Deep Convolution Neural Network Based System for Early Diagnosis of Alzheimer's Disease

ObjectivesAlzheimer's Disease (AD) is the most general type of dementia. In all leading countries, it is one of the primary reasons of death in senior citizens. Currently, it is diagnosed by calculating the MSME score and by the manual study of MRI Scan. Also, different machine learning methods are utilized for automatic diagnosis but existing has some limitations in terms of accuracy. So, main objective of this paper to include a preprocessing method before CNN model to increase the accuracy of classification.Materials and methodIn this paper, we present a deep learning-based approach for detection of Alzheimer's Disease from ADNI database of Alzheimer's disease patients, the dataset contains fMRI and PET images of Alzheimer's patients along with normal person's image. We have applied 3D to 2D conversion and resizing of images before applying VGG-16 architecture of Convolution neural network for feature extraction. Finally, for classification SVM, Linear Discriminate, K means clustering, and Decision tree classifiers are used.ResultsThe experimental result shows that the average accuracy of 99.95% is achieved for the classification of the fMRI dataset, while the average accuracy of 73.46% is achieved with the PET dataset. On comparing results on the basis of accuracy, specificity, sensitivity and on some other parameters we found that these results are better than existing methods.Conclusionsthis paper, suggested a unique way to increase the performance of CNN models by applying some preprocessing on image dataset before sending to CNN architecture for feature extraction. We applied this method on ADNI database and on comparing the accuracies with other similar approaches it shows better results.     

Cluster analysis of pressure pain threshold maps from the trapezius muscle

The aim of this study was to investigate and present a new mapping method to describe muscle pain sensitivity based on the assessment of pressure pain threshold (PPT) over the trapezius muscle. PPT data were recorded from 36 points in 20 healthy males using a standardised grid. Points were clustered using the K-means algorithm with a fixed initialisation procedure. The total number of clusters was determined on the basis of (1) R ² evaluation of the clustering outcome compared against a desired 95% reduction in variance criterion and (2) the number of empty clusters. A minimum of three clusters were found which fulfilled the criteria. The proposed method enables the identification of a relation between the muscle subdivisions and pressure pain sensitivity within the trapezius. Further, the cluster analysis will enable the study of differences in pain sensitivity distributions between patients and controls and quantify the effect of intervention (physical or pharmacological treatments).     

An effective method for accurate nymphal-stage delimitation of the cicada Hyalessa fuscata

Subterranean nymphal development in cicadas presents challenges to researchers in accurately estimating the number of their developmental stages, although such information is crucial to understanding and predicting their population dynamics. While most studies have relied on head width as an attribute for life-stage determination to date, such character in cicadas can be highly variable and thus differentiation solely based on such morphology is prone to subjectivity in practice. Here, we propose a reliable method for instar estimation that is applicable to Hyalessa fuscata nymphs. We first obtained morphometrics of nymphs in all stages. Second, we computed logarithm-transformation and principal component analysis to extract a transformed variable that captures most of the variance of morphological characteristics. Third, k-means were computed to divide the dataset into distinct clusters assuming four-, five- and six life-stage scenarios for the best interferences of life stages. Finally, simple linear regression analysis was conducted to compare and select the best fit model. Our result shows that five nymphal stages best fit for H. fuscata nymphs. This method is expected to provide an easy-to-handle ecological tool for the study of life history of cicadas as well as other insects that have long life cycles and multiple developmental stages.     

森林转基因林木抽样策略的研究

转基因技术加快了林木遗传改良的进程。由于转基因林木存在潜在的生态风险,有必要对森林中转基因林木的分布等情况进行监测。基于无人机(unmanned aerial vehicles,UAV)获取的森林平面影像,提出了真实地形下森林转基因林木抽样监测的策略。首先根据检验参数来确定需要抽样的样本数,然后改进了两种算法分别进行不规则区域内常规的和有孔洞情况下的均匀抽样。与之前的抽样方法相比,本方法提供了更好的抽样准确性和均匀性,更适应森林区域的实际情况。     

Clustering hybrid regression: a novel computational approach to study and model biohydrogen production through dark fermentation

Clustering hybrid regression (CHR) approach was developed and evaluated using data from H(2)-producing glucose-based, suspended-cell bioreactor operated for 5 months. The aim was to describe the relationship between metabolic end products and H(2)-production rate. Self-organizing maps (SOM) were used to better visualize the dataset and to detect main metabolic patterns in bioprocess data. SOM detected three distinct metabolic patterns with butyrate, acetate and ethanol as dominant metabolites, respectively. Butyrate dominated metabolism was related to high H(2) production, while acetate and ethanol dominated metabolisms resulted in low H(2) production. CHR models performed well [mean square error (MSE) 0.55 and 0.56] in modeling the H(2)-production rate. The results validate the suitability of the CHR approach in describing the bioprocess behavior and in the modeling of H(2) production rate. The developed model can help in discovering key metabolic interactions and suitable process parameters from complex datasets, and increase the understanding of the bioprocesses occurring in engineered and natural environments.     

