基于四类运动想象任务的脑电信号识别

背景：脑-机接口是在大脑与外部设备之间建立的直接的交流通路,基于运动想象的脑-机接口研究已经从两类运动想象任务的识别发展到多类任务的识别。 目的：探寻准确有效的对多任务运动想象脑电信号进行特征提取及模式识别的方法。 方法：首先采用公共平均参考法减小多通道中各导联间的相关性,提高脑电信号的信噪比。并对公共空间模式算法进行扩展,采用“一对多”的策略,对4类任务的脑电信号进行特征提取,在模式识别过程中,采用基于决策树法的支持向量机进行分类。对于实验对象样本不充足,结合支持向量机和贝叶斯分类器,将分类结果中具有大概率的测试样本扩充到训练集,最后再次运用支持向量机进行分类。 结果与结论：最佳正确率达到92.78%,“一对多”的公共空间模式和基于决策树的支持向量机可以有效地进行多任务脑电信号识别,扩充样本可以提高分类正确率。     

模糊聚类支持向量机在步态分类中的应用

背景：对于患有神经系统或骨骼肌肉系统疾病的患者,分析步态数据可以评定康复程度,制定治疗方案。如何有效地分类小样本步态数据成为重要的研究课题。 目的：用改进的支持向量机算法对小样本步态数据进行分类,准确诊断疾病。 方法：建立加入模糊C均值聚类的支持向量机算法,选用Gait Dynamics in Neuro-Degenerative Disease Data Base 40~59岁年龄段的6组数据,共720个样本数据,采用左摆间隔和左支撑间隔两维参数对步态数据建模。数据归一化后,通过模糊C均值聚类对数据进行预处理;然后用支持向量机对数据进行分类。采用不同核函数的支持向量机算法验证分类能力。 结果与结论：实验结果表明,利用改进的支持向量机算法,可以有效地对信号进行分类,有助于疾病的诊断和治疗方案的制定。     

基于改进的径向基函数网络的冠心病心电信号识别方法

研究一种利用径向基函数(RBF)神经网络识别冠心病心电信号模式的方法。讨论了径向基函数中心的选取，构造了改进的RBF网络对训练样本和测试样本进行识别。结果表明，此项研究中采用的神经网络能对训练样本和测试样本正确地进行模式识别，训练方法能够自适应的确定聚类个数，从而确定聚类中心，避免了K均值聚类方法中因K值选取的不同而造成的误差。此方法收敛速度快，是一种有效的识别冠心痛心电信号的方法。     

相关向量机在肿瘤表达谱分类问题中的应用

基因芯片技术能够检测大量基因的表达水平,在肿瘤研究中得到日益广泛的应用。基于基因芯片表达谱的肿瘤分类诊断是肿瘤表达谱研究的一个热点,肿瘤表达谱分类是一个典型的高维度小样本分类问题,描述一个两步策略的分类方法。在测试的基因表达谱中存在大量的非差异表达冗余基因,通过一个有效的基因预选择策略得到一个较小的候选基因子集,然后建立基于相关向量机的分类预测模型。在4个真实的肿瘤表达谱数据上,与几种不同的方法进行比较,结果显示该方法可以得到更好的分类精度,同时表现出很好的稳定性。     

一种基于复合稀疏表示的阿尔茨海默病的诊断方法

利用磁共振影像数据实现对阿尔茨海默病的准确诊断。将常规稀疏表示中的单层字典分解为两层,分别使用各类别的典型样本和类内差异作为两层字典的元素;设计一种两层字典协调工作的复合稀疏表示形式,以期利用训练样本更为精确地表示待识别样本,并构建分类器用于阿尔茨海默病的分类识别。在ADNI数据库的对比实验表明,该方法的识别性能优于支持向量机和同类的稀疏表示分类器。     

基于流形学习的聚类方法在基因芯片表达谱分析中的应用

对基因芯片表达谱的聚类分析有助于发现共表达的基因,而共表达的特性往往是共调控基因所拥有的性质。因此,对基因表达谱的准确聚类将有利于更加准确地发现基因之间的调控关系。本研究使用机器学习中的等度规映射、局部线性嵌入、拉普拉斯特征根映射等流形学习方法处理基因表达谱数据,得到非线性降维后的数据。在此基础上应用K均值聚类、模糊聚类、自组织映射神经网络等聚类方法,根据给定的阈值,从酵母基因表达数据的382个聚类结果中得到了117个共表达基因对,而从人类血清组织细胞的基因表达数据的132个聚类结果中得到了89个共表达基因对。使用的判别准则表明,基于流形学习的聚类方法与以往的方法相当,且能够被用以发现高维基因芯片表达数据中的低维的流形结构。     

