A wavelet-based optimal texture feature set for classification of brain tumours

In this work, the classification of brain tumours in magnetic resonance images is studied by using optimal texture features. These features are used to classify three sets of brain images—normal brain, benign tumour and malignant tumour. A wavelet-based texture feature set is derived from the region of interest. Each selected brain region of interest is characterized with both its energy and texture features extracted from the selected high frequency subband. An artificial neural network classifier is employed to evaluate the performance of these features. Feature selection is performed by a genetic algorithm. Principal component analysis and classical sequential methods are compared against the genetic approach in terms of the best recognition rate achieved and the optimal number of features. A classification performance of 98% is achieved in a genetic algorithm with only four of the available 29 features. Principal component analysis and classical sequential methods require a larger feature set to attain the similar classification accuracy of 98%. The optimal texture features such as range of angular second moment, range of sum variance, range of information measure of correlation II and energy selected by the genetic algorithm provide best classification performance with lower computational effort.     

Comprehensive microRNA profiling reveals a unique human embryonic stem cell signature dominated by a single seed sequence

Embryonic stem cells are unique among cultured cells in their ability to self-renew and differentiate into a wide diversity of cell types, suggesting that a specific molecular control network underlies these features. Human embryonic stem cells (hESCs) are known to have distinct mRNA expression, global DNA methylation, and chromatin profiles, but the involvement of high-level regulators, such as microRNAs (miRNA), in the hESC-specific molecular network is poorly understood. We report that global miRNA expression profiling of hESCs and a variety of stem cell and differentiated cell types using a novel microarray platform revealed a unique set of miRNAs differentially regulated in hESCs, including numerous miRNAs not previously linked to hESCs. These hESC-associated miRNAs were more likely to be located in large genomic clusters, and less likely to be located in introns of coding genes. hESCs had higher expression of oncogenic miRNAs and lower expression of tumor suppressor miRNAs than the other cell types. Many miRNAs upregulated in hESCs share a common consensus seed sequence, suggesting that there is cooperative regulation of a critical set of target miRNAs. We propose that miRNAs are coordinately controlled in hESCs, and are key regulators of pluripotence and differentiation. Disclosure of potential conflicts of interest is found at the end of this article.     

A wavelet-based optimal texture feature set for classification of brain tumours

In this work, the classification of brain tumours in magnetic resonance images is studied by using optimal texture features. These features are used to classify three sets of brain images - normal brain, benign tumour and malignant tumour. A wavelet-based texture feature set is derived from the region of interest. Each selected brain region of interest is characterized with both its energy and texture features extracted from the selected high frequency subband. An artificial neural network classifier is employed to evaluate the performance of these features. Feature selection is performed by a genetic algorithm. Principal component analysis and classical sequential methods are compared against the genetic approach in terms of the best recognition rate achieved and the optimal number of features. A classification performance of 98% is achieved in a genetic algorithm with only four of the available 29 features. Principal component analysis and classical sequential methods require a larger feature set to attain the similar classification accuracy of 98%. The optimal texture features such as range of angular second moment, range of sum variance, range of information measure of correlation II and energy selected by the genetic algorithm provide best classification performance with lower computational effort.     

Automatic detection of erthemato-squamous diseases using adaptive neuro- fuzzy inference systems

A new approach based on adaptive neuro-fuzzy inference system (ANFIS) was presented for detection of erythemato-squamous diseases. The domain contained records of patients with known diagnosis. Given a training set of such records, the ANFIS classifiers learned how to differentiate a new case in the domain. The six ANFIS classifiers were used to detect the six erythemato-squamous diseases when 34 features defining six disease indications were used as inputs. To improve diagnostic accuracy, the seventh ANFIS classifier (combining ANFIS) was trained using the outputs of the six ANFIS classifiers as input data. The proposed ANFIS model combined the neural network adaptive capabilities and the fuzzy logic qualitative approach. Some conclusions concerning the impacts of features on the detection of erythemato-squamous diseases were obtained through analysis of the ANFIS. The performances of the ANFIS model were evaluated in terms of training performances and classification accuracies and the results confirmed that the proposed ANFIS model has some potential in detecting the erythemato-squamous diseases. The ANFIS model achieved accuracy rates which were higher than that of the stand-alone neural network model.     

