太赫兹技术在医学成像方面的应用和进展

近年来太赫兹波(terahenz radiation THz),作为一种无创性的显像源以及在显像方面潜在的优势,逐渐被应用于医学影像中。本文就太赫兹显像的原理、医学显像方面的应用以及其在分子影像学方面的进展进行综述。     

基于特征向量法和支持向量机的抑郁症脑电信号分类

目的:从抑郁症患者EEG信号中提取与疾病相关的信息以实现对抑郁症患者与健康人的自动分类.方法:用特征向量法对抑郁症患者与健康人脑电进行特征提取,得到脑电信号功率谱幅度的最大值、最小值、平均值和标准偏差等特征参数,然后用支持向量机分类器进行训练和分类,并进行测试验证.结果:相对于用小波变换提取的频率相关参数作为分类特征,采用本文特征向量法功率谱估计提取的特征参数为分类特征的分类器具有更好的分类效果,其抑郁症患者和健康人脑电信号的分类准确率可以达到95.6%.结论:该研究成果为抑郁症疾病的物理诊断提供了一种新的途径.     

基于修正四阶Cole-Cole模型的太赫兹波段生物组织介电特性估算方法探索

为了获得生物组织在太赫兹波段的相对介电常数和电导率,在四阶Cole-Cole模型基础上进行修正,使修正后的模型不仅能描述生物组织在10 Hz~100 GHz频段的介电特性,还能描述生物组织在太赫兹波段的介电特性。在四阶Cole-Cole模型上新增了一项ωk/i,对正常皮肤组织在10 Hz~20 GHz和0.15~1.95 THz频段的介电特性的实验数据按照文献方法进行处理,提取四阶Cole-Cole修正模型的参数。在10 Hz~100 GHz频段内,原有模型和修正模型计算值的相对误差在5%以内,并且在太赫兹波段正常皮肤组织修正模型的相对介电常数和电导率的计算值与实验数据的相对误差基本在10%以内,结果验证了四阶Cole-Cole修正模型的正确性。将该修正模型应用在脑白质和脑灰质太赫兹波段的介电特性的计算中,得到脑白质和脑灰质在0.15~1.95 THz频段的相对介电常数和电导率。所提出的四阶Cole-Cole修正模型还可以对其他生物组织在太赫兹波段的介电特性进行预测。     

基于加权自相关函数特征提取法的多类蛋白质同源寡聚体分类研究

我们提出一种新的特征提取方法,即用蛋白质序列的氨基酸组成成分和一系列的氨基酸残基指数加权自相关函数构成特征向量,表示蛋白质序列,与支持向量机算法组合对蛋白质同源二聚体、同源三聚体、同源四聚体、同源六聚体进行分类研究,得到较好的分类结果。在Jackknife检验下,采用支持向量机算法,基于此新特征提取法所构成的参数集QIANA、QIANB、MEEJ、ROBB和SNEP的总分类精度分别为77.63%、77.16%、76.46%、76.70%、75.06%,分别比传统氨基酸组成成分特征提取法(参数集为COMP)提高6.39、5.92、5.22、5.46、3.82个百分点。对于参数集QIANA,支持向量机的总分类精度为77.63%,比协方差算法提高16.29个百分点。这些结果表明:(1新特征提取法是有效和可行的,基于此特征提取法构成的特征向量包含蛋白质四级结构信息,且可能捕获了埋藏在缔合亚基作用部位接触表面的基本信息;(2)对于蛋白质同源寡聚体分类研究,支持向量机是非常有效的。     

基于支持向量机的中医望诊颧色分类研究

目的在中医望诊理论中,人面部各区域颜色能反映人体健康状况。针对传统的颧色判别方法主要依靠医生目视判断,本文提出一种基于支持向量机的颧色自动分类方法。方法首先根据中医专家经验分别选取训练集中颧红和颧非红图像中的8×8像素块作为训练样本,将测试图像划分成8×8像素块作为测试样本,然后提取每块(包括训练样本和测试样本)中64个像素的R、G、B值作为特征,使用训练后的支持向量机分类器将测试样本中的块分类为红块和非红块。最后根据每幅两颧图像中被判为红色的块所占的比例来对该图像进行分类,若两颧区域中所占的红块的比例超过预设的阈值,则将其判为两颧红,否则判为两颧非红。结果在专家鉴别的图像库上进行了测试,该算法对颧色分类的准确率接近83%。结论基于支持向量机的颧色分类方法能取得较好的分类效果。     

