基于SFS-SVM的乳腺癌预测模型的构建

目的:构建基于序列前向选择算法(SFS)与支持向量机算法(SVM)分类器融合的乳腺癌预测模型,提高计算机辅助诊断技术对乳腺癌细针穿刺细胞病理的准确率。方法:对456组乳腺肿瘤病理数据作为训练集,利用SFS-SVM算法对30个特征进行筛选,得到最优的特征组合,再用112组乳腺肿瘤病理数据作为测试集验证,构建乳腺癌预测模型。该模型的预测精度通过5折交叉验证进行评价。评价指标包括:受试者工作特性曲线(ROC)下面积(AUC)、准确率(ACC)、敏感度和特异度。结果:构建了基于SFS-SVM的乳腺癌预测模型,该模型(AUC为98.39%,ACC为97.35%)相对于单独SVM算法(AUC为97.00%, ACC为92.42%)有一定的提高。结论:基于SFS特征选择的SVM分类器乳腺癌预测模型能较好地对乳腺癌进行辅助诊断。     

基于Lasso特征选择的自闭症预测

目的寻找自闭症(autism spectrum disorders,ASD)的客观生物标记以辅助临床诊断。静息态功能连接(resting-state functional connectivity,RSFC)反映了大脑不同脑区神经活动模式间的时间相关性,研究者常从RSFC中探索识别ASD的生物标记,然而大部分方法还不能有效选出具有识别力的RSFC。本文采用最小绝对值收缩与选择算子(least absolute shrinkage and selection operator,Lasso)来选择ASD组和正常发育组之间具有显著性差异的RSFC特征。方法首先基于Pearson相关分析提取RSFC特征,并进行阈值化处理保留具有较大正相关值的RSFC。然后采用Lasso方法提取有识别能力的RSFC,最后利用支持向量机进行ASD分类,并主要以分类准确率指标对分类性能进行评估。结果基于Lasso方法进行特征选择后,ASD分类准确率为81.52%,同时找出了能显著区分ASD儿童与正常儿童的RSFC。结论基于Lasso特征选择的方法提高了对ASD的识别准确率,识别的生物标记有潜力应用于临床诊断。     

基于定量影像组学的乳腺肿瘤良恶性诊断

乳腺癌是女性致死率最高的恶性肿瘤之一。为提高诊断效率,提供给医生更加客观和准确的诊断结果。借助影像组学的方法,利用公开数据集BreaKHis中82例患者的乳腺肿瘤病理图像,提取乳腺肿瘤病理图像的灰度特征、Haralick纹理特征、局部二值模式(LBP)特征和Gabor特征共139维影像组学特征,并用主成分分析(PCA)对影像组学特征进行降维,然后利用随机森林(RF)、极限学习机(ELM)、支持向量机(SVM)、k最近邻(kNN)等4种不同的分类器构建乳腺肿瘤良恶性的诊断模型,并对上述不同的特征集进行评估。结果表明,基于支持向量机的影像组学特征的分类效果最好,准确率能达到88.2%,灵敏性达到86.62%,特异性达到89.82%。影像组学方法可为乳腺肿瘤良恶性预测提供一种新型的检测手段,使乳腺肿瘤良恶性临床诊断的准确率得到很大提升。     

基于偏序拓扑图的帕金森病语音障碍分析方法

从形式概念分析角度,提出将偏序拓扑图用于帕金森病语音障碍分析与诊断。首先,在属性拓扑的基础上,结合偏序结构表示,构造偏序拓扑图的形式背景表示方法,并利用偏序拓扑图进行概念本体计算,获得原始形式背景的层次化概念树结构。进而结合决策属性,对概念树进行着色与约简,获得约简概念树。根据约简概念树的偏序关系,可获得分析对象的概念分类结构。将该方法应用于帕金森病语音特征数据集进行概念提取,实验表明不但可以在概念层面分析帕金森病与语音特征的关系,同时可以作为诊断依据进行数据诊断。将该方法应用于多个帕金森病数据集(样本数分别为197、5 875、1 040、220)进行分类精度对比测试,表明基于偏序拓扑图的帕金森病语音障碍分析在不同的帕金森病语音数据集下的平均诊断精度达到76.64%,高于LDA(67.36%)、QDA(70.83%)、kNN(71.83%)、parzen窗(70.24%)、SVM(74.61%)等经典分类器的诊断精度,高出经典分类器SVM 2.72%,表明该方法能有效应用于帕金森病语音障碍分析。     

