemical (ATC) code of drugs is of vital importance for drug classification and repositioning. Discovering new association information related to drugs and ATC codes is still difficult for this topic. We propose a novel method named drug–domain hybrid (dD-Hybrid) incorporating drug–domain interaction network information into prediction models to predict drug’s ATC codes. It is based on the assumption that drugs interacting with the same domain tend to share therapeutic effects. The results demonstrated dD-Hybrid has comparable performance to other methods on the gold standard dataset. Further, several new predicted drug-ATC pairs have been verified by experiments, which offer a novel way to utilize drugs for new purposes effectively.