基于压缩感知算法的基因表达谱数据分析

目的 基因表达谱数据分析是生物信息学领域最重要的研究内容之一.其可实现对不同病理分型的肿瘤的正确分类,对肿瘤诊断和治疗具有重大意义.方法 本文应用压缩感知算法实现对胃癌基因表达谱数据的分类,运用训练数据构造冗余字典,采用随机分布的规范行矢量高斯矩阵构造感知矩阵,对训练数据和测试数据进行感知,利用正交l2-范数算法对基因表达谱数据进行重建,在变换域中采用近邻法测试判断数据类别,与样本的实际类别相比较.结果 实验结果表明,压缩感知算法与K均值聚类、SVM等其他分类算法相比有较高的分类正确率,且分类速度快,能避免特征选取的问题.结论 本文方法对疾病的临床诊断和生物信息学研究有重要的参考和借鉴作用.     

基于迁移学习和支持向量机的乳腺癌分子分型预测

乳腺癌分子分型对乳腺癌的治疗具有决定性的参考作用,传统的分型方法有创且可能存在假阳性问题,而已有的基于影像学的分型方法准确率较低。本文提出一种利用迁移学习提取特征并结合支持向量机的分型预测方法,对乳腺癌PET/CT标记图像进行融合和归一化,再使用Xception迁移学习网络进行特征提取,最后使用支持向量机进行分类实现分型。对样本测试集进行性能评估表明,Xception+SVM模型的准确率达到0.687,AUC为0.787,优于现有基于影像学的方法,验证了本文方法的有效性。     

基于异方差混合转移分布模型和支持向量机的脑电分类研究

为解决线性分析和单一非线性动力学指标方法无法准确描述脑电信号的问题,本研究提出基于异方差混合转移分布模型脑电特征提取方法。首先对采集到的脑电信号依据条件期望最大化(ECM)算法建立异方差混合转移分布模型,求得模型条件方差序列的均值及方差作为脑电信号的特征,将得到的脑电信号特征采用支持向量机进行分类。通过对6个人的正常脑电信号和带有眼电伪迹脑电信号进行分类仿真实验,其结果表明该方法能很好地拟合出脑电信号,且分类精确度能达到99.166 7%,说明此方法可有效提取脑电特征并准确识别出眼电伪迹。     

肌电信号运动模式识别中典型样本集的选取

训练样本的质量直接影响神经网络的识别能力,我们针对手部运动模式分类问题,提出了一种典型样本的提取方法。首先,利用“类属函数”对原始样本进行预处理,以提高聚类样本的质量;然后利用聚类分析方法求得各类样本的聚类中心,得到典型样本。在此基础上,使用该方法来获得手部肢体张开、合拢动作的典型样本,将其作为训练样本BP神经网络进行训练,完成运动模式的识别。实验表明,通过提取运动模式的典型样本的方式可提高模式识别准确率。     

Tumor classification by tissue microarray profiling: random forest clustering applied to renal cell carcinoma.

We describe a novel strategy (random forest clustering) for tumor profiling based on tissue microarray data. Random forest clustering is attractive for tissue microarray and other immunohistochemistry data since it handles highly skewed tumor marker expressions well and weighs the contribution of each marker according to its relatedness with other tumor markers. This is the first tumor class discovery analysis of renal cell carcinoma patients based on protein expression profiles. The tissue array data contained at least three tumor samples from each of 366 renal cell carcinoma patients. The eight tumor markers explore tumor proliferation, cell cycle abnormalities, cell mobility, and the hypoxia pathway. Since the procedure is unsupervised, no clinicopathological data or traditional classifications are used a priori. To explore whether the tissue microarray data can be used to identify fundamental subtypes of renal cell carcinoma patients, we first carried out random forest clustering of all 366 patients. By analyzing the tumor markers simultaneously, the procedure automatically detected classes that correspond to clear- vs non-clear cell tumors (demonstration of proof-of-principle). The resulting molecular grouping provides better prediction of survival (logrank P=0.000090) than this classical pathological grouping (logrank P=0.023). We then sought to extend the class discovery by searching for finer subclasses of clear cell patients. The procedure automatically discovered: (a) two classes corresponding to low- and high-grade patients (demonstration of proof-of-principle); (b) a subgroup of long-surviving clear cell patients with a distinct molecular profile and (c) two novel tumor subclasses in low-grade clear cell patients that could not be explained by any clinicopathological variables (demonstration of discovery).     