基于双分支多视角深度自注意力编码器的乳腺肿瘤分类方法

自动乳腺全容积超声成像(ABVS)系统因其高效、无辐射等特性成为筛查乳腺癌的重要方式。针对ABVS图像进行计算机辅助乳腺肿瘤良恶性分类的研究,有利于帮助临床医生准确、快速地进行乳腺癌的诊断,甚至可辅助提高低年资医生的诊断水平。ABVS系统产生的三维乳腺图像数据量较大,造成常规的深度学习方式训练时间长、占用资源巨大。本研究设计了一种基于ABVS数据的多视角图像提取方式,替代常规的三维数据输入,在降低参数量的同时弥补二维深度学习中的空间关联性;其次,基于交叉视角图像的空间位置关系,提出一种深度自注意力编码器(Transformer)网络,用于获得图像的有效特征表达。实验是基于自有ABVS数据库的153例容积图像,良恶性分类的准确率为86.88%,F1-评分为81.70%,AUC达到0.831 6。所提出的方法有望应用于ABVS图像的乳腺肿瘤良恶性筛查。     

基于Skip-gram词嵌入算法的结构化患者特征表示方法研究

目的基于表示学习中的Skip-gram词嵌入算法,寻找能够克服电子病历中结构化特征的高维性并在语义层次上表示特征的方法。方法本文的数据来源于北京市某三甲医院的电子病历系统,从中提取患者的结构化特征,包括疾病、药物和实验室指标,其中实验室指标通过正常值范围离散化;利用Skip-gram算法,将电子病历中离散型患者特征(疾病和药物)和离散后的连续型患者特征(实验室指标)嵌入到同一个低维实数向量空间中。通过t-SNE降维可视化方法显示低维实数空间中特征向量的关系,并与特征向量间的余弦距离计算结果相互印证,从而评价特征表示的有效性和揭示特征向量间的潜在联系。结果患者特征的低维实数向量既降低了患者特征的维度,又很好地表征了特征间的潜在联系,临床含义相关的特征表示成的低维实数向量也很相近。结论基于Skip-gram算法将患者结构化特征表示成低维实数向量取得了较好的效果,为解决EMR数据表示的高维性以及结构化特征间潜在关系分析提供一种思路。     

基于机器学习方法的胃癌分型标志基因提取

基于机器学习方法寻找和发现新的胃癌亚型分类的相关基因,可以为探讨胃癌发生的分子机制及其基因水平的诊断和治疗提供标志和依据.试验选用33例中国人的胃癌Oligo基因芯片数据,数据包括13例弥漫型胃癌样本、20例肠型胃癌样本,基因向量为21 378个.采用基因表达差异显著性分析方法(SAM)、偏最小二乘VIP系数法(PLS)和基于巴氏距离的顺序前向搜索方法(BD-SFS)结合的多步骤降维方法,提取到20个能将弥漫型样本和肠型样本有效分开的特征基因.这些特征基因基于支持向量机(SVM),分类准确率可达到89.43%;基于分层聚类分析,准确率可达到93.94%.同时,基因生物学意义的分析结果显示,所选的大部分标志基因对于人类恶性肿瘤的诊断和分型有很重要的意义.     