Comparison of Shallow and Deep Learning Methods on Classifying the Regional Pattern of Diffuse Lung Disease

This study aimed to compare shallow and deep learning of classifying the patterns of interstitial lung diseases (ILDs). Using high-resolution computed tomography images, two experienced radiologists marked 1200 regions of interest (ROIs), in which 600 ROIs were each acquired using a GE or Siemens scanner and each group of 600 ROIs consisted of 100 ROIs for subregions that included normal and five regional pulmonary disease patterns (ground-glass opacity, consolidation, reticular opacity, emphysema, and honeycombing). We employed the convolution neural network (CNN) with six learnable layers that consisted of four convolution layers and two fully connected layers. The classification results were compared with the results classified by a shallow learning of a support vector machine (SVM). The CNN classifier showed significantly better performance for accuracy compared with that of the SVM classifier by 6–9%. As the convolution layer increases, the classification accuracy of the CNN showed better performance from 81.27 to 95.12%. Especially in the cases showing pathological ambiguity such as between normal and emphysema cases or between honeycombing and reticular opacity cases, the increment of the convolution layer greatly drops the misclassification rate between each case. Conclusively, the CNN classifier showed significantly greater accuracy than the SVM classifier, and the results implied structural characteristics that are inherent to the specific ILD patterns.     

基于双通道神经网络的疾病文本分类方法

医疗疾病文本的准确分类对医疗信息化的发展具有重要的推进作用,本研究提出一种基于双通道学习的神经网络模型研究疾病文本分类方法。该模型分别使用卷积神经网络和双向长短期记忆网络对患者输入的疾病症状文本进行局部特征以及时序特征学习。此外,在双向长短期记忆网络上引入自注意力机制区分特征对类别预测的贡献值,增强模型的学习能力和可解释性。为使两个通道提取到的特征能够共同决定分类结果,该模型将两种特征进行拼接融合,最后利用softmax分类器得到最终的分类结果。实验结果表明,在疾病文本分类的性能方面,该模型相比其他分类模型具有较高的精确率、召回率和F1值,分别可达90.61%、90.48%和90.51%。     

MicroRNAs in gene regulation: when the smallest governs it all

Encoded by the genome of most eukaryotes examined so far, microRNAs (miRNAs) are small approximately 21-nucleotide (nt) noncoding RNAs (ncRNAs) derived from a biosynthetic cascade involving sequential processing steps executed by the ribonucleases (RNases) III Drosha and Dicer. Following their recent identification, miRNAs have rapidly taken the center stage as key regulators of gene expression. In this review, we will summarize our current knowledge of the miRNA biosynthetic pathway and its protein components, as well as the processes it regulates via miRNAs, which are known to exert a variety of biological functions in eukaryotes. Although the relative importance of miRNAs remains to be fully appreciated, deregulated protein expression resulting from either dysfunctional miRNA biogenesis or abnormal miRNA-based gene regulation may represent a key etiologic factor in several, as yet unidentified, diseases. Hence is our need to better understand the complexity of the basic mechanisms underlying miRNA biogenesis and function.     

MicroRNAs: new regulators of immune cell development and function

Decades of research went into understanding immune cell development and function without awareness that consideration of a key element, microRNA (miRNA), was lacking. The discovery of miRNAs as regulators of developmental events in model organisms suggested to many investigators that miRNA might be involved in the immune system. In the past few years, widespread examination of this possibility has produced notable results. Results have shown that miRNAs affect mammalian immune cell differentiation, the outcome of immune responses to infection and the development of diseases of immunological origin. Some miRNAs repress expression of target proteins with well established functions in hematopoiesis. Here we bring together much of this work, which has so far only scratched the surface of this very fertile field of investigation, and show how the results illuminate many historic questions about hematopoiesis and immune function.     