Genome-wide analysis of acute myeloid leukemia with normal karyotype reveals a unique pattern of homeobox gene expression distinct from those with translocation-mediated fusion events

Gene expression profiles were determined from presentation peripheral blood and bone marrow samples of 28 patients with acute myeloid leukemia (AML). Hierarchical clustering sorted the profiles into separate groups, each representing one of the major cytogenetic classes in AML [i.e., t(8;21), t(15;17), inv(16), 11q23, and normal karyotype]. Statistical group comparison identified genes whose expression was strongly correlated with these chromosomal classes. Moreover, the normal karyotype AMLs were characterized by distinctive up-regulation of certain members of the class I homeobox A and B gene families, implying a common underlying genetic lesion. These data reveal novel diagnostic and therapeutic targets and demonstrate the potential of microarray-based dissection of AML.     

Computational selection of distinct class- and subclass-specific gene expression signatures

In this investigation we used statistical methods to select genes with expression profiles that partition classes and subclasses of biological samples. Gene expression data corresponding to liver samples from rats treated for 24 h with an enzyme inducer (phenobarbital) or a peroxisome proliferator (clofibrate, gemfibrozil or Wyeth 14,643) were subjected to a modified Z-score test to identify gene outliers and a binomial distribution to reduce the probability of detecting genes as differentially expressed by chance. Hierarchical clustering of 238 statistically valid differentially expressed genes partitioned class-specific gene expression signatures into groups that clustered samples exposed to the enzyme inducer or to peroxisome proliferators. Using analysis of variance (ANOVA) and linear discriminant analysis methods we identified single genes as well as coupled gene expression profiles that separated the phenobarbital from the peroxisome proliferator treated samples and discerned the fibrate (gemfibrozil and clofibrate) subclass of peroxisome proliferators. A comparison of genes ranked by ANOVA with genes assessed as significant by mixedlinear models analysis [J. Comput. Biol. 8 (2001) 625] or ranked by information gain revealed good congruence with the top 10 genes from each statistical method in the contrast between phenobarbital and peroxisome proliferators expression profiles. We propose building upon a classification regimen comprised of analysis of replicate data, outlier diagnostics and gene selection procedures to utilize cDNA microarray data to categorize subclasses of samples exposed to pharmacologic agents.     

Systematic learning of gene functional classes from DNA array expression data by using multilayer perceptrons

Recent advances in microarray technology have opened new ways for functional annotation of previously uncharacterised genes on a genomic scale. This has been demonstrated by unsupervised clustering of co-expressed genes and, more importantly, by supervised learning algorithms. Using prior knowledge, these algorithms can assign functional annotations based on more complex expression signatures found in existing functional classes. Previously, support vector machines (SVMs) and other machine-learning methods have been applied to a limited number of functional classes for this purpose. Here we present, for the first time, the comprehensive application of supervised neural networks (SNNs) for functional annotation. Our study is novel in that we report systematic results for ~100 classes in the Munich Information Center for Protein Sequences (MIPS) functional catalog. We found that only ~10% of these are learnable (based on the rate of false negatives). A closer analysis reveals that false positives (and negatives) in a machine-learning context are not necessarily "false" in a biological sense. We show that the high degree of interconnections among functional classes confounds the signatures that ought to be learned for a unique class. We term this the "Borges effect" and introduce two new numerical indices for its quantification. Our analysis indicates that classification systems with a lower Borges effect are better suitable for machine learning. Furthermore, we introduce a learning procedure for combining false positives with the original class. We show that in a few iterations this process converges to a gene set that is learnable with considerably low rates of false positives and negatives and contains genes that are biologically related to the original class, allowing for a coarse reconstruction of the interactions between associated biological pathways. We exemplify this methodology using the well-studied tricarboxylic acid cycle.     