基于静息态功能磁共振成像的精神分裂症脑网络特征分类研究

如何从复杂的静息态功能核磁共振成像(rs-fMRI)数据中提取高鉴别性特征,是提升精神分裂症识别精度的关键。本文使用一种加权稀疏脑网络构建方法,采用肯德尔相关系数(KCC)从脑网络中提取连接特征,并基于线性支持向量机对57例精神分裂症患者与64例健康受试者进行分类研究,最终得到了较高的分类精度(81.82%)。本文研究结果表明,相较于传统的皮尔逊相关和基于稀疏表示的脑网络构建方法,以及常用的双样本t检验(t-test)和最小绝对收缩与选择算子(Lasso)特征选择方法,本文提出的算法可以更有效地提取出能够区分精神分裂症患者与健康人群的脑功能网络连接特征,进而提升分类精度;同时本研究中所提取的鉴别性连接特征或可作为潜在的临床生物学标志物,用以辅助精神分裂症的诊断。     

基于Relief算法的心血管疾病辅助诊断研究

本文研究了Relief特征选择方法在光电容积脉搏波(PPG)中的应用,分析寻找区分心血管疾病的指标,提出了一种辅助心血管疾病诊断的方法。通过收集40位志愿者的生理病理信息,并实时采集血压与指尖PPG波形数据,形成样本数据集。基于PPG波形,定义并提取了52个特征参数,通过特征选择Relief算法筛选出10个核心特征参数,形成最优特征子集,并分析它们对心血管疾病的影响。最后使用分类算法建模,对心血管疾病做出了辅助诊断,k邻近算法(k NN)模型对心血管疾病的预测正确率达到66.67%,支持向量机(SVM)模型对心血管疾病的预测正确率达到83.33%。结果表明:(1)年龄对心血管疾病辅助诊断最为重要;(2)最优特征子集元素特征为心血管健康状况评价与预测提供了重要依据。本研究表明,经Relief算法选择得到的最优特征子集为心血管疾病辅助诊断提供了更高的准确性。     

方剂配伍规律的可视化表示方法与知识发现

方剂配伍规律研究是中医现代化研究的核心问题之一。随着数据挖掘技术的发展和中医信息化的逐渐深入,很多数据挖掘方法已被应用到方剂配伍规律研究领域。基于形式概念分析理论,提出一种偏序结构图分层表示的方剂配伍知识可视化方法。以《张仲景方方族》中小青龙汤类方剂为例,说明知识发现过程。以该类方剂中的方剂与药物、证候与药物为对象和属性分别构建偏序结构图,依据属性特征定义及偏序结构图的层次关系分析方剂配伍规律。结果表明,根据方剂与药物偏序结构图的层次和涵盖支路情况,可以直观地发现小青龙汤类方剂中包含1味核心药五味子;高频药物包括细辛和半夏,其中细辛出现频次为13次,半夏出现频次为10次;常用的药对有8对,药组有3组。从不同簇集角度分析可以发现,小青龙汤类方剂可以聚类为5大簇集,每个簇集的方剂组成、主治功效等具有共性。根据证候与药物偏序结构图可以发现,除小青龙汤证候外,12个证候均是在小青龙汤证候基础上加减变化而成的。可见,偏序结构图可视化表示方法可清晰地反映出方剂与药物、药物与证候之间的配伍群结构。     

An intelligent framework for the classification of the 12-lead ECG.

An intelligent framework has been proposed to classify an unknown 12-Lead electrocardiogram into one of a possible number of mutually exclusive and combined diagnostic classes. The framework segregates the classification problem into a number of bi-dimensional classification problems, requiring individual bi-group classifiers for each individual diagnostic class. The bi-group classifiers were generated employing Neural Networks (NN), combined with a combination framework containing an Evidential Reasoning framework to accommodate for any conflicting situations between the bi-group classifiers. A number of different feature selection techniques were investigated with the aim of generating the most appropriate input vector for the bi-group classifiers. It was found that by reducing the original input feature vector, the generalisation ability of the classifiers, when exposed to unseen data, was enhanced and subsequently this reduced the computational requirements of the network itself. The entire framework was compared with a conventional approach to NN classification and a rule based classification approach. The framework attained a significantly higher level of classification in comparison with the other methods; 80.0% compared with 66.7% for the rule based technique and 68.00% for the conventional neural approach.     