Integrated analysis of the molecular mechanisms in idiopathic pulmonary fibrosis

Rationale: Idiopathic pulmonary fibrosis (IPF) is one of the most aggressive forms of idiopathic interstitial pneumonia. Some miRNAs may be associated with IPF and may affect the occurrence and development of IPF in various pathways. Many miRNAs and genes that may be involved in the development of IPF have been discovered using chip and high throughput technologies.Methods: We analyzed one miRNA and four mRNA databases. We identified hub genes and pathways related to IPF using GO, KEGG enrichment analysis, gene set variation analysis (GSVA), PPI network construction, and hub gene analysis. A comprehensive analysis of differentially expressed miRNAs (DEMs), predicted miRNA target genes, and differentially expressed genes (DEGs) led to the creation of a miRNA-mRNA regulatory network in IPF.Results: We found 203 DEGs and 165 DEMs that were associated with IPF. The findings of enrichment analyses showed that these DEGs were mainly involved in antimicrobial humoral response, antimicrobial humoral immune response mediated by antimicrobial peptide, extracellular matrix organization, cell killing, and organ or tissue specific immune response. The VEGFA, CDH5, and WNT3A genes overlapped between hub genes and the miRNA-mRNA regulatory network. The miRNAs including miR-199b-5p, miR-140-5p, miR-199a-5p, miR-125A-5p, and miR-107 that we predicted would regulate the VEGFA, CDH5, and WNT3A genes, which were also associated with IPF or other fibrosis-related diseases. GSVA indicated that metabolic processes of UTP and IMP, immune response, regulation of Th2 cell cytokine production, and positive regulation of NK cell-mediated immunity are associated with the pathogenesis and treatment of IPF. These pathways also interact with VEGFA, CDH5, and WNT3A.Conclusion: These findings provide a new research direction for the diagnosis and treatment of IPF.     

Not miR-ly small RNAs: big potential for microRNAs in therapy

RNA interference (RNAi) describes a set of natural processes in which genes are silenced by small RNAs. RNAi has been widely used as an experimental tool that has recently become the focus of drug development efforts to treat a variety of diseases and disorders. Like all molecular therapies, in vivo delivery is the major hurdle to realizing therapeutic RNAi. Several strategies have been developed that increase small RNA half-life in the blood, facilitate transduction across biological membranes, and mediate cell-specific delivery. Importantly, these strategies permit targeting of mRNAs as well as microRNAs (miRNAs), a class of small RNAs encoded in the genome. miRNAs are required for multiple developmental and cellular processes. Dysfunction of miRNAs can result in a host of pathologies, suggesting that miRNAs are potential targets of therapy. Recent studies of miRNA function in immune-specific pathways indicate that specific miRNAs might be exploited as therapeutic targets to treat immune disorders, including autoimmunity, allergy, and hematopoietic cancers.     

MicroRNAs grow up in the immune system

MicroRNA (miRNA) target predictions support a view in which each miRNA regulates translation and stability of several hundred messenger RNAs (mRNAs). Studies that overexpress individual miRNAs typically uncover relative subtle inhibition of the predicted targets. Accordingly, most miRNAs expressed in a given cell type may serve the function to broadly inhibit cell-type-inappropriate gene expression and deepen a pre-existing differentiation program. However, recent functional analyses of miRNAs in the immune system reveal that many cellular decisions are controlled by single miRNAs that entail significant downregulation of one or few target proteins. Investigations of these miRNA/mRNA pairs showed that miRNA-adjusted target protein levels are crucial at specific cellular transition points. Here, we will review recent advances in the regulation of the miRNA pathway and discuss how miRNAs control immune functions.     

Bispectral analysis and genetic algorithm for congestive heart failure recognition based on heart rate variability

This paper proposes a congestive heart failure (CHF) recognition method that includes features calculated from the bispectrum of heart rate variability (HRV) diagrams and a genetic algorithm (GA) for feature selection. The roles of the bispectrum-related features and the GA feature selector are investigated. Features calculated from the subband regions of the HRV bispectrum are added into a feature set containing only regular time-domain and frequency-domain features. A support vector machine (SVM) is employed as the classifier. A feature selector based on genetic algorithm proceeds to select the most effective features for the classifier. The results confirm the effectiveness of including bispectrum-related features for promoting the discrimination power of the classifier. When compared with the other two methods in the literature, the proposed method (without GA) outperforms both of them with a high accuracy of 96.38%. More than 3.14% surpluses in accuracies are observed. The application of GA as a feature selector further elevates the recognition accuracy from 96.38% to 98.79%. When compared to the Isler and Kuntalp's impressive results recently published in the literature that also uses GA for feature selection, the proposed method (with GA) outperforms them with more than 2.4% surpass in the recognition accuracy. These results confirm the significance of recruiting bispectrum-related features in a CHF classification system. Moreover, the application of GA as feature selector can further improve the performance of the classifier.     