基于SARS RNA序列可视化方法的进一步分析与研究

我们提出了一种新的基因可视化分析方法。该可视化方法基于元胞自动机(CA)原理,将RNA序列的一维信息转化为二维可视图谱;在将此方法应用到SARS RNA序列的分析中,可以发现SARS-CoV病毒不同于其他非SARS的冠状病毒的RNA序列可视特征;在此基础上,对得到的RNA特征序列进行片断提取,用片断可视化取代全序列可视化,并用主成分分析(PCA),支持向量机(SVM)等算法对RNA片断进行分析。结果证明,SARS-CoV的该类特征具有很好的模式可分性,与模式分析方法的结合则可以作为可视化的佐证,可以更充分地利用特征片断作为判别SARS-CoV的一种非常规的手段。     

基于遗传算法与支持向量机的基因微阵列分析

遗传算法和支持向量机是近年来发展迅速的机器学习算法,对样本量较小而变量数较大的基因微阵列数据支持向量机具有很好的分类效果。而遗传算法通过初始种群的不断进化（交叉,变异和选择）,从而收敛到最优解,达到降维的目的。本文将二者结合,采用遗传算法并以支持向量机的分类准确率作为适应度函数,提高分类准确度。结果显示这种方法对分类更加有效。本文同时也对特征基因在代谢通路上的分布和功能作了一定的研究。     

基于主成份分析和支持向量机的MRI图像多目标分割

在MRI图像中,颅内各组织的边界极其复杂且不规则,这对传统的分割算法提出严峻的挑战.主成份分析（PCA）可达到降维和消除冗余信息的目的,为使支持向量机（SVM）准备的样本空间更为紧凑、合理.本研究采用PCA将图像的57维特征向量处理后,研究多分类SVM对MRI图像进行多目标分割,成功提取颅内各组织不规则边界.理论和实验表明,基于PCA和SVM相结合的分割性能优于仅采用SVM的分割性能.     

基于全脑功能连接的网络游戏障碍分类研究

目的：利用静息态磁共振数据构建全脑功能连接网络,通过多元模式分析建立诊断模型,实现网络游戏障碍(internet gaming disorder, IGD)和正常对照组之间的分类识别。方法：采集71例IGD患者及88例正常对照的大脑静息态磁共振数据,采用功能连接分析技术构建全脑功能连接网络。将大脑功能连接作为分类特征,采用支持向量机和多种特征选择方法,探索IGD患者区别于正常人的异常网络模式。综合多种特征选择方法筛选的共有特征最终确定IGD客观识别的影像学标志。结果：基于全脑功能连接建立的分类模型准确率最高可达80.6%（敏感性为78.5%,特异性为81.2%）。用于区分IGD患者和正常对照的神经影像学标记主要位于左侧背外侧前额叶、右侧前扣带回、左侧眶额回、右侧海马旁回和双侧颞叶等负责认知控制、动机和学习记忆的脑区。结论：基于静息态全脑功能连接的诊断模型对IGD有较好的区分能力,未来可以为临床智能诊断提供补充手段。     

基于极限学习机的跌倒检测分类识别研究

由于年龄和身体条件的限制,在老年人群中跌倒是非常普遍的现象。因此,根据老年人跌倒的运动特征,远程监测他们在各个时间段的状态,以便在其摔倒或突发状况时及时采取措施显得尤为重要。针对人体运动状态进行监测,分析人体运动学特征,提出基于极限学习机的跌倒检测算法。运用三维加速度传感器采集人体的三维加速度值,建立跌倒检测特征模型。在此基础上,建立基于极限学习机的跌倒检测分类器,完成对老年人的计算机辅助跌倒检测。实验数据共540例样本,选用了不同数量的训练集和测试集,其中440例作为训练数据,其余100例为测试数据。测试结果表明,准确率为93%,敏感度为87.5%,特异性为91.7%,具有良好的分类性能。在对分类训练的运行时间方面,基于极限学习机的跌倒检测方法与传统的机器学习方法相比具有明显优势。     

Combining scanning haptic microscopy and fibre optic Raman spectroscopy for tissue characterization