基于支持向量机的乳腺癌预后状态预测和疗效评估

乳腺癌是危害妇女健康的主要恶性肿瘤.目前基因与疾病关系的研究取得了一系列的成果,使得利用乳腺癌患者的基因信息来预测预后状态和评估治疗效果成为了可能.支持向量机(support vector machine,SVM)分类方法在实际二类分类问题的应用中显示出良好的学习和泛化能力,已被广泛地应用于诸多研究领域.本文采用支持向量机SVM、K-近邻法(K-nearest neighbor,K-NN)、概率神经网络(probabilistic neural network,PNN)、决策树(decision tree,DT)分类器,结合乳腺癌患者基因数据来预测患者的预后状态和评估治疗效果.结果表明:当使用高斯径向基核函数时,SVM通过5次交叉验证的最佳平均分类准确率达到了88.44%,优于K-NN(81.69%)、PNN(80.68%)和DT(71.19%)等分类器,表明该方法有望成为一种有效、实用的乳腺癌预后状态预测和治疗效果客观评价的工具.     

Genetic diagnosis of cancer by fuzzy-rough gene selection and the complementary hierarchical fuzzy classifier

Gene expression data have extremely high dimensionality with respect to traditional classifiers which causes not to be used efficiently. In this paper a Fuzzy-Rough Gene Selection and Complementary Hierarchical Fuzzy classifier (FRGS-CHF) to classify the gene expression data as a new methodology is proposed. First, some relevant genes are selected using fuzzy-rough attribute selection method. After removing redundant genes, a new complementary hierarchical fuzzy classifier is proposed. The complementary learning mechanism refers to positive and negative learning which are found in the human brain hippocampus. FRGS-CHF is made-up of two parallel hierarchical fuzzy systems; the first is trained with positive samples whilst the other is treated with negative samples. In contrast to many other methods such as statistical or neural networks, FRGS-CHF provides greater interpretability. It does not rely on the assumption of underlying data distribution. Using complementary and hierarchical approaches, the proposed method exploits the lateral inhibition between output classes and considers the problem as a multidimensional problem. Benchmarked datasets are used to demonstrate the validity and advantages of the proposed method over the other existing methods in terms of the accuracy, better transparency, time efficiency together with fewer fuzzy rules and parameters.     

Characterization of breast tumors using machine learning based upon multiparametric magnetic resonance imaging features

Magnetic resonance imaging (MRI) is playing an important role in the classification of breast tumors. MRI can be used to obtain multiparametric (mp) information, such as structural, hemodynamic, and physiological information. Quantitative analysis of mp-MRI data has shown potential in improving the accuracy of breast tumor classification. In general, a large set of quantitative and texture features can be generated depending upon the type of methodology used. A suitable combination of selected quantitative and texture features can further improve the accuracy of tumor classification. Machine learning (ML) classifiers based upon features derived from MRI data have shown potential in tumor classification. There is a need for further research studies on selecting an appropriate combination of features and evaluating the performance of different ML classifiers for accurate classification of breast tumors. The objective of the current study was to develop and optimize an ML framework based upon mp-MRI features for the characterization of breast tumors (malignant vs. benign and low- vs. high-grade). This study included the breast mp-MRI data of 60 female patients with histopathology results. A total of 128 features were extracted from the mp-MRI tumor data followed by features selection. Five ML classifiers were evaluated for tumor classification using 10-fold crossvalidation with 10 repetitions. The support vector machine (SVM) classifier based on optimum features selected using a wrapper method with an adaptive boosting (AdaBoost) technique provided the highest sensitivity (0.96 ± 0.03), specificity (0.92 ± 0.09), and accuracy (94% ± 2.91%) in the classification of malignant versus benign tumors. This method also provided the highest sensitivity (0.94 ± 0.07), specificity (0.80 ± 0.05), and accuracy (90% ± 5.48%) in the classification of low- versus high-grade tumors. These findings suggest that the SVM classifier outperformed other ML methods in the binary classification of breast tumors.     