Reducing microarray data via nonnegative matrix factorization for visualization and clustering analysis

In microarray data analysis, each gene expression sample has thousands of genes and reducing such high dimensionality is useful for both visualization and further clustering of samples. Traditional principal component analysis (PCA) is a commonly used method which has problems. Nonnegative Matrix Factorization (NMF) is a new dimension reduction method. In this paper we compare NMF and PCA for dimension reduction. The reduced data is used for visualization, and clustering analysis via k-means on 11 real gene expression datasets. Before the clustering analysis, we apply NMF and PCA for reduction in visualization. The results on one leukemia dataset show that NMF can discover natural clusters and clearly detect one mislabeled sample while PCA cannot. For clustering analysis via k-means, NMF most typically outperforms PCA. Our results demonstrate the superiority of NMF over PCA in reducing microarray data.     

Epitope clustering in regions undergoing efficient proteasomal processing defines immunodominant CTL regions of a tumor antigen

Identification of immunodominant CD8(+) T cell responses to frequently expressed tumor antigens across MHC class I polymorphism is essential for the implementation of cancer immunotherapy. However, the key factors that determine immunodominance are not fully understood. Because of its frequent expression in tumors and its spontaneous immunogenicity, NY-ESO-1 is a prime target of cancer vaccines and an ideal model antigen for elucidating the molecular basis of immunodominant tumor-specific CD8(+) T cell responses. Here, we have assessed CD8(+)T cell responses to full-length NY-ESO-1 in cancer patients. We identified 3 immunodominant regions of the protein located within 3 distinct clusters of MHC class I binding sequences that co-localize with previously defined clusters of MHC class II binding sequences, are predicted to be hydrophobic and undergo efficient proteasomal processing. Our results support the concept that epitope clustering within defined protein regions identifies tumor antigen immunodominant regions and suggest a general strategy for their identification.     

Melanoma cell lines as the basis for antitumor vaccine preparation

The aim of the study was to obtain cell lines from tumor samples, and to determine phenotypic cell characteristics in order to choose the optimal line for vaccine preparation. 15 cell lines with stable growth, varying in cultural growth character and cytomorphology, were obtained from samples taken from patients with metastatic skin melanoma. Immunofluorescense method was used to determine the expression of T- and B-lymphocyte markers, antigens of major histocompatibility complex (MHC) class I and II, and CD86 co-stimulating molecule in the cell lines. The expression of melanocyte differentiation antigens and cancer/testicular antigens was evaluated using immunocytochemical assay. The results allowed the authors to distinguish three types of melanoma cell lines according to the expression of MHC molecules: MHC-negative; MHC class I positive; MHC classes I and II positive.     

Myoepithelial mRNA expression profiling reveals a common tumor-suppressor phenotype

A series of myoepithelial cell lines and xenografts derived from benign human myoepithelial tumors of diverse sources (salivary gland, breast, and lung) exhibit common mRNA expression profiles indicative of a tumor-suppressor phenotype. Previously established myoepithelial cell lines and xenografts (HMS-#; HMS-#X) were compared to nonmyoepithelial breast carcinoma cells (MDA-MB-231 and MDA-MB-468, and inflammatory breast carcinoma samples, IBCr, and IBCw), a normal mammary epithelial cell line (HMEC) and individual cases of human breast cancer (zcBT#T), and matched normal human breast tissues (zcBT#N) (overall samples = 22). The global gene expression profile (22,000 genes) of these individual samples was examined using Affymetrix Microarray Gene Chips and subsequently analyzed with both Affymetrix and DChip algorithms. The myoepithelial cell lines/xenografts were distinct and very different from the nonmyoepithelial breast carcinoma cells and the normal breast and breast tumor biopsies. Two hundred and seven specifically selected genes represented a subset of genes that distinguished (P < 0.05) all the myoepithelial cell lines/xenografts from all the other samples and which themselves exhibited hierarchical clustering. Further analysis of these genes revealed increased expression in genes belonging to the classes of extracellular matrix proteins, angiogenic inhibitors, and proteinase inhibitors and decreased expression belonging to the classes of angiogenic factors and proteinases. Developmental genes were also differentially expressed (either over or underexpressed). These studies confirm our previous impression that human myoepithelial cells express a distinct tumor-suppressor phenotype.