The tactile resonance method (TRM) and Raman spectroscopy (RS) are promising for tissue characterization in vivo. Our goal is to combine these techniques into one instrument, to use TRM for swift scanning, and RS for increasing the diagnostic power. The aim of this study was to determine the classification accuracy, using support vector machines, for measurements on porcine tissue and also produce preliminary data on human prostate tissue. This was done by developing a new experimental set-up combining micro-scale TRM—scanning haptic microscopy (SHM)—for assessing stiffness on a micro-scale, with fibre optic RS measurements for assessing biochemical content. We compared the accuracy using SHM alone versus SHM combined with RS, for different degrees of tissue homogeneity. The cross-validation classification accuracy for healthy porcine tissue types using SHM alone was 65–81%, and when RS was added it increased to 81–87%. The accuracy for healthy and cancerous human tissue was 67–70% when only SHM was used, and increased to 72–77% for the combined measurements. This shows that the potential for swift and accurate classification of healthy and cancerous prostate tissue is high. This is promising for developing a tool for probing the surgical margins during prostate cancer surgery.     

Support vector methods for survival analysis: a comparison between ranking and regression approaches

Objective

To compare and evaluate ranking, regression and combined machine learning approaches for the analysis of survival data.

Methods

The literature describes two approaches based on support vector machines to deal with censored observations. In the first approach the key idea is to rephrase the task as a ranking problem via the concordance index, a problem which can be solved efficiently in a context of structural risk minimization and convex optimization techniques. In a second approach, one uses a regression approach, dealing with censoring by means of inequality constraints. The goal of this paper is then twofold: (i) introducing a new model combining the ranking and regression strategy, which retains the link with existing survival models such as the proportional hazards model via transformation models; and (ii) comparison of the three techniques on 6 clinical and 3 high-dimensional datasets and discussing the relevance of these techniques over classical approaches fur survival data.

Results

We compare svm-based survival models based on ranking constraints, based on regression constraints and models based on both ranking and regression constraints. The performance of the models is compared by means of three different measures: (i) the concordance index, measuring the model’s discriminating ability; (ii) the logrank test statistic, indicating whether patients with a prognostic index lower than the median prognostic index have a significant different survival than patients with a prognostic index higher than the median; and (iii) the hazard ratio after normalization to restrict the prognostic index between 0 and 1. Our results indicate a significantly better performance for models including regression constraints above models only based on ranking constraints.

Conclusions

This work gives empirical evidence that svm-based models using regression constraints perform significantly better than svm-based models based on ranking constraints. Our experiments show a comparable performance for methods including only regression or both regression and ranking constraints on clinical data. On high dimensional data, the former model performs better. However, this approach does not have a theoretical link with standard statistical models for survival data. This link can be made by means of transformation models when ranking constraints are included.     

Supervised machine learning-based classification of oral malodor based on the microbiota in saliva samples

ObjectiveThis study presents an effective method of classifying oral malodor from oral microbiota in saliva by using a support vector machine (SVM), an artificial neural network (ANN), and a decision tree. This approach uses concentrations of methyl mercaptan in mouth air as an indicator of oral malodor, and peak areas of terminal restriction fragment (T-RF) length polymorphisms (T-RFLPs) of the 16S rRNA gene as data for supervised machine-learning methods, without identifying specific species producing oral malodorous compounds.Methods16S rRNA genes were amplified from saliva samples from 309 subjects, and T-RFLP analysis was carried out with the DNA fragments. T-RFLP analysis provides information on microbiota consisting of fragment lengths and peak areas corresponding to bacterial strains. The peak area is equivalent to the frequency of a specific fragment when one molecule is selected from terminal fragments. Another frequency is obtained by dividing the number of species-containing samples by the total number of samples. An SVM, an ANN, and a decision tree were trained based on these two frequencies in 308 samples and classified the presence or absence of methyl mercaptan in mouth air from the remaining subject.ResultsThe proportion that trained SVM expressed as entropy achieved the highest classification accuracy, with a sensitivity of 51.1% and specificity of 95.0%. The ANN and decision tree provided lower classification accuracies, and only classification by the ANN was improved by weighting with entropy from the frequency of appearance in samples, which increased the accuracy to 81.9% with a sensitivity of 60.2% and a specificity of 90.5%. The decision tree showed low classification accuracy under all conditions.ConclusionsUsing T-RF proportions and frequencies, models to classify the presence of methyl mercaptan, a volatile sulfur-containing compound that causes oral malodor, were developed. SVM classifiers successfully classified the presence of methyl mercaptan with high specificity, and this classification is expected to be useful for screening saliva for oral malodor before visits to specialist clinics. Classification by a SVM and an ANN does not require the identification of the oral microbiota species responsible for the malodor, and the ANN also does not require the proportions of T-RFs.     