Improving breast cancer risk stratification using resonance-frequency electrical impedance spectroscopy through fusion of multiple classifiers

This study aims to improve breast cancer risk stratification. A seven-probe resonance-frequency-based electrical impedance spectroscopy (REIS) system was designed, assembled, and utilized to establish a data set of examinations from 174 women. Three classifiers, including artificial neural network (ANN), support vector machine (SVM), and Gaussian mixture model (GMM), were independently developed to predict the likelihood of each woman to be recommended for biopsy. The performances of these classifiers were compared, and seven fusion methods for integrating these classifiers were investigated. The results showed that among the three classifiers, the ANN yielded the highest performance with an area under the curve (AUC) of 0.81 for the receiver operating characteristic (ROC), while SVM and GMM achieved AUCs of 0.80 and 0.78, respectively. Improvements of up to 3% were obtained using fusion of the three classifiers, with the largest improvement obtained using either a “minimum score” rule or a “weighted sum” rule. Comparing different combinations of two out of the three classifiers, the weighted sum rule provided the most robust and consistent results, with AUCs of 0.81, 0.83, and 0.82 for the different combinations of ANN and SVM, ANN and GMM, and SVM and GMM, respectively. Furthermore, at 90% specificity, the ANN, the weighted sum- and min rule-based classifiers, all detected 67% of the verified cancer cases as compared with 50, 50, and 60% detection of the high risk cases, respectively. The study demonstrated that REIS examinations provide relevant information for developing breast cancer risk stratification tools and that using fusion of several not-fully-correlated classifiers can improve classification performance.     

Can–Evo–Ens: Classifier stacking based evolutionary ensemble system for prediction of human breast cancer using amino acid sequences

The diagnostic of human breast cancer is an intricate process and specific indicators may produce negative results. In order to avoid misleading results, accurate and reliable diagnostic system for breast cancer is indispensable. Recently, several interesting machine-learning (ML) approaches are proposed for prediction of breast cancer. To this end, we developed a novel classifier stacking based evolutionary ensemble system “Can–Evo–Ens” for predicting amino acid sequences associated with breast cancer. In this paper, first, we selected four diverse-type of ML algorithms of Naïve Bayes, K-Nearest Neighbor, Support Vector Machines, and Random Forest as base-level classifiers. These classifiers are trained individually in different feature spaces using physicochemical properties of amino acids. In order to exploit the decision spaces, the preliminary predictions of base-level classifiers are stacked. Genetic programming (GP) is then employed to develop a meta-classifier that optimal combine the predictions of the base classifiers. The most suitable threshold value of the best-evolved predictor is computed using Particle Swarm Optimization technique. Our experiments have demonstrated the robustness of Can–Evo–Ens system for independent validation dataset. The proposed system has achieved the highest value of Area Under Curve (AUC) of ROC Curve of 99.95% for cancer prediction. The comparative results revealed that proposed approach is better than individual ML approaches and conventional ensemble approaches of AdaBoostM1, Bagging, GentleBoost, and Random Subspace. It is expected that the proposed novel system would have a major impact on the fields of Biomedical, Genomics, Proteomics, Bioinformatics, and Drug Development.     

An Efficient Approach for Automated Mass Segmentation and Classification in Mammograms

Breast cancer is becoming a leading death of women all over the world; clinical experiments demonstrate that early detection and accurate diagnosis can increase the potential of treatment. In order to improve the breast cancer diagnosis precision, this paper presents a novel automated segmentation and classification method for mammograms. We conduct the experiment on both DDSM database and MIAS database, firstly extract the region of interests (ROIs) with chain codes and using the rough set (RS) method to enhance the ROIs, secondly segment the mass region from the location ROIs with an improved vector field convolution (VFC) snake and following extract features from the mass region and its surroundings, and then establish features database with 32 dimensions; finally, these features are used as input to several classification techniques. In our work, the random forest is used and compared with support vector machine (SVM), genetic algorithm support vector machine (GA-SVM), particle swarm optimization support vector machine (PSO-SVM), and decision tree. The effectiveness of our method is evaluated by a comprehensive and objective evaluation system; also, Matthew’s correlation coefficient (MCC) indicator is used. Among the state-of-the-art classifiers, our method achieves the best performance with best accuracy of 97.73 %, and the MCC value reaches 0.8668 and 0.8652 in unique DDSM database and both two databases, respectively. Experimental results prove that the proposed method outperforms the other methods; it could consider applying in CAD systems to assist the physicians for breast cancer diagnosis.