保护区周边居民薪柴采集对乔木分布的影响——以四川王朗大熊猫自然保护区为例

保护区内周边地区的自然村庄中人们的日常生活对于村庄附近的森林群落结构有着重要的影响。通过分析乔木蓄积量和“乔木-灌木”复合系统中乔木的相对密度随着距离村落中心远近变化的函数关系,揭示当地居民薪柴采集活动对于乔木分布影响的规律。首先对于四川省平武县木座乡、木皮乡和白马乡里10个村庄进行随机入户农村家庭经济调查,掌握基本经济条件;然后对于调查结果进行聚类分析,选定小河、厄里和详述加3个典型村庄作为研究对象。按照距村落中心距离远近作梯度样带调查,计算出各个样带中的木材蓄积量和“乔木一灌木”复合系统中乔木的相对密度,利用数理统计方法进行模型拟合。拟合结果显示：（1）贫富差距对于乔木分布规律没有明显的影响,3个村落的数据分析结果类似。说明这3个村落的贫富差距还没有足以影响村落周围乔木分布。（2）木材蓄积量和“乔木．灌木”复合系统中乔木的相对密度与距村落中心位置距离的函数关系分别符合Logistic模型和Growth模型。在距离小于第1域值（3,4km）的时候,因变量随着自变量——距离的增加而缓慢增加。这说明这一区域内乔木分布受到薪柴采集活动影响很大,虽然也有所增加但是趋势不明显。当距离在第1域值（3～4km）至第2域值（7—8km）的时候,曲线切线斜率突然增大。这说明薪柴采集活动频率和强度迅速增加,因而导致乔木的蓄积量和密度都快速恢复。当距离继续增大的时候,因变量增长速率又逐渐回落,乔木分布也已经与原始林中的状态相似。     

基于狄利克雷多项式过程模型与K-means 结合的菌群分析

群体分型是一种有助于更好的理解人类身心健康等复杂生物学问题的有效方法,聚类是一种为了对样本分组来降低复杂性的定义肠型的方法,而传统K-means聚类算法的K值选取无法确定,本文在传统K-means聚类算法的基础上进行了改进,并公开数据集上进行了验证,实验表明改进算法能够解决K值选取无法确定的问题,且聚类结果的稳定性、准确性和聚类质量都得到显著提高。将改进后的模型运用于肠道菌群OTUs数据,发现不仅能够有效地区分2-型糖尿病患者样本间的相似性,而且能鉴定出影响菌群结构异质性最大的OTUs菌,为临床解决2-型糖尿病问题提供了一种新的思路。     

A Combinatorial approach: To design inhibitory molecules on Hemagglutinin protein of H1N1 virus (Swine Flu)

The Hemagglutinin (HA) is a protein of influenza A virus. It is present on the surface of influenza A virus and it is a glycoprotein. The HA is identified as potential drug target. H1N1 thiazolides, proved to be a potent drug in the inhibition of H1N1 replication. It is also known as inhibitor of other strains of influenza A virus. Thiazolide drug represses viral HA''s maturation at a level which exists just before the resistance from digestion of endoglycosidase-H and thereby it hampers, HA insertion in host membrane. Blocking the appropriate active site of hemagglutinin protein helps in the disease control. In the present work, we have generated diverse combinatorial library based ligands on known inhibitor thiazolides and they were used for virtual screening by Molegro virtual docker program. K-means clustering approach was used for finding new inhibitory molecules with more appropriate features. These resulted molecules are may be helpful in the treatment of swine flu and many other related diseases.     

Identification of new candidate drugs for lung cancer using chemical–chemical interactions,chemical–protein interactions and a K-means clustering algorithm

Lung cancer, characterized by uncontrolled cell growth in the lung tissue, is the leading cause of global cancer deaths. Until now, effective treatment of this disease is limited. Many synthetic compounds have emerged with the advancement of combinatorial chemistry. Identification of effective lung cancer candidate drug compounds among them is a great challenge. Thus, it is necessary to build effective computational methods that can assist us in selecting for potential lung cancer drug compounds. In this study, a computational method was proposed to tackle this problem. The chemical–chemical interactions and chemical–protein interactions were utilized to select candidate drug compounds that have close associations with approved lung cancer drugs and lung cancer-related genes. A permutation test and K-means clustering algorithm were employed to exclude candidate drugs with low possibilities to treat lung cancer. The final analysis suggests that the remaining drug compounds have potential anti-lung cancer activities and most of them have structural dissimilarity with approved drugs for lung cancer.