Analysis of SPECT brain images for the diagnosis of Alzheimer's disease based on NMF for feature extraction

This letter presents a novel computer-aided diagnosis (CAD) technique for the early diagnosis of Alzheimer's disease (AD) based on non-negative matrix factorization (NMF) analysis applied to single photon emission computed tomography (SPECT) images. A baseline normalized SPECT database containing normalized data for both AD patients and healthy reference patients is selected for this study. The SPECT database is analyzed by applying the Fisher discriminant ratio (FDR) for feature selection and NMF for feature extraction of relevant components of each subject. The main goal of these preprocessing steps is to reduce the large dimensionality of the input data and to relieve the so called “curse of dimensionality” problem. The resulting NMF-transformed set of data, which contains a reduced number of features, is classified by means of a support vector machines based classification technique (SVM). The proposed NMF + SVM method yields up to 94% classification accuracy, with high sensitivity and specificity values (upper than 90%), becoming an accurate method for SPECT image classification. For the sake of completeness, comparison between another recently developed principal component analysis (PCA) plus SVM method and the proposed method is also provided, yielding results for the NMF + SVM approach that outperform the behavior of the reference PCA + SVM or conventional voxel-as-feature (VAF) plus SVM methods.     

Robust prediction of protein subcellular localization combining PCA and WSVMs

Automated prediction of protein subcellular localization is an important tool for genome annotation and drug discovery, and Support Vector Machines (SVMs) can effectively solve this problem in a supervised manner. However, the datasets obtained from real experiments are likely to contain outliers or noises, which can lead to poor generalization ability and classification accuracy. To explore this problem, we adopt strategies to lower the effect of outliers. First we design a method based on Weighted SVMs, different weights are assigned to different data points, so the training algorithm will learn the decision boundary according to the relative importance of the data points. Second we analyse the influence of Principal Component Analysis (PCA) on WSVM classification, propose a hybrid classifier combining merits of both PCA and WSVM. After performing dimension reduction operations on the datasets, kernel-based possibilistic c-means algorithm can generate more suitable weights for the training, as PCA transforms the data into a new coordinate system with largest variances affected greatly by the outliers. Experiments on benchmark datasets show promising results, which confirms the effectiveness of the proposed method in terms of prediction accuracy.     

基于SVM的便携式睡眠监测系统设计

目的睡眠监测是睡眠质量分析中重要的环节,但目前的睡眠监测系统复杂而且难以携带。本文提出基于支持向量机的便携式睡眠监测系统,以方便地实时监控睡眠。方法该系统硬件部分由服务器和用户端设备构成,其中用户端设备负责数据采集和数据传输,服务器端负责数据分析及相关的资源管理。睡眠分析软件采用支持向量机(support vector machines,SVM)作为分析算法,在提取特征值的基础上,以有向无环图作为多分类策略分析得到睡眠的时相。结果对于患者的睡眠脑电实验表明分析正确率高,所需的分析时间短。结论该系统用户端设备体积小,方便携带,分析正确率高,实时性好,在睡眠监测领域具有良好的应用前景。     

Electrocardiogram analysis using a combination of statistical, geometric, and nonlinear heart rate variability features

Objective

The paper addresses a common and recurring problem of electrocardiogram (ECG) classification based on heart rate variability (HRV) analysis. Current understanding of the limits of HRV analysis in diagnosing different cardiac conditions is not complete. Existing research suggests that a combination of carefully selected linear and nonlinear HRV features should significantly improve the accuracy for both binary and multiclass classification problems. The primary goal of this work is to evaluate a proposed combination of HRV features. Other explored objectives are the comparison of different machine learning algorithms in the HRV analysis and the inspection of the most suitable period T between two consecutively analyzed R-R intervals for nonlinear features.

Methods and material

We extracted 11 features from 5 min of R-R interval recordings: SDNN, RMSSD, pNN20, HRV triangular index (HTI), spatial filling index (SFI), correlation dimension, central tendency measure (CTM), and four approximate entropy features (ApEn1-ApEn4). Analyzed heart conditions included normal heart rhythm, arrhythmia (any), supraventricular arrhythmia, and congestive heart failure. One hundred patient records from six online databases were analyzed, 25 for each condition. Feature vectors were extracted by a platform designed for this purpose, named ECG Chaos Extractor. The vectors were then analyzed by seven clustering and classification algorithms in the Weka system: K-means, expectation maximization (EM), C4.5 decision tree, Bayesian network, artificial neural network (ANN), support vector machines (SVM) and random forest (RF). Four-class and two-class (normal vs. abnormal) classification was performed. Relevance of particular features was evaluated using 1-Rule and C4.5 decision tree in the cases of individual features classification and classification with features’ pairs.

Results

Average total classification accuracy obtained for top three classification methods in the two classes’ case was: RF 99.7%, ANN 99.1%, SVM 98.9%. In the four classes’ case the best results were: RF 99.6%, Bayesian network 99.4%, SVM 98.4%. The best overall method was RF. C4.5 decision tree was successful in the construction of useful classification rules for the two classes’ case. EM and K-means showed comparable clustering results: around 50% for the four classes’ case and around 75% for the two classes’ case. HTI, pNN20, RMSSD, ApEn3, ApEn4 and SFI were shown to be the most relevant features. HTI in particular appears in most of the top-ranked pairs of features and is the best analyzed feature. The choice of the period T for nonlinear features was shown to be arbitrary. However, a combination of five different periods significantly improved classification accuracy, from 70% for a single period up to 99% for five periods.

Conclusions

Analysis shows that the proposed combination of 11 linear and nonlinear HRV features gives high classification accuracy when nonlinear features are extracted for five periods. The features’ combination was thoroughly analyzed using several machine learning algorithms. In particular, RF algorithm proved to be highly efficient and accurate in both binary and multiclass classification of HRV records. Interpretable and useful rules were obtained with C4.5 decision tree. Further work in this area should elucidate which features should be extracted for the best classification results for specific types of cardiac disorders.     

基于光腔衰荡光谱的呼吸丙酮分析仪设计与实现

目的 利用光腔衰荡光谱（CRDS）搭建了一套呼吸丙酮分析仪,以研究人体呼吸中的糖尿病潜在生物标志物丙酮,推动呼吸分析在临床疾病诊断或代谢监测中的应用.方法 选择266 nm紫外脉冲激光器作为光源,反射率为99.956％的高反镜组成衰荡腔,以光电倍增管（PMT）为探测器,在尺寸为60 cm×20 cm的铝板上搭建了一套可移动式呼吸分析仪.用不同体积分数的标准丙酮气体验证仪器的准确度与线性响应后,将其用于健康人体与糖尿病患者的呼吸样本测量.结果 实验测得高反射镜等效反射率为99.93％,仪器典型的衰荡时间基线平均值为2.386 4μs,稳定度为0.22％,对不同体积分数的丙酮样本具有线性响应（R2=0.99）,仪器检测极限为0.13×10-6（3σ准则）.结论 该呼吸分析仪具有良好的稳定性与重复性,可满足临床上人体呼吸的测量要求,并即将用于大量的临床呼吸测试以研究呼吸标志物与疾病之间的相关性.     

Identification of potential drugs for diffuse large b-cell lymphoma based on bioinformatics and Connectivity Map database

Diffuse large B-cell lymphoma (DLBCL) is the most main subtype in non-Hodgkin lymphoma. After chemotherapy, about 30% of patients with DLBCL develop resistance and relapse. This study was to identify potential therapeutic drugs for DLBCL using the bioinformatics method. The differentially expressed genes (DEGs) between DLBCL and non-cancer samples were downloaded from the Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO). Gene ontology enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of DEGs were analyzed using the Database for Annotation, Visualization, and Integrated Discovery. The R software package (SubpathwayMiner) was used to perform pathway analysis on DEGs affected by drugs found in the Connectivity Map (CMap) database. Protein–protein interaction (PPI) networks of DEGs were constructed using the Search Tool for the Retrieval of Interacting Genes online database and Cytoscape software. In order to identify potential novel drugs for DLBCL, the DLBCL-related pathways and drug-affected pathways were integrated. The results showed that 1927 DEGs were identified from TCGA and GEO. We found 54 significant pathways of DLBCL using KEGG pathway analysis. By integrating pathways, we identified five overlapping pathways and 47 drugs that affected these pathways. The PPI network analysis results showed that the CDK2 is closely associated with three overlapping pathways (cell cycle, p53 signaling pathway, and small cell lung cancer). The further literature verification results showed that etoposide, rinotecan, methotrexate, resveratrol, and irinotecan have been used as classic clinical drugs for DLBCL. Anisomycin, naproxen, gossypol, vorinostat, emetine, mycophenolic acid and daunorubicin also act on DLBCL. It was found through bioinformatics analysis that paclitaxel in the drug-pathway network can be used as a potential novel drug for DLBCL.