基于极限梯度提升和深度神经网络共同决策的心音分类方法

心音是诊断心血管疾病常用的医学信号之一。本文对心音正常/异常的二分类问题进行了研究,提出了一种基于极限梯度提升(XGBoost)和深度神经网络共同决策的心音分类算法,实现了对特征的选择和模型准确率的进一步提升。首先,本文对预处理后的心音信号进行心音分割,在此基础上提取了5个大类的特征,前4类特征采用递归特征消除法进行特征选择,作为XGBoost分类器的输入,最后一类为梅尔频率倒谱系数(MFCC),作为长短时记忆网络(LSTM)的输入。考虑到数据集的不平衡性,本文在两种分类器中皆使用了加权改进的方法。最后采用异质集成决策方法得到预测结果。将本文所提心音分类算法应用于PhysioNet网站在2016年发起的PhysioNet心脏病学挑战赛(CINC)所用公开心音数据库,以测试灵敏度、特异性、修正后的准确率以及F得分,结果分别为93%、89.4%、91.2%、91.3%,通过与其他研究者应用机器学习、卷积神经网络(CNN)等方法的结果比较,在准确率和灵敏度上有明显提高,证明了本文方法能有效地提高心音信号分类的准确性,在部分心血管疾病的临床辅助诊断应用中有很大的潜力。     

microRNAs与骨骼肌疾病

MicroRNA(miRN A)是一类长度约为22个核苷酸的非编码单链RNA分子,它能与mRNA的特定位点结合,参与转录后基因表达调控,从而抑制蛋白质的合成。因此,它在调节基因转录与表达,调控生物体正常发育等各个生理过程中扮演重要角色,同时也对人类疾病的防治以及生物进化探索有着重要意义。有关miRNAs和骨骼肌之间的研究主要集中在miRNAs与骨骼肌生长发育、骨骼肌损伤、骨骼肌疾病。一些普遍表达的miRNAs在胚胎骨骼肌发育和成体骨骼肌再生过程中起着促进作用,且miRNAs调节异常是骨骼肌疾病的一个共同特征。本文主要就miRNAs的生物合成过程和生物学特性,miRNAs在骨骼肌发育过程中的主要功能,以及miRNA对骨骼肌相关疾病的影响和作用作一综述。     

miRNAs can be generally associated with human pathologies as exemplified for miR-144*

Background

miRNA profiles are promising biomarker candidates for a manifold of human pathologies, opening new avenues for diagnosis and prognosis. Beyond studies that describe miRNAs frequently as markers for specific traits, we asked whether a general pattern for miRNAs across many diseases exists.

Methods

We evaluated genome-wide circulating profiles of 1,049 patients suffering from 19 different cancer and non-cancer diseases as well as unaffected controls. The results were validated on 319 individuals using qRT-PCR.

Results

We discovered 34 miRNAs with strong disease association. Among those, we found substantially decreased levels of hsa-miR-144* and hsa-miR-20b with AUC of 0.751 (95% CI: 0.703–0.799), respectively. We also discovered a set of miRNAs, including hsa-miR-155*, as rather stable markers, offering reasonable control miRNAs for future studies. The strong downregulation of hsa-miR-144* and the less variable pattern of hsa-miR-155* has been validated in a cohort of 319 samples in three different centers. Here, breast cancer as an additional disease phenotype not included in the screening phase has been included as the 20^th trait.

Conclusions

Our study on 1,368 patients including 1,049 genome-wide miRNA profiles and 319 qRT-PCR validations further underscores the high potential of specific blood-borne miRNA patterns as molecular biomarkers. Importantly, we highlight 34 miRNAs that are generally dysregulated in human pathologies. Although these markers are not specific to certain diseases they may add to the diagnosis in combination with other markers, building a specific signature. Besides these dysregulated miRNAs, we propose a set of constant miRNAs that may be used